diff --git "a/checkpoint-2745/trainer_state.json" "b/checkpoint-2745/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-2745/trainer_state.json" @@ -0,0 +1,20761 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 2.701771653543307, + "eval_steps": 153, + "global_step": 2745, + "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": 10.4072, + "step": 1 + }, + { + "epoch": 0.001968503937007874, + "grad_norm": NaN, + "learning_rate": 0.0, + "loss": 11.0865, + "step": 2 + }, + { + "epoch": 0.002952755905511811, + "grad_norm": 60.69786071777344, + "learning_rate": 9.940357852882705e-10, + "loss": 9.5114, + "step": 3 + }, + { + "epoch": 0.003937007874015748, + "grad_norm": 59.147647857666016, + "learning_rate": 1.988071570576541e-09, + "loss": 9.9584, + "step": 4 + }, + { + "epoch": 0.004921259842519685, + "grad_norm": NaN, + "learning_rate": 1.988071570576541e-09, + "loss": 10.068, + "step": 5 + }, + { + "epoch": 0.005905511811023622, + "grad_norm": 65.82320404052734, + "learning_rate": 2.9821073558648116e-09, + "loss": 11.0224, + "step": 6 + }, + { + "epoch": 0.006889763779527559, + "grad_norm": 59.096107482910156, + "learning_rate": 3.976143141153082e-09, + "loss": 9.7703, + "step": 7 + }, + { + "epoch": 0.007874015748031496, + "grad_norm": 61.43330764770508, + "learning_rate": 4.970178926441353e-09, + "loss": 10.5005, + "step": 8 + }, + { + "epoch": 0.008858267716535433, + "grad_norm": 61.25212860107422, + "learning_rate": 5.964214711729623e-09, + "loss": 10.1987, + "step": 9 + }, + { + "epoch": 0.00984251968503937, + "grad_norm": 61.6600341796875, + "learning_rate": 6.9582504970178946e-09, + "loss": 10.0277, + "step": 10 + }, + { + "epoch": 0.010826771653543307, + "grad_norm": 61.71075439453125, + "learning_rate": 7.952286282306164e-09, + "loss": 10.6965, + "step": 11 + }, + { + "epoch": 0.011811023622047244, + "grad_norm": 59.34035873413086, + "learning_rate": 8.946322067594435e-09, + "loss": 10.0609, + "step": 12 + }, + { + "epoch": 0.012795275590551181, + "grad_norm": 64.14221954345703, + "learning_rate": 9.940357852882705e-09, + "loss": 11.6214, + "step": 13 + }, + { + "epoch": 0.013779527559055118, + "grad_norm": 57.682823181152344, + "learning_rate": 1.0934393638170978e-08, + "loss": 9.4053, + "step": 14 + }, + { + "epoch": 0.014763779527559055, + "grad_norm": 62.256858825683594, + "learning_rate": 1.1928429423459246e-08, + "loss": 10.4014, + "step": 15 + }, + { + "epoch": 0.015748031496062992, + "grad_norm": 64.56417083740234, + "learning_rate": 1.2922465208747517e-08, + "loss": 10.4119, + "step": 16 + }, + { + "epoch": 0.01673228346456693, + "grad_norm": 56.612098693847656, + "learning_rate": 1.3916500994035789e-08, + "loss": 9.4658, + "step": 17 + }, + { + "epoch": 0.017716535433070866, + "grad_norm": 60.02314758300781, + "learning_rate": 1.4910536779324056e-08, + "loss": 9.2169, + "step": 18 + }, + { + "epoch": 0.018700787401574805, + "grad_norm": 66.38670349121094, + "learning_rate": 1.590457256461233e-08, + "loss": 11.2337, + "step": 19 + }, + { + "epoch": 0.01968503937007874, + "grad_norm": 64.9875259399414, + "learning_rate": 1.68986083499006e-08, + "loss": 11.0572, + "step": 20 + }, + { + "epoch": 0.02066929133858268, + "grad_norm": 59.36448287963867, + "learning_rate": 1.789264413518887e-08, + "loss": 11.0452, + "step": 21 + }, + { + "epoch": 0.021653543307086614, + "grad_norm": 59.18537902832031, + "learning_rate": 1.888667992047714e-08, + "loss": 10.31, + "step": 22 + }, + { + "epoch": 0.022637795275590553, + "grad_norm": 54.053524017333984, + "learning_rate": 1.988071570576541e-08, + "loss": 9.1395, + "step": 23 + }, + { + "epoch": 0.023622047244094488, + "grad_norm": 49.979618072509766, + "learning_rate": 2.087475149105368e-08, + "loss": 8.4201, + "step": 24 + }, + { + "epoch": 0.024606299212598427, + "grad_norm": 51.654518127441406, + "learning_rate": 2.1868787276341955e-08, + "loss": 8.6036, + "step": 25 + }, + { + "epoch": 0.025590551181102362, + "grad_norm": 63.55083465576172, + "learning_rate": 2.2862823061630224e-08, + "loss": 11.7579, + "step": 26 + }, + { + "epoch": 0.0265748031496063, + "grad_norm": 59.30263137817383, + "learning_rate": 2.3856858846918493e-08, + "loss": 10.1307, + "step": 27 + }, + { + "epoch": 0.027559055118110236, + "grad_norm": 50.75270462036133, + "learning_rate": 2.4850894632206765e-08, + "loss": 9.2915, + "step": 28 + }, + { + "epoch": 0.028543307086614175, + "grad_norm": 51.7747917175293, + "learning_rate": 2.5844930417495034e-08, + "loss": 9.0208, + "step": 29 + }, + { + "epoch": 0.02952755905511811, + "grad_norm": 46.81666564941406, + "learning_rate": 2.6838966202783303e-08, + "loss": 8.6867, + "step": 30 + }, + { + "epoch": 0.03051181102362205, + "grad_norm": 44.82905578613281, + "learning_rate": 2.7833001988071578e-08, + "loss": 8.0925, + "step": 31 + }, + { + "epoch": 0.031496062992125984, + "grad_norm": 47.148406982421875, + "learning_rate": 2.8827037773359847e-08, + "loss": 8.6617, + "step": 32 + }, + { + "epoch": 0.03248031496062992, + "grad_norm": 47.153053283691406, + "learning_rate": 2.982107355864811e-08, + "loss": 8.3374, + "step": 33 + }, + { + "epoch": 0.03346456692913386, + "grad_norm": 45.4912223815918, + "learning_rate": 3.081510934393639e-08, + "loss": 7.8566, + "step": 34 + }, + { + "epoch": 0.0344488188976378, + "grad_norm": 51.92588806152344, + "learning_rate": 3.180914512922466e-08, + "loss": 9.0698, + "step": 35 + }, + { + "epoch": 0.03543307086614173, + "grad_norm": 43.18679428100586, + "learning_rate": 3.280318091451293e-08, + "loss": 7.7727, + "step": 36 + }, + { + "epoch": 0.03641732283464567, + "grad_norm": 43.12812805175781, + "learning_rate": 3.37972166998012e-08, + "loss": 7.6128, + "step": 37 + }, + { + "epoch": 0.03740157480314961, + "grad_norm": 44.7684211730957, + "learning_rate": 3.479125248508947e-08, + "loss": 7.8762, + "step": 38 + }, + { + "epoch": 0.038385826771653545, + "grad_norm": 36.34556198120117, + "learning_rate": 3.578528827037774e-08, + "loss": 7.5191, + "step": 39 + }, + { + "epoch": 0.03937007874015748, + "grad_norm": 39.9330940246582, + "learning_rate": 3.6779324055666005e-08, + "loss": 7.5638, + "step": 40 + }, + { + "epoch": 0.040354330708661415, + "grad_norm": 39.49993133544922, + "learning_rate": 3.777335984095428e-08, + "loss": 7.1878, + "step": 41 + }, + { + "epoch": 0.04133858267716536, + "grad_norm": 34.661251068115234, + "learning_rate": 3.8767395626242556e-08, + "loss": 6.8878, + "step": 42 + }, + { + "epoch": 0.04232283464566929, + "grad_norm": 40.655540466308594, + "learning_rate": 3.976143141153082e-08, + "loss": 7.5775, + "step": 43 + }, + { + "epoch": 0.04330708661417323, + "grad_norm": 36.12670135498047, + "learning_rate": 4.0755467196819094e-08, + "loss": 7.1076, + "step": 44 + }, + { + "epoch": 0.04429133858267716, + "grad_norm": 34.178226470947266, + "learning_rate": 4.174950298210736e-08, + "loss": 6.5589, + "step": 45 + }, + { + "epoch": 0.045275590551181105, + "grad_norm": 36.6577033996582, + "learning_rate": 4.274353876739563e-08, + "loss": 7.4456, + "step": 46 + }, + { + "epoch": 0.04625984251968504, + "grad_norm": 32.548709869384766, + "learning_rate": 4.373757455268391e-08, + "loss": 6.8233, + "step": 47 + }, + { + "epoch": 0.047244094488188976, + "grad_norm": 34.000553131103516, + "learning_rate": 4.4731610337972176e-08, + "loss": 6.7633, + "step": 48 + }, + { + "epoch": 0.04822834645669291, + "grad_norm": 32.247859954833984, + "learning_rate": 4.572564612326045e-08, + "loss": 6.6024, + "step": 49 + }, + { + "epoch": 0.04921259842519685, + "grad_norm": 28.947786331176758, + "learning_rate": 4.6719681908548713e-08, + "loss": 6.2778, + "step": 50 + }, + { + "epoch": 0.05019685039370079, + "grad_norm": 30.279062271118164, + "learning_rate": 4.7713717693836986e-08, + "loss": 6.1026, + "step": 51 + }, + { + "epoch": 0.051181102362204724, + "grad_norm": 31.13785171508789, + "learning_rate": 4.870775347912525e-08, + "loss": 6.632, + "step": 52 + }, + { + "epoch": 0.05216535433070866, + "grad_norm": 29.648237228393555, + "learning_rate": 4.970178926441353e-08, + "loss": 6.6962, + "step": 53 + }, + { + "epoch": 0.0531496062992126, + "grad_norm": 28.224645614624023, + "learning_rate": 5.06958250497018e-08, + "loss": 5.8514, + "step": 54 + }, + { + "epoch": 0.054133858267716536, + "grad_norm": 28.693328857421875, + "learning_rate": 5.168986083499007e-08, + "loss": 5.9951, + "step": 55 + }, + { + "epoch": 0.05511811023622047, + "grad_norm": 24.777812957763672, + "learning_rate": 5.268389662027834e-08, + "loss": 5.4554, + "step": 56 + }, + { + "epoch": 0.05610236220472441, + "grad_norm": 26.11226463317871, + "learning_rate": 5.3677932405566605e-08, + "loss": 6.0147, + "step": 57 + }, + { + "epoch": 0.05708661417322835, + "grad_norm": 24.698375701904297, + "learning_rate": 5.467196819085488e-08, + "loss": 5.215, + "step": 58 + }, + { + "epoch": 0.058070866141732284, + "grad_norm": 26.616317749023438, + "learning_rate": 5.5666003976143156e-08, + "loss": 6.4525, + "step": 59 + }, + { + "epoch": 0.05905511811023622, + "grad_norm": 26.09321403503418, + "learning_rate": 5.666003976143142e-08, + "loss": 5.4048, + "step": 60 + }, + { + "epoch": 0.060039370078740155, + "grad_norm": 19.83713150024414, + "learning_rate": 5.7654075546719694e-08, + "loss": 5.0424, + "step": 61 + }, + { + "epoch": 0.0610236220472441, + "grad_norm": 26.56923484802246, + "learning_rate": 5.864811133200796e-08, + "loss": 6.2646, + "step": 62 + }, + { + "epoch": 0.06200787401574803, + "grad_norm": 23.580089569091797, + "learning_rate": 5.964214711729623e-08, + "loss": 5.0847, + "step": 63 + }, + { + "epoch": 0.06299212598425197, + "grad_norm": 23.453126907348633, + "learning_rate": 6.06361829025845e-08, + "loss": 5.4415, + "step": 64 + }, + { + "epoch": 0.0639763779527559, + "grad_norm": 21.229190826416016, + "learning_rate": 6.163021868787278e-08, + "loss": 5.2469, + "step": 65 + }, + { + "epoch": 0.06496062992125984, + "grad_norm": 19.477190017700195, + "learning_rate": 6.262425447316104e-08, + "loss": 5.1378, + "step": 66 + }, + { + "epoch": 0.06594488188976377, + "grad_norm": 19.40647315979004, + "learning_rate": 6.361829025844931e-08, + "loss": 5.1636, + "step": 67 + }, + { + "epoch": 0.06692913385826772, + "grad_norm": 22.20977210998535, + "learning_rate": 6.461232604373759e-08, + "loss": 5.5596, + "step": 68 + }, + { + "epoch": 0.06791338582677166, + "grad_norm": 19.186826705932617, + "learning_rate": 6.560636182902586e-08, + "loss": 4.9508, + "step": 69 + }, + { + "epoch": 0.0688976377952756, + "grad_norm": 20.190908432006836, + "learning_rate": 6.660039761431412e-08, + "loss": 5.2355, + "step": 70 + }, + { + "epoch": 0.06988188976377953, + "grad_norm": 18.122196197509766, + "learning_rate": 6.75944333996024e-08, + "loss": 4.7359, + "step": 71 + }, + { + "epoch": 0.07086614173228346, + "grad_norm": 17.524765014648438, + "learning_rate": 6.858846918489067e-08, + "loss": 4.8947, + "step": 72 + }, + { + "epoch": 0.0718503937007874, + "grad_norm": 18.821767807006836, + "learning_rate": 6.958250497017893e-08, + "loss": 4.6269, + "step": 73 + }, + { + "epoch": 0.07283464566929133, + "grad_norm": 18.19922637939453, + "learning_rate": 7.057654075546721e-08, + "loss": 4.6072, + "step": 74 + }, + { + "epoch": 0.07381889763779527, + "grad_norm": 16.908899307250977, + "learning_rate": 7.157057654075548e-08, + "loss": 4.9125, + "step": 75 + }, + { + "epoch": 0.07480314960629922, + "grad_norm": 19.90263557434082, + "learning_rate": 7.256461232604374e-08, + "loss": 4.5856, + "step": 76 + }, + { + "epoch": 0.07578740157480315, + "grad_norm": 17.92584800720215, + "learning_rate": 7.355864811133201e-08, + "loss": 4.7879, + "step": 77 + }, + { + "epoch": 0.07677165354330709, + "grad_norm": 16.29261589050293, + "learning_rate": 7.455268389662029e-08, + "loss": 4.5348, + "step": 78 + }, + { + "epoch": 0.07775590551181102, + "grad_norm": 16.3350887298584, + "learning_rate": 7.554671968190855e-08, + "loss": 4.3554, + "step": 79 + }, + { + "epoch": 0.07874015748031496, + "grad_norm": 14.408184051513672, + "learning_rate": 7.654075546719683e-08, + "loss": 4.2984, + "step": 80 + }, + { + "epoch": 0.0797244094488189, + "grad_norm": 16.71326446533203, + "learning_rate": 7.753479125248511e-08, + "loss": 4.5505, + "step": 81 + }, + { + "epoch": 0.08070866141732283, + "grad_norm": 14.62590217590332, + "learning_rate": 7.852882703777338e-08, + "loss": 4.5325, + "step": 82 + }, + { + "epoch": 0.08169291338582677, + "grad_norm": 17.189268112182617, + "learning_rate": 7.952286282306164e-08, + "loss": 4.2725, + "step": 83 + }, + { + "epoch": 0.08267716535433071, + "grad_norm": 16.960248947143555, + "learning_rate": 8.051689860834992e-08, + "loss": 4.3054, + "step": 84 + }, + { + "epoch": 0.08366141732283465, + "grad_norm": 15.114398956298828, + "learning_rate": 8.151093439363819e-08, + "loss": 4.5536, + "step": 85 + }, + { + "epoch": 0.08464566929133858, + "grad_norm": 16.153371810913086, + "learning_rate": 8.250497017892645e-08, + "loss": 4.0265, + "step": 86 + }, + { + "epoch": 0.08562992125984252, + "grad_norm": 15.731820106506348, + "learning_rate": 8.349900596421472e-08, + "loss": 4.7453, + "step": 87 + }, + { + "epoch": 0.08661417322834646, + "grad_norm": 14.69382381439209, + "learning_rate": 8.4493041749503e-08, + "loss": 4.071, + "step": 88 + }, + { + "epoch": 0.08759842519685039, + "grad_norm": 13.735575675964355, + "learning_rate": 8.548707753479126e-08, + "loss": 4.1582, + "step": 89 + }, + { + "epoch": 0.08858267716535433, + "grad_norm": 16.017065048217773, + "learning_rate": 8.648111332007953e-08, + "loss": 4.1131, + "step": 90 + }, + { + "epoch": 0.08956692913385826, + "grad_norm": 17.237276077270508, + "learning_rate": 8.747514910536782e-08, + "loss": 3.6582, + "step": 91 + }, + { + "epoch": 0.09055118110236221, + "grad_norm": 15.59334945678711, + "learning_rate": 8.846918489065609e-08, + "loss": 4.143, + "step": 92 + }, + { + "epoch": 0.09153543307086615, + "grad_norm": 14.918270111083984, + "learning_rate": 8.946322067594435e-08, + "loss": 4.2273, + "step": 93 + }, + { + "epoch": 0.09251968503937008, + "grad_norm": 14.899909019470215, + "learning_rate": 9.045725646123262e-08, + "loss": 3.9321, + "step": 94 + }, + { + "epoch": 0.09350393700787402, + "grad_norm": 18.112892150878906, + "learning_rate": 9.14512922465209e-08, + "loss": 4.2061, + "step": 95 + }, + { + "epoch": 0.09448818897637795, + "grad_norm": 15.854629516601562, + "learning_rate": 9.244532803180916e-08, + "loss": 4.1042, + "step": 96 + }, + { + "epoch": 0.09547244094488189, + "grad_norm": 16.44801139831543, + "learning_rate": 9.343936381709743e-08, + "loss": 3.9513, + "step": 97 + }, + { + "epoch": 0.09645669291338582, + "grad_norm": 14.854127883911133, + "learning_rate": 9.44333996023857e-08, + "loss": 3.8627, + "step": 98 + }, + { + "epoch": 0.09744094488188976, + "grad_norm": 17.02035903930664, + "learning_rate": 9.542743538767397e-08, + "loss": 4.3613, + "step": 99 + }, + { + "epoch": 0.0984251968503937, + "grad_norm": 15.354338645935059, + "learning_rate": 9.642147117296224e-08, + "loss": 3.8513, + "step": 100 + }, + { + "epoch": 0.09940944881889764, + "grad_norm": 16.35565757751465, + "learning_rate": 9.74155069582505e-08, + "loss": 3.5866, + "step": 101 + }, + { + "epoch": 0.10039370078740158, + "grad_norm": 16.66194725036621, + "learning_rate": 9.840954274353878e-08, + "loss": 3.5239, + "step": 102 + }, + { + "epoch": 0.10137795275590551, + "grad_norm": 15.875446319580078, + "learning_rate": 9.940357852882706e-08, + "loss": 3.5921, + "step": 103 + }, + { + "epoch": 0.10236220472440945, + "grad_norm": 14.344114303588867, + "learning_rate": 1.0039761431411533e-07, + "loss": 3.5962, + "step": 104 + }, + { + "epoch": 0.10334645669291338, + "grad_norm": 18.503963470458984, + "learning_rate": 1.013916500994036e-07, + "loss": 4.0001, + "step": 105 + }, + { + "epoch": 0.10433070866141732, + "grad_norm": 16.944435119628906, + "learning_rate": 1.0238568588469187e-07, + "loss": 4.1374, + "step": 106 + }, + { + "epoch": 0.10531496062992125, + "grad_norm": 16.46833038330078, + "learning_rate": 1.0337972166998014e-07, + "loss": 3.9049, + "step": 107 + }, + { + "epoch": 0.1062992125984252, + "grad_norm": 14.921700477600098, + "learning_rate": 1.043737574552684e-07, + "loss": 3.2511, + "step": 108 + }, + { + "epoch": 0.10728346456692914, + "grad_norm": 15.574972152709961, + "learning_rate": 1.0536779324055668e-07, + "loss": 3.2479, + "step": 109 + }, + { + "epoch": 0.10826771653543307, + "grad_norm": 16.810884475708008, + "learning_rate": 1.0636182902584495e-07, + "loss": 3.6414, + "step": 110 + }, + { + "epoch": 0.10925196850393701, + "grad_norm": 17.074661254882812, + "learning_rate": 1.0735586481113321e-07, + "loss": 3.6429, + "step": 111 + }, + { + "epoch": 0.11023622047244094, + "grad_norm": 18.52947235107422, + "learning_rate": 1.0834990059642149e-07, + "loss": 3.423, + "step": 112 + }, + { + "epoch": 0.11122047244094488, + "grad_norm": 18.681869506835938, + "learning_rate": 1.0934393638170976e-07, + "loss": 3.4967, + "step": 113 + }, + { + "epoch": 0.11220472440944881, + "grad_norm": 20.385292053222656, + "learning_rate": 1.1033797216699802e-07, + "loss": 3.7649, + "step": 114 + }, + { + "epoch": 0.11318897637795275, + "grad_norm": 18.912538528442383, + "learning_rate": 1.1133200795228631e-07, + "loss": 3.2845, + "step": 115 + }, + { + "epoch": 0.1141732283464567, + "grad_norm": 17.856229782104492, + "learning_rate": 1.1232604373757458e-07, + "loss": 3.356, + "step": 116 + }, + { + "epoch": 0.11515748031496063, + "grad_norm": 17.08562469482422, + "learning_rate": 1.1332007952286284e-07, + "loss": 3.2086, + "step": 117 + }, + { + "epoch": 0.11614173228346457, + "grad_norm": 17.54237937927246, + "learning_rate": 1.1431411530815111e-07, + "loss": 3.5561, + "step": 118 + }, + { + "epoch": 0.1171259842519685, + "grad_norm": 19.936498641967773, + "learning_rate": 1.1530815109343939e-07, + "loss": 3.7353, + "step": 119 + }, + { + "epoch": 0.11811023622047244, + "grad_norm": 17.135496139526367, + "learning_rate": 1.1630218687872765e-07, + "loss": 3.403, + "step": 120 + }, + { + "epoch": 0.11909448818897637, + "grad_norm": 17.260093688964844, + "learning_rate": 1.1729622266401592e-07, + "loss": 3.1009, + "step": 121 + }, + { + "epoch": 0.12007874015748031, + "grad_norm": 17.075611114501953, + "learning_rate": 1.182902584493042e-07, + "loss": 3.2139, + "step": 122 + }, + { + "epoch": 0.12106299212598425, + "grad_norm": 23.433874130249023, + "learning_rate": 1.1928429423459245e-07, + "loss": 3.3339, + "step": 123 + }, + { + "epoch": 0.1220472440944882, + "grad_norm": 18.25501251220703, + "learning_rate": 1.2027833001988073e-07, + "loss": 2.9464, + "step": 124 + }, + { + "epoch": 0.12303149606299213, + "grad_norm": 18.079578399658203, + "learning_rate": 1.21272365805169e-07, + "loss": 3.3366, + "step": 125 + }, + { + "epoch": 0.12401574803149606, + "grad_norm": 16.392736434936523, + "learning_rate": 1.222664015904573e-07, + "loss": 3.0618, + "step": 126 + }, + { + "epoch": 0.125, + "grad_norm": 15.782499313354492, + "learning_rate": 1.2326043737574557e-07, + "loss": 3.0092, + "step": 127 + }, + { + "epoch": 0.12598425196850394, + "grad_norm": 14.74819278717041, + "learning_rate": 1.2425447316103382e-07, + "loss": 2.7152, + "step": 128 + }, + { + "epoch": 0.12696850393700787, + "grad_norm": 17.743946075439453, + "learning_rate": 1.2524850894632207e-07, + "loss": 2.9423, + "step": 129 + }, + { + "epoch": 0.1279527559055118, + "grad_norm": 15.759135246276855, + "learning_rate": 1.2624254473161035e-07, + "loss": 2.6569, + "step": 130 + }, + { + "epoch": 0.12893700787401574, + "grad_norm": 18.54253387451172, + "learning_rate": 1.2723658051689863e-07, + "loss": 2.8469, + "step": 131 + }, + { + "epoch": 0.12992125984251968, + "grad_norm": 18.318775177001953, + "learning_rate": 1.282306163021869e-07, + "loss": 2.9089, + "step": 132 + }, + { + "epoch": 0.1309055118110236, + "grad_norm": 17.751266479492188, + "learning_rate": 1.2922465208747519e-07, + "loss": 2.5809, + "step": 133 + }, + { + "epoch": 0.13188976377952755, + "grad_norm": 21.29975128173828, + "learning_rate": 1.3021868787276344e-07, + "loss": 2.6987, + "step": 134 + }, + { + "epoch": 0.1328740157480315, + "grad_norm": 19.9519100189209, + "learning_rate": 1.3121272365805172e-07, + "loss": 3.2518, + "step": 135 + }, + { + "epoch": 0.13385826771653545, + "grad_norm": 19.99887466430664, + "learning_rate": 1.3220675944333997e-07, + "loss": 2.9145, + "step": 136 + }, + { + "epoch": 0.13484251968503938, + "grad_norm": 17.0355167388916, + "learning_rate": 1.3320079522862825e-07, + "loss": 2.4809, + "step": 137 + }, + { + "epoch": 0.13582677165354332, + "grad_norm": 18.628467559814453, + "learning_rate": 1.3419483101391653e-07, + "loss": 2.8264, + "step": 138 + }, + { + "epoch": 0.13681102362204725, + "grad_norm": 19.511394500732422, + "learning_rate": 1.351888667992048e-07, + "loss": 2.5724, + "step": 139 + }, + { + "epoch": 0.1377952755905512, + "grad_norm": 21.264480590820312, + "learning_rate": 1.3618290258449306e-07, + "loss": 2.6949, + "step": 140 + }, + { + "epoch": 0.13877952755905512, + "grad_norm": 19.76134490966797, + "learning_rate": 1.3717693836978134e-07, + "loss": 2.6925, + "step": 141 + }, + { + "epoch": 0.13976377952755906, + "grad_norm": 20.930673599243164, + "learning_rate": 1.381709741550696e-07, + "loss": 2.9311, + "step": 142 + }, + { + "epoch": 0.140748031496063, + "grad_norm": 21.966018676757812, + "learning_rate": 1.3916500994035787e-07, + "loss": 2.5667, + "step": 143 + }, + { + "epoch": 0.14173228346456693, + "grad_norm": 21.916505813598633, + "learning_rate": 1.4015904572564615e-07, + "loss": 3.2471, + "step": 144 + }, + { + "epoch": 0.14271653543307086, + "grad_norm": 20.081771850585938, + "learning_rate": 1.4115308151093443e-07, + "loss": 2.2441, + "step": 145 + }, + { + "epoch": 0.1437007874015748, + "grad_norm": 22.893489837646484, + "learning_rate": 1.421471172962227e-07, + "loss": 2.75, + "step": 146 + }, + { + "epoch": 0.14468503937007873, + "grad_norm": 23.95358657836914, + "learning_rate": 1.4314115308151096e-07, + "loss": 2.9669, + "step": 147 + }, + { + "epoch": 0.14566929133858267, + "grad_norm": 21.101062774658203, + "learning_rate": 1.4413518886679924e-07, + "loss": 2.736, + "step": 148 + }, + { + "epoch": 0.1466535433070866, + "grad_norm": 25.240341186523438, + "learning_rate": 1.451292246520875e-07, + "loss": 3.104, + "step": 149 + }, + { + "epoch": 0.14763779527559054, + "grad_norm": 18.358688354492188, + "learning_rate": 1.4612326043737577e-07, + "loss": 2.2175, + "step": 150 + }, + { + "epoch": 0.1486220472440945, + "grad_norm": 21.986661911010742, + "learning_rate": 1.4711729622266402e-07, + "loss": 2.7415, + "step": 151 + }, + { + "epoch": 0.14960629921259844, + "grad_norm": 20.64093017578125, + "learning_rate": 1.4811133200795232e-07, + "loss": 1.8707, + "step": 152 + }, + { + "epoch": 0.15059055118110237, + "grad_norm": 20.602142333984375, + "learning_rate": 1.4910536779324058e-07, + "loss": 2.5961, + "step": 153 + }, + { + "epoch": 0.15059055118110237, + "eval_Qnli-dev_cosine_accuracy": 0.634765625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8508153557777405, + "eval_Qnli-dev_cosine_ap": 0.6461335447626624, + "eval_Qnli-dev_cosine_f1": 0.6505636070853462, + "eval_Qnli-dev_cosine_f1_threshold": 0.7770615816116333, + "eval_Qnli-dev_cosine_precision": 0.5246753246753246, + "eval_Qnli-dev_cosine_recall": 0.8559322033898306, + "eval_Qnli-dev_dot_accuracy": 0.634765625, + "eval_Qnli-dev_dot_accuracy_threshold": 653.7443237304688, + "eval_Qnli-dev_dot_ap": 0.6461682282377894, + "eval_Qnli-dev_dot_f1": 0.6505636070853462, + "eval_Qnli-dev_dot_f1_threshold": 597.0731811523438, + "eval_Qnli-dev_dot_precision": 0.5246753246753246, + "eval_Qnli-dev_dot_recall": 0.8559322033898306, + "eval_Qnli-dev_euclidean_accuracy": 0.634765625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.141305923461914, + "eval_Qnli-dev_euclidean_ap": 0.6461382925406688, + "eval_Qnli-dev_euclidean_f1": 0.6505636070853462, + "eval_Qnli-dev_euclidean_f1_threshold": 18.50943946838379, + "eval_Qnli-dev_euclidean_precision": 0.5246753246753246, + "eval_Qnli-dev_euclidean_recall": 0.8559322033898306, + "eval_Qnli-dev_manhattan_accuracy": 0.6328125, + "eval_Qnli-dev_manhattan_accuracy_threshold": 331.46282958984375, + "eval_Qnli-dev_manhattan_ap": 0.6431949026371255, + "eval_Qnli-dev_manhattan_f1": 0.6501650165016502, + "eval_Qnli-dev_manhattan_f1_threshold": 404.6050109863281, + "eval_Qnli-dev_manhattan_precision": 0.5324324324324324, + "eval_Qnli-dev_manhattan_recall": 0.8347457627118644, + "eval_Qnli-dev_max_accuracy": 0.634765625, + "eval_Qnli-dev_max_accuracy_threshold": 653.7443237304688, + "eval_Qnli-dev_max_ap": 0.6461682282377894, + "eval_Qnli-dev_max_f1": 0.6505636070853462, + "eval_Qnli-dev_max_f1_threshold": 597.0731811523438, + "eval_Qnli-dev_max_precision": 0.5324324324324324, + "eval_Qnli-dev_max_recall": 0.8559322033898306, + "eval_allNLI-dev_cosine_accuracy": 0.67578125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9452645182609558, + "eval_allNLI-dev_cosine_ap": 0.4264736612515921, + "eval_allNLI-dev_cosine_f1": 0.512, + "eval_allNLI-dev_cosine_f1_threshold": 0.8565204739570618, + "eval_allNLI-dev_cosine_precision": 0.39143730886850153, + "eval_allNLI-dev_cosine_recall": 0.7398843930635838, + "eval_allNLI-dev_dot_accuracy": 0.67578125, + "eval_allNLI-dev_dot_accuracy_threshold": 726.30615234375, + "eval_allNLI-dev_dot_ap": 0.42647535250956575, + "eval_allNLI-dev_dot_f1": 0.512, + "eval_allNLI-dev_dot_f1_threshold": 658.1103515625, + "eval_allNLI-dev_dot_precision": 0.39143730886850153, + "eval_allNLI-dev_dot_recall": 0.7398843930635838, + "eval_allNLI-dev_euclidean_accuracy": 0.67578125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.171283721923828, + "eval_allNLI-dev_euclidean_ap": 0.4264736612515921, + "eval_allNLI-dev_euclidean_f1": 0.512, + "eval_allNLI-dev_euclidean_f1_threshold": 14.84876823425293, + "eval_allNLI-dev_euclidean_precision": 0.39143730886850153, + "eval_allNLI-dev_euclidean_recall": 0.7398843930635838, + "eval_allNLI-dev_manhattan_accuracy": 0.67578125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 201.49061584472656, + "eval_allNLI-dev_manhattan_ap": 0.4252213828672732, + "eval_allNLI-dev_manhattan_f1": 0.5107692307692308, + "eval_allNLI-dev_manhattan_f1_threshold": 417.52728271484375, + "eval_allNLI-dev_manhattan_precision": 0.3480083857442348, + "eval_allNLI-dev_manhattan_recall": 0.9595375722543352, + "eval_allNLI-dev_max_accuracy": 0.67578125, + "eval_allNLI-dev_max_accuracy_threshold": 726.30615234375, + "eval_allNLI-dev_max_ap": 0.42647535250956575, + "eval_allNLI-dev_max_f1": 0.512, + "eval_allNLI-dev_max_f1_threshold": 658.1103515625, + "eval_allNLI-dev_max_precision": 0.39143730886850153, + "eval_allNLI-dev_max_recall": 0.9595375722543352, + "eval_loss": 2.2652623653411865, + "eval_runtime": 50.7627, + "eval_samples_per_second": 32.78, + "eval_sequential_score": 0.6461682282377894, + "eval_steps_per_second": 0.138, + "eval_sts-test_pearson_cosine": 0.2749904272806095, + "eval_sts-test_pearson_dot": 0.27496363262371837, + "eval_sts-test_pearson_euclidean": 0.2934483033082174, + "eval_sts-test_pearson_manhattan": 0.2923996087310511, + "eval_sts-test_pearson_max": 0.2934483033082174, + "eval_sts-test_spearman_cosine": 0.31159390381099095, + "eval_sts-test_spearman_dot": 0.31138581044552094, + "eval_sts-test_spearman_euclidean": 0.3115817314678925, + "eval_sts-test_spearman_manhattan": 0.3095556181083969, + "eval_sts-test_spearman_max": 0.31159390381099095, + "step": 153 + }, + { + "epoch": 0.1515748031496063, + "grad_norm": 22.330442428588867, + "learning_rate": 1.5009940357852886e-07, + "loss": 3.1149, + "step": 154 + }, + { + "epoch": 0.15255905511811024, + "grad_norm": 23.656953811645508, + "learning_rate": 1.510934393638171e-07, + "loss": 2.2976, + "step": 155 + }, + { + "epoch": 0.15354330708661418, + "grad_norm": 20.271608352661133, + "learning_rate": 1.5208747514910539e-07, + "loss": 2.4436, + "step": 156 + }, + { + "epoch": 0.1545275590551181, + "grad_norm": 25.410293579101562, + "learning_rate": 1.5308151093439367e-07, + "loss": 2.8826, + "step": 157 + }, + { + "epoch": 0.15551181102362205, + "grad_norm": 23.772783279418945, + "learning_rate": 1.5407554671968192e-07, + "loss": 2.3664, + "step": 158 + }, + { + "epoch": 0.15649606299212598, + "grad_norm": 23.44937515258789, + "learning_rate": 1.5506958250497022e-07, + "loss": 2.2485, + "step": 159 + }, + { + "epoch": 0.15748031496062992, + "grad_norm": 23.024261474609375, + "learning_rate": 1.5606361829025848e-07, + "loss": 2.5167, + "step": 160 + }, + { + "epoch": 0.15846456692913385, + "grad_norm": 20.63090705871582, + "learning_rate": 1.5705765407554675e-07, + "loss": 1.7183, + "step": 161 + }, + { + "epoch": 0.1594488188976378, + "grad_norm": 21.573190689086914, + "learning_rate": 1.58051689860835e-07, + "loss": 2.1003, + "step": 162 + }, + { + "epoch": 0.16043307086614172, + "grad_norm": 23.774974822998047, + "learning_rate": 1.5904572564612329e-07, + "loss": 2.5785, + "step": 163 + }, + { + "epoch": 0.16141732283464566, + "grad_norm": 27.151123046875, + "learning_rate": 1.6003976143141154e-07, + "loss": 2.8789, + "step": 164 + }, + { + "epoch": 0.1624015748031496, + "grad_norm": 23.50958824157715, + "learning_rate": 1.6103379721669984e-07, + "loss": 2.3425, + "step": 165 + }, + { + "epoch": 0.16338582677165353, + "grad_norm": 23.77661895751953, + "learning_rate": 1.620278330019881e-07, + "loss": 2.0966, + "step": 166 + }, + { + "epoch": 0.1643700787401575, + "grad_norm": 22.070526123046875, + "learning_rate": 1.6302186878727637e-07, + "loss": 2.1126, + "step": 167 + }, + { + "epoch": 0.16535433070866143, + "grad_norm": 22.653602600097656, + "learning_rate": 1.6401590457256465e-07, + "loss": 2.1824, + "step": 168 + }, + { + "epoch": 0.16633858267716536, + "grad_norm": 21.470808029174805, + "learning_rate": 1.650099403578529e-07, + "loss": 2.2009, + "step": 169 + }, + { + "epoch": 0.1673228346456693, + "grad_norm": 25.822694778442383, + "learning_rate": 1.6600397614314118e-07, + "loss": 2.3796, + "step": 170 + }, + { + "epoch": 0.16830708661417323, + "grad_norm": 22.609458923339844, + "learning_rate": 1.6699801192842944e-07, + "loss": 2.3096, + "step": 171 + }, + { + "epoch": 0.16929133858267717, + "grad_norm": 24.10075569152832, + "learning_rate": 1.6799204771371774e-07, + "loss": 2.7897, + "step": 172 + }, + { + "epoch": 0.1702755905511811, + "grad_norm": 22.21641731262207, + "learning_rate": 1.68986083499006e-07, + "loss": 2.2097, + "step": 173 + }, + { + "epoch": 0.17125984251968504, + "grad_norm": 17.717933654785156, + "learning_rate": 1.6998011928429427e-07, + "loss": 1.7508, + "step": 174 + }, + { + "epoch": 0.17224409448818898, + "grad_norm": 22.352798461914062, + "learning_rate": 1.7097415506958253e-07, + "loss": 2.353, + "step": 175 + }, + { + "epoch": 0.1732283464566929, + "grad_norm": 23.421472549438477, + "learning_rate": 1.719681908548708e-07, + "loss": 2.4276, + "step": 176 + }, + { + "epoch": 0.17421259842519685, + "grad_norm": 20.41706657409668, + "learning_rate": 1.7296222664015906e-07, + "loss": 2.1016, + "step": 177 + }, + { + "epoch": 0.17519685039370078, + "grad_norm": 19.39253807067871, + "learning_rate": 1.7395626242544734e-07, + "loss": 1.8461, + "step": 178 + }, + { + "epoch": 0.17618110236220472, + "grad_norm": 19.994935989379883, + "learning_rate": 1.7495029821073564e-07, + "loss": 1.8104, + "step": 179 + }, + { + "epoch": 0.17716535433070865, + "grad_norm": 24.119571685791016, + "learning_rate": 1.759443339960239e-07, + "loss": 2.6023, + "step": 180 + }, + { + "epoch": 0.1781496062992126, + "grad_norm": 28.111419677734375, + "learning_rate": 1.7693836978131217e-07, + "loss": 2.5261, + "step": 181 + }, + { + "epoch": 0.17913385826771652, + "grad_norm": 26.244142532348633, + "learning_rate": 1.7793240556660042e-07, + "loss": 2.1053, + "step": 182 + }, + { + "epoch": 0.18011811023622049, + "grad_norm": 24.29507827758789, + "learning_rate": 1.789264413518887e-07, + "loss": 1.9712, + "step": 183 + }, + { + "epoch": 0.18110236220472442, + "grad_norm": 26.753236770629883, + "learning_rate": 1.7992047713717695e-07, + "loss": 2.4693, + "step": 184 + }, + { + "epoch": 0.18208661417322836, + "grad_norm": 23.031953811645508, + "learning_rate": 1.8091451292246523e-07, + "loss": 2.1119, + "step": 185 + }, + { + "epoch": 0.1830708661417323, + "grad_norm": 24.926044464111328, + "learning_rate": 1.819085487077535e-07, + "loss": 2.4797, + "step": 186 + }, + { + "epoch": 0.18405511811023623, + "grad_norm": 19.047605514526367, + "learning_rate": 1.829025844930418e-07, + "loss": 2.1587, + "step": 187 + }, + { + "epoch": 0.18503937007874016, + "grad_norm": 26.785459518432617, + "learning_rate": 1.8389662027833004e-07, + "loss": 1.9578, + "step": 188 + }, + { + "epoch": 0.1860236220472441, + "grad_norm": 22.257556915283203, + "learning_rate": 1.8489065606361832e-07, + "loss": 2.1368, + "step": 189 + }, + { + "epoch": 0.18700787401574803, + "grad_norm": 24.006427764892578, + "learning_rate": 1.8588469184890657e-07, + "loss": 2.4212, + "step": 190 + }, + { + "epoch": 0.18799212598425197, + "grad_norm": 22.14805793762207, + "learning_rate": 1.8687872763419485e-07, + "loss": 1.9591, + "step": 191 + }, + { + "epoch": 0.1889763779527559, + "grad_norm": 19.438581466674805, + "learning_rate": 1.8787276341948313e-07, + "loss": 1.5816, + "step": 192 + }, + { + "epoch": 0.18996062992125984, + "grad_norm": 19.473068237304688, + "learning_rate": 1.888667992047714e-07, + "loss": 1.4029, + "step": 193 + }, + { + "epoch": 0.19094488188976377, + "grad_norm": 22.895261764526367, + "learning_rate": 1.898608349900597e-07, + "loss": 1.9385, + "step": 194 + }, + { + "epoch": 0.1919291338582677, + "grad_norm": 22.117504119873047, + "learning_rate": 1.9085487077534794e-07, + "loss": 1.5596, + "step": 195 + }, + { + "epoch": 0.19291338582677164, + "grad_norm": 25.059682846069336, + "learning_rate": 1.9184890656063622e-07, + "loss": 1.6663, + "step": 196 + }, + { + "epoch": 0.19389763779527558, + "grad_norm": 31.338993072509766, + "learning_rate": 1.9284294234592447e-07, + "loss": 2.0026, + "step": 197 + }, + { + "epoch": 0.19488188976377951, + "grad_norm": 25.27398681640625, + "learning_rate": 1.9383697813121275e-07, + "loss": 2.0046, + "step": 198 + }, + { + "epoch": 0.19586614173228348, + "grad_norm": 22.62946128845215, + "learning_rate": 1.94831013916501e-07, + "loss": 1.5016, + "step": 199 + }, + { + "epoch": 0.1968503937007874, + "grad_norm": 26.856422424316406, + "learning_rate": 1.958250497017893e-07, + "loss": 2.184, + "step": 200 + }, + { + "epoch": 0.19783464566929135, + "grad_norm": 27.561872482299805, + "learning_rate": 1.9681908548707756e-07, + "loss": 2.3442, + "step": 201 + }, + { + "epoch": 0.19881889763779528, + "grad_norm": 29.985654830932617, + "learning_rate": 1.9781312127236584e-07, + "loss": 2.6981, + "step": 202 + }, + { + "epoch": 0.19980314960629922, + "grad_norm": 22.171070098876953, + "learning_rate": 1.9880715705765412e-07, + "loss": 2.5481, + "step": 203 + }, + { + "epoch": 0.20078740157480315, + "grad_norm": 26.214073181152344, + "learning_rate": 1.9980119284294237e-07, + "loss": 2.9798, + "step": 204 + }, + { + "epoch": 0.2017716535433071, + "grad_norm": 22.690643310546875, + "learning_rate": 2.0079522862823065e-07, + "loss": 2.287, + "step": 205 + }, + { + "epoch": 0.20275590551181102, + "grad_norm": 21.528181076049805, + "learning_rate": 2.017892644135189e-07, + "loss": 1.9393, + "step": 206 + }, + { + "epoch": 0.20374015748031496, + "grad_norm": 24.61495018005371, + "learning_rate": 2.027833001988072e-07, + "loss": 2.892, + "step": 207 + }, + { + "epoch": 0.2047244094488189, + "grad_norm": 21.661781311035156, + "learning_rate": 2.0377733598409546e-07, + "loss": 2.26, + "step": 208 + }, + { + "epoch": 0.20570866141732283, + "grad_norm": 21.17841339111328, + "learning_rate": 2.0477137176938374e-07, + "loss": 2.5911, + "step": 209 + }, + { + "epoch": 0.20669291338582677, + "grad_norm": 22.297616958618164, + "learning_rate": 2.05765407554672e-07, + "loss": 2.1239, + "step": 210 + }, + { + "epoch": 0.2076771653543307, + "grad_norm": 18.447376251220703, + "learning_rate": 2.0675944333996027e-07, + "loss": 2.0683, + "step": 211 + }, + { + "epoch": 0.20866141732283464, + "grad_norm": 19.692792892456055, + "learning_rate": 2.0775347912524852e-07, + "loss": 1.768, + "step": 212 + }, + { + "epoch": 0.20964566929133857, + "grad_norm": 24.793012619018555, + "learning_rate": 2.087475149105368e-07, + "loss": 2.5468, + "step": 213 + }, + { + "epoch": 0.2106299212598425, + "grad_norm": 21.60909652709961, + "learning_rate": 2.097415506958251e-07, + "loss": 1.8956, + "step": 214 + }, + { + "epoch": 0.21161417322834647, + "grad_norm": 22.114286422729492, + "learning_rate": 2.1073558648111336e-07, + "loss": 2.044, + "step": 215 + }, + { + "epoch": 0.2125984251968504, + "grad_norm": 19.02837562561035, + "learning_rate": 2.1172962226640164e-07, + "loss": 1.5721, + "step": 216 + }, + { + "epoch": 0.21358267716535434, + "grad_norm": 19.785751342773438, + "learning_rate": 2.127236580516899e-07, + "loss": 1.6278, + "step": 217 + }, + { + "epoch": 0.21456692913385828, + "grad_norm": 21.3282470703125, + "learning_rate": 2.1371769383697817e-07, + "loss": 1.7754, + "step": 218 + }, + { + "epoch": 0.2155511811023622, + "grad_norm": 25.80916404724121, + "learning_rate": 2.1471172962226642e-07, + "loss": 1.8594, + "step": 219 + }, + { + "epoch": 0.21653543307086615, + "grad_norm": 21.912315368652344, + "learning_rate": 2.1570576540755473e-07, + "loss": 1.8309, + "step": 220 + }, + { + "epoch": 0.21751968503937008, + "grad_norm": 24.051366806030273, + "learning_rate": 2.1669980119284298e-07, + "loss": 2.0619, + "step": 221 + }, + { + "epoch": 0.21850393700787402, + "grad_norm": 24.29237174987793, + "learning_rate": 2.1769383697813126e-07, + "loss": 2.3335, + "step": 222 + }, + { + "epoch": 0.21948818897637795, + "grad_norm": 25.850160598754883, + "learning_rate": 2.186878727634195e-07, + "loss": 2.023, + "step": 223 + }, + { + "epoch": 0.2204724409448819, + "grad_norm": 27.208112716674805, + "learning_rate": 2.196819085487078e-07, + "loss": 2.1975, + "step": 224 + }, + { + "epoch": 0.22145669291338582, + "grad_norm": 22.276878356933594, + "learning_rate": 2.2067594433399604e-07, + "loss": 1.9228, + "step": 225 + }, + { + "epoch": 0.22244094488188976, + "grad_norm": 30.213895797729492, + "learning_rate": 2.2166998011928432e-07, + "loss": 2.3565, + "step": 226 + }, + { + "epoch": 0.2234251968503937, + "grad_norm": 22.16749382019043, + "learning_rate": 2.2266401590457263e-07, + "loss": 1.896, + "step": 227 + }, + { + "epoch": 0.22440944881889763, + "grad_norm": 21.131744384765625, + "learning_rate": 2.2365805168986088e-07, + "loss": 2.0912, + "step": 228 + }, + { + "epoch": 0.22539370078740156, + "grad_norm": 25.1036376953125, + "learning_rate": 2.2465208747514916e-07, + "loss": 2.7703, + "step": 229 + }, + { + "epoch": 0.2263779527559055, + "grad_norm": 25.316814422607422, + "learning_rate": 2.256461232604374e-07, + "loss": 1.6988, + "step": 230 + }, + { + "epoch": 0.22736220472440946, + "grad_norm": 25.363996505737305, + "learning_rate": 2.266401590457257e-07, + "loss": 2.0406, + "step": 231 + }, + { + "epoch": 0.2283464566929134, + "grad_norm": 21.906835556030273, + "learning_rate": 2.2763419483101394e-07, + "loss": 1.9288, + "step": 232 + }, + { + "epoch": 0.22933070866141733, + "grad_norm": 21.407150268554688, + "learning_rate": 2.2862823061630222e-07, + "loss": 2.0457, + "step": 233 + }, + { + "epoch": 0.23031496062992127, + "grad_norm": 21.2374210357666, + "learning_rate": 2.296222664015905e-07, + "loss": 1.7061, + "step": 234 + }, + { + "epoch": 0.2312992125984252, + "grad_norm": 20.94179344177246, + "learning_rate": 2.3061630218687878e-07, + "loss": 1.6244, + "step": 235 + }, + { + "epoch": 0.23228346456692914, + "grad_norm": 21.845712661743164, + "learning_rate": 2.3161033797216703e-07, + "loss": 2.0241, + "step": 236 + }, + { + "epoch": 0.23326771653543307, + "grad_norm": 19.496191024780273, + "learning_rate": 2.326043737574553e-07, + "loss": 1.567, + "step": 237 + }, + { + "epoch": 0.234251968503937, + "grad_norm": 21.819353103637695, + "learning_rate": 2.335984095427436e-07, + "loss": 1.8084, + "step": 238 + }, + { + "epoch": 0.23523622047244094, + "grad_norm": 27.17051124572754, + "learning_rate": 2.3459244532803184e-07, + "loss": 2.4363, + "step": 239 + }, + { + "epoch": 0.23622047244094488, + "grad_norm": 24.850723266601562, + "learning_rate": 2.3558648111332012e-07, + "loss": 1.7532, + "step": 240 + }, + { + "epoch": 0.2372047244094488, + "grad_norm": 24.120052337646484, + "learning_rate": 2.365805168986084e-07, + "loss": 2.0797, + "step": 241 + }, + { + "epoch": 0.23818897637795275, + "grad_norm": 23.708179473876953, + "learning_rate": 2.3757455268389668e-07, + "loss": 1.9562, + "step": 242 + }, + { + "epoch": 0.23917322834645668, + "grad_norm": 20.58576774597168, + "learning_rate": 2.385685884691849e-07, + "loss": 1.6751, + "step": 243 + }, + { + "epoch": 0.24015748031496062, + "grad_norm": 25.161970138549805, + "learning_rate": 2.395626242544732e-07, + "loss": 2.0265, + "step": 244 + }, + { + "epoch": 0.24114173228346455, + "grad_norm": 22.079387664794922, + "learning_rate": 2.4055666003976146e-07, + "loss": 1.6065, + "step": 245 + }, + { + "epoch": 0.2421259842519685, + "grad_norm": 20.125717163085938, + "learning_rate": 2.4155069582504976e-07, + "loss": 1.7439, + "step": 246 + }, + { + "epoch": 0.24311023622047245, + "grad_norm": 22.855445861816406, + "learning_rate": 2.42544731610338e-07, + "loss": 2.0237, + "step": 247 + }, + { + "epoch": 0.2440944881889764, + "grad_norm": 22.16312026977539, + "learning_rate": 2.4353876739562627e-07, + "loss": 1.6128, + "step": 248 + }, + { + "epoch": 0.24507874015748032, + "grad_norm": 21.942440032958984, + "learning_rate": 2.445328031809146e-07, + "loss": 1.6581, + "step": 249 + }, + { + "epoch": 0.24606299212598426, + "grad_norm": 24.027729034423828, + "learning_rate": 2.4552683896620283e-07, + "loss": 2.1538, + "step": 250 + }, + { + "epoch": 0.2470472440944882, + "grad_norm": 23.1455078125, + "learning_rate": 2.4652087475149113e-07, + "loss": 2.049, + "step": 251 + }, + { + "epoch": 0.24803149606299213, + "grad_norm": 22.017520904541016, + "learning_rate": 2.475149105367794e-07, + "loss": 1.2573, + "step": 252 + }, + { + "epoch": 0.24901574803149606, + "grad_norm": 18.747554779052734, + "learning_rate": 2.4850894632206764e-07, + "loss": 1.5619, + "step": 253 + }, + { + "epoch": 0.25, + "grad_norm": 18.277063369750977, + "learning_rate": 2.495029821073559e-07, + "loss": 1.2611, + "step": 254 + }, + { + "epoch": 0.25098425196850394, + "grad_norm": 19.781038284301758, + "learning_rate": 2.5049701789264414e-07, + "loss": 1.3443, + "step": 255 + }, + { + "epoch": 0.25196850393700787, + "grad_norm": 21.605199813842773, + "learning_rate": 2.5149105367793245e-07, + "loss": 1.3436, + "step": 256 + }, + { + "epoch": 0.2529527559055118, + "grad_norm": 29.64748764038086, + "learning_rate": 2.524850894632207e-07, + "loss": 2.8117, + "step": 257 + }, + { + "epoch": 0.25393700787401574, + "grad_norm": 19.245962142944336, + "learning_rate": 2.53479125248509e-07, + "loss": 1.7563, + "step": 258 + }, + { + "epoch": 0.2549212598425197, + "grad_norm": 18.752405166625977, + "learning_rate": 2.5447316103379726e-07, + "loss": 1.3148, + "step": 259 + }, + { + "epoch": 0.2559055118110236, + "grad_norm": 24.133914947509766, + "learning_rate": 2.554671968190855e-07, + "loss": 2.0278, + "step": 260 + }, + { + "epoch": 0.25688976377952755, + "grad_norm": 19.490673065185547, + "learning_rate": 2.564612326043738e-07, + "loss": 1.2403, + "step": 261 + }, + { + "epoch": 0.2578740157480315, + "grad_norm": 20.14169692993164, + "learning_rate": 2.5745526838966207e-07, + "loss": 1.588, + "step": 262 + }, + { + "epoch": 0.2588582677165354, + "grad_norm": 26.948959350585938, + "learning_rate": 2.5844930417495037e-07, + "loss": 2.0071, + "step": 263 + }, + { + "epoch": 0.25984251968503935, + "grad_norm": 22.330224990844727, + "learning_rate": 2.5944333996023857e-07, + "loss": 1.5312, + "step": 264 + }, + { + "epoch": 0.2608267716535433, + "grad_norm": 21.379629135131836, + "learning_rate": 2.604373757455269e-07, + "loss": 1.8641, + "step": 265 + }, + { + "epoch": 0.2618110236220472, + "grad_norm": 20.703140258789062, + "learning_rate": 2.614314115308152e-07, + "loss": 1.2933, + "step": 266 + }, + { + "epoch": 0.26279527559055116, + "grad_norm": 22.9117488861084, + "learning_rate": 2.6242544731610343e-07, + "loss": 1.6262, + "step": 267 + }, + { + "epoch": 0.2637795275590551, + "grad_norm": 23.842002868652344, + "learning_rate": 2.634194831013917e-07, + "loss": 1.721, + "step": 268 + }, + { + "epoch": 0.26476377952755903, + "grad_norm": 20.449384689331055, + "learning_rate": 2.6441351888667994e-07, + "loss": 1.4713, + "step": 269 + }, + { + "epoch": 0.265748031496063, + "grad_norm": 25.885969161987305, + "learning_rate": 2.6540755467196824e-07, + "loss": 1.4625, + "step": 270 + }, + { + "epoch": 0.26673228346456695, + "grad_norm": 24.52666473388672, + "learning_rate": 2.664015904572565e-07, + "loss": 1.7254, + "step": 271 + }, + { + "epoch": 0.2677165354330709, + "grad_norm": 23.4957275390625, + "learning_rate": 2.6739562624254475e-07, + "loss": 1.5108, + "step": 272 + }, + { + "epoch": 0.2687007874015748, + "grad_norm": 23.828855514526367, + "learning_rate": 2.6838966202783305e-07, + "loss": 2.1126, + "step": 273 + }, + { + "epoch": 0.26968503937007876, + "grad_norm": 21.05967903137207, + "learning_rate": 2.693836978131213e-07, + "loss": 1.3967, + "step": 274 + }, + { + "epoch": 0.2706692913385827, + "grad_norm": 24.555776596069336, + "learning_rate": 2.703777335984096e-07, + "loss": 1.7067, + "step": 275 + }, + { + "epoch": 0.27165354330708663, + "grad_norm": 22.135860443115234, + "learning_rate": 2.7137176938369786e-07, + "loss": 1.4847, + "step": 276 + }, + { + "epoch": 0.27263779527559057, + "grad_norm": 22.105913162231445, + "learning_rate": 2.723658051689861e-07, + "loss": 1.6515, + "step": 277 + }, + { + "epoch": 0.2736220472440945, + "grad_norm": 16.617225646972656, + "learning_rate": 2.7335984095427437e-07, + "loss": 0.9367, + "step": 278 + }, + { + "epoch": 0.27460629921259844, + "grad_norm": 26.01727867126465, + "learning_rate": 2.743538767395627e-07, + "loss": 2.0267, + "step": 279 + }, + { + "epoch": 0.2755905511811024, + "grad_norm": 22.522462844848633, + "learning_rate": 2.75347912524851e-07, + "loss": 1.5023, + "step": 280 + }, + { + "epoch": 0.2765748031496063, + "grad_norm": 20.646358489990234, + "learning_rate": 2.763419483101392e-07, + "loss": 1.1248, + "step": 281 + }, + { + "epoch": 0.27755905511811024, + "grad_norm": 23.3087158203125, + "learning_rate": 2.773359840954275e-07, + "loss": 1.6224, + "step": 282 + }, + { + "epoch": 0.2785433070866142, + "grad_norm": 24.115968704223633, + "learning_rate": 2.7833001988071574e-07, + "loss": 1.7969, + "step": 283 + }, + { + "epoch": 0.2795275590551181, + "grad_norm": 27.229188919067383, + "learning_rate": 2.7932405566600404e-07, + "loss": 2.2498, + "step": 284 + }, + { + "epoch": 0.28051181102362205, + "grad_norm": 25.797216415405273, + "learning_rate": 2.803180914512923e-07, + "loss": 1.7477, + "step": 285 + }, + { + "epoch": 0.281496062992126, + "grad_norm": 23.666858673095703, + "learning_rate": 2.8131212723658055e-07, + "loss": 1.6261, + "step": 286 + }, + { + "epoch": 0.2824803149606299, + "grad_norm": 26.826387405395508, + "learning_rate": 2.8230616302186885e-07, + "loss": 2.0911, + "step": 287 + }, + { + "epoch": 0.28346456692913385, + "grad_norm": 23.288511276245117, + "learning_rate": 2.833001988071571e-07, + "loss": 1.9519, + "step": 288 + }, + { + "epoch": 0.2844488188976378, + "grad_norm": 19.76810073852539, + "learning_rate": 2.842942345924454e-07, + "loss": 1.3132, + "step": 289 + }, + { + "epoch": 0.2854330708661417, + "grad_norm": 24.65778923034668, + "learning_rate": 2.852882703777336e-07, + "loss": 2.3292, + "step": 290 + }, + { + "epoch": 0.28641732283464566, + "grad_norm": 19.230430603027344, + "learning_rate": 2.862823061630219e-07, + "loss": 1.3781, + "step": 291 + }, + { + "epoch": 0.2874015748031496, + "grad_norm": 20.665143966674805, + "learning_rate": 2.8727634194831017e-07, + "loss": 1.5753, + "step": 292 + }, + { + "epoch": 0.28838582677165353, + "grad_norm": 18.584993362426758, + "learning_rate": 2.8827037773359847e-07, + "loss": 1.4158, + "step": 293 + }, + { + "epoch": 0.28937007874015747, + "grad_norm": 20.731081008911133, + "learning_rate": 2.892644135188867e-07, + "loss": 2.1661, + "step": 294 + }, + { + "epoch": 0.2903543307086614, + "grad_norm": 20.006755828857422, + "learning_rate": 2.90258449304175e-07, + "loss": 1.4928, + "step": 295 + }, + { + "epoch": 0.29133858267716534, + "grad_norm": 20.01453971862793, + "learning_rate": 2.912524850894633e-07, + "loss": 2.2825, + "step": 296 + }, + { + "epoch": 0.29232283464566927, + "grad_norm": 18.137697219848633, + "learning_rate": 2.9224652087475153e-07, + "loss": 1.7261, + "step": 297 + }, + { + "epoch": 0.2933070866141732, + "grad_norm": 20.046411514282227, + "learning_rate": 2.9324055666003984e-07, + "loss": 1.8635, + "step": 298 + }, + { + "epoch": 0.29429133858267714, + "grad_norm": 15.859444618225098, + "learning_rate": 2.9423459244532804e-07, + "loss": 0.974, + "step": 299 + }, + { + "epoch": 0.2952755905511811, + "grad_norm": 17.976015090942383, + "learning_rate": 2.9522862823061634e-07, + "loss": 1.53, + "step": 300 + }, + { + "epoch": 0.296259842519685, + "grad_norm": 20.52704429626465, + "learning_rate": 2.9622266401590465e-07, + "loss": 1.5985, + "step": 301 + }, + { + "epoch": 0.297244094488189, + "grad_norm": 19.45134925842285, + "learning_rate": 2.972166998011929e-07, + "loss": 1.2169, + "step": 302 + }, + { + "epoch": 0.29822834645669294, + "grad_norm": 22.01141357421875, + "learning_rate": 2.9821073558648115e-07, + "loss": 1.771, + "step": 303 + }, + { + "epoch": 0.2992125984251969, + "grad_norm": 17.591617584228516, + "learning_rate": 2.992047713717694e-07, + "loss": 1.4506, + "step": 304 + }, + { + "epoch": 0.3001968503937008, + "grad_norm": 25.382244110107422, + "learning_rate": 3.001988071570577e-07, + "loss": 1.9496, + "step": 305 + }, + { + "epoch": 0.30118110236220474, + "grad_norm": 21.026479721069336, + "learning_rate": 3.01192842942346e-07, + "loss": 1.2436, + "step": 306 + }, + { + "epoch": 0.30118110236220474, + "eval_Qnli-dev_cosine_accuracy": 0.671875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.822678804397583, + "eval_Qnli-dev_cosine_ap": 0.6991833375897509, + "eval_Qnli-dev_cosine_f1": 0.6688, + "eval_Qnli-dev_cosine_f1_threshold": 0.7263889312744141, + "eval_Qnli-dev_cosine_precision": 0.5372750642673522, + "eval_Qnli-dev_cosine_recall": 0.885593220338983, + "eval_Qnli-dev_dot_accuracy": 0.671875, + "eval_Qnli-dev_dot_accuracy_threshold": 632.18896484375, + "eval_Qnli-dev_dot_ap": 0.6993450442740953, + "eval_Qnli-dev_dot_f1": 0.6688, + "eval_Qnli-dev_dot_f1_threshold": 558.1864013671875, + "eval_Qnli-dev_dot_precision": 0.5372750642673522, + "eval_Qnli-dev_dot_recall": 0.885593220338983, + "eval_Qnli-dev_euclidean_accuracy": 0.671875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 16.508346557617188, + "eval_Qnli-dev_euclidean_ap": 0.6991985293193381, + "eval_Qnli-dev_euclidean_f1": 0.6688, + "eval_Qnli-dev_euclidean_f1_threshold": 20.50628089904785, + "eval_Qnli-dev_euclidean_precision": 0.5372750642673522, + "eval_Qnli-dev_euclidean_recall": 0.885593220338983, + "eval_Qnli-dev_manhattan_accuracy": 0.66796875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 362.6116638183594, + "eval_Qnli-dev_manhattan_ap": 0.6988100851084365, + "eval_Qnli-dev_manhattan_f1": 0.669826224328594, + "eval_Qnli-dev_manhattan_f1_threshold": 456.92926025390625, + "eval_Qnli-dev_manhattan_precision": 0.5340050377833753, + "eval_Qnli-dev_manhattan_recall": 0.8983050847457628, + "eval_Qnli-dev_max_accuracy": 0.671875, + "eval_Qnli-dev_max_accuracy_threshold": 632.18896484375, + "eval_Qnli-dev_max_ap": 0.6993450442740953, + "eval_Qnli-dev_max_f1": 0.669826224328594, + "eval_Qnli-dev_max_f1_threshold": 558.1864013671875, + "eval_Qnli-dev_max_precision": 0.5372750642673522, + "eval_Qnli-dev_max_recall": 0.8983050847457628, + "eval_allNLI-dev_cosine_accuracy": 0.68359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9261605739593506, + "eval_allNLI-dev_cosine_ap": 0.47961880692471753, + "eval_allNLI-dev_cosine_f1": 0.5101351351351351, + "eval_allNLI-dev_cosine_f1_threshold": 0.7954977750778198, + "eval_allNLI-dev_cosine_precision": 0.360381861575179, + "eval_allNLI-dev_cosine_recall": 0.8728323699421965, + "eval_allNLI-dev_dot_accuracy": 0.68359375, + "eval_allNLI-dev_dot_accuracy_threshold": 711.72412109375, + "eval_allNLI-dev_dot_ap": 0.4796043382357695, + "eval_allNLI-dev_dot_f1": 0.5101351351351351, + "eval_allNLI-dev_dot_f1_threshold": 611.3052978515625, + "eval_allNLI-dev_dot_precision": 0.360381861575179, + "eval_allNLI-dev_dot_recall": 0.8728323699421965, + "eval_allNLI-dev_euclidean_accuracy": 0.68359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.652993202209473, + "eval_allNLI-dev_euclidean_ap": 0.47952222164725533, + "eval_allNLI-dev_euclidean_f1": 0.5101351351351351, + "eval_allNLI-dev_euclidean_f1_threshold": 17.728565216064453, + "eval_allNLI-dev_euclidean_precision": 0.360381861575179, + "eval_allNLI-dev_euclidean_recall": 0.8728323699421965, + "eval_allNLI-dev_manhattan_accuracy": 0.6796875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 236.265380859375, + "eval_allNLI-dev_manhattan_ap": 0.4807847983836923, + "eval_allNLI-dev_manhattan_f1": 0.5103668261562998, + "eval_allNLI-dev_manhattan_f1_threshold": 417.0115966796875, + "eval_allNLI-dev_manhattan_precision": 0.3524229074889868, + "eval_allNLI-dev_manhattan_recall": 0.9248554913294798, + "eval_allNLI-dev_max_accuracy": 0.68359375, + "eval_allNLI-dev_max_accuracy_threshold": 711.72412109375, + "eval_allNLI-dev_max_ap": 0.4807847983836923, + "eval_allNLI-dev_max_f1": 0.5103668261562998, + "eval_allNLI-dev_max_f1_threshold": 611.3052978515625, + "eval_allNLI-dev_max_precision": 0.360381861575179, + "eval_allNLI-dev_max_recall": 0.9248554913294798, + "eval_loss": 1.5213453769683838, + "eval_runtime": 50.241, + "eval_samples_per_second": 33.12, + "eval_sequential_score": 0.6993450442740953, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.4516880351405699, + "eval_sts-test_pearson_dot": 0.4516761779594392, + "eval_sts-test_pearson_euclidean": 0.4615962658267084, + "eval_sts-test_pearson_manhattan": 0.46201541629939036, + "eval_sts-test_pearson_max": 0.46201541629939036, + "eval_sts-test_spearman_cosine": 0.46725208205491164, + "eval_sts-test_spearman_dot": 0.4672651046719766, + "eval_sts-test_spearman_euclidean": 0.46725722844997175, + "eval_sts-test_spearman_manhattan": 0.4677601878767496, + "eval_sts-test_spearman_max": 0.4677601878767496, + "step": 306 + }, + { + "epoch": 0.3021653543307087, + "grad_norm": 27.924283981323242, + "learning_rate": 3.021868787276342e-07, + "loss": 2.2057, + "step": 307 + }, + { + "epoch": 0.3031496062992126, + "grad_norm": 19.933040618896484, + "learning_rate": 3.031809145129225e-07, + "loss": 1.6786, + "step": 308 + }, + { + "epoch": 0.30413385826771655, + "grad_norm": 20.79754638671875, + "learning_rate": 3.0417495029821077e-07, + "loss": 1.748, + "step": 309 + }, + { + "epoch": 0.3051181102362205, + "grad_norm": 21.448503494262695, + "learning_rate": 3.051689860834991e-07, + "loss": 1.5541, + "step": 310 + }, + { + "epoch": 0.3061023622047244, + "grad_norm": 23.68476676940918, + "learning_rate": 3.0616302186878733e-07, + "loss": 2.2968, + "step": 311 + }, + { + "epoch": 0.30708661417322836, + "grad_norm": 19.669784545898438, + "learning_rate": 3.071570576540756e-07, + "loss": 1.585, + "step": 312 + }, + { + "epoch": 0.3080708661417323, + "grad_norm": 21.109407424926758, + "learning_rate": 3.0815109343936384e-07, + "loss": 1.8371, + "step": 313 + }, + { + "epoch": 0.3090551181102362, + "grad_norm": 17.614957809448242, + "learning_rate": 3.0914512922465214e-07, + "loss": 1.1129, + "step": 314 + }, + { + "epoch": 0.31003937007874016, + "grad_norm": 18.3952693939209, + "learning_rate": 3.1013916500994045e-07, + "loss": 1.5495, + "step": 315 + }, + { + "epoch": 0.3110236220472441, + "grad_norm": 17.686391830444336, + "learning_rate": 3.1113320079522865e-07, + "loss": 1.4327, + "step": 316 + }, + { + "epoch": 0.31200787401574803, + "grad_norm": 21.255008697509766, + "learning_rate": 3.1212723658051695e-07, + "loss": 1.4801, + "step": 317 + }, + { + "epoch": 0.31299212598425197, + "grad_norm": 19.738698959350586, + "learning_rate": 3.131212723658052e-07, + "loss": 1.7096, + "step": 318 + }, + { + "epoch": 0.3139763779527559, + "grad_norm": 21.54611587524414, + "learning_rate": 3.141153081510935e-07, + "loss": 1.6717, + "step": 319 + }, + { + "epoch": 0.31496062992125984, + "grad_norm": 21.207399368286133, + "learning_rate": 3.151093439363818e-07, + "loss": 1.7151, + "step": 320 + }, + { + "epoch": 0.3159448818897638, + "grad_norm": 23.244253158569336, + "learning_rate": 3.1610337972167e-07, + "loss": 1.7081, + "step": 321 + }, + { + "epoch": 0.3169291338582677, + "grad_norm": 19.656099319458008, + "learning_rate": 3.170974155069583e-07, + "loss": 1.431, + "step": 322 + }, + { + "epoch": 0.31791338582677164, + "grad_norm": 22.234607696533203, + "learning_rate": 3.1809145129224657e-07, + "loss": 1.5734, + "step": 323 + }, + { + "epoch": 0.3188976377952756, + "grad_norm": 23.064062118530273, + "learning_rate": 3.190854870775349e-07, + "loss": 1.7307, + "step": 324 + }, + { + "epoch": 0.3198818897637795, + "grad_norm": 17.646806716918945, + "learning_rate": 3.200795228628231e-07, + "loss": 1.0644, + "step": 325 + }, + { + "epoch": 0.32086614173228345, + "grad_norm": 18.901668548583984, + "learning_rate": 3.210735586481114e-07, + "loss": 1.0651, + "step": 326 + }, + { + "epoch": 0.3218503937007874, + "grad_norm": 22.49507713317871, + "learning_rate": 3.220675944333997e-07, + "loss": 1.4805, + "step": 327 + }, + { + "epoch": 0.3228346456692913, + "grad_norm": 17.89360809326172, + "learning_rate": 3.2306163021868794e-07, + "loss": 0.839, + "step": 328 + }, + { + "epoch": 0.32381889763779526, + "grad_norm": 20.79975128173828, + "learning_rate": 3.240556660039762e-07, + "loss": 1.1801, + "step": 329 + }, + { + "epoch": 0.3248031496062992, + "grad_norm": 18.659631729125977, + "learning_rate": 3.2504970178926444e-07, + "loss": 1.36, + "step": 330 + }, + { + "epoch": 0.3257874015748031, + "grad_norm": 21.92584800720215, + "learning_rate": 3.2604373757455275e-07, + "loss": 1.3371, + "step": 331 + }, + { + "epoch": 0.32677165354330706, + "grad_norm": 17.514934539794922, + "learning_rate": 3.27037773359841e-07, + "loss": 1.1707, + "step": 332 + }, + { + "epoch": 0.327755905511811, + "grad_norm": 18.580942153930664, + "learning_rate": 3.280318091451293e-07, + "loss": 1.2572, + "step": 333 + }, + { + "epoch": 0.328740157480315, + "grad_norm": 22.246044158935547, + "learning_rate": 3.2902584493041756e-07, + "loss": 1.3537, + "step": 334 + }, + { + "epoch": 0.3297244094488189, + "grad_norm": 26.39656639099121, + "learning_rate": 3.300198807157058e-07, + "loss": 1.7096, + "step": 335 + }, + { + "epoch": 0.33070866141732286, + "grad_norm": 24.07045555114746, + "learning_rate": 3.310139165009941e-07, + "loss": 1.5137, + "step": 336 + }, + { + "epoch": 0.3316929133858268, + "grad_norm": 18.82029914855957, + "learning_rate": 3.3200795228628237e-07, + "loss": 1.1989, + "step": 337 + }, + { + "epoch": 0.33267716535433073, + "grad_norm": 19.72603416442871, + "learning_rate": 3.330019880715706e-07, + "loss": 1.3559, + "step": 338 + }, + { + "epoch": 0.33366141732283466, + "grad_norm": 19.72848129272461, + "learning_rate": 3.3399602385685887e-07, + "loss": 1.3643, + "step": 339 + }, + { + "epoch": 0.3346456692913386, + "grad_norm": 19.329681396484375, + "learning_rate": 3.349900596421472e-07, + "loss": 1.2283, + "step": 340 + }, + { + "epoch": 0.33562992125984253, + "grad_norm": 25.018512725830078, + "learning_rate": 3.359840954274355e-07, + "loss": 1.5829, + "step": 341 + }, + { + "epoch": 0.33661417322834647, + "grad_norm": 18.037017822265625, + "learning_rate": 3.369781312127237e-07, + "loss": 1.1866, + "step": 342 + }, + { + "epoch": 0.3375984251968504, + "grad_norm": 21.839269638061523, + "learning_rate": 3.37972166998012e-07, + "loss": 1.531, + "step": 343 + }, + { + "epoch": 0.33858267716535434, + "grad_norm": 21.1947021484375, + "learning_rate": 3.3896620278330024e-07, + "loss": 1.5581, + "step": 344 + }, + { + "epoch": 0.3395669291338583, + "grad_norm": 22.119661331176758, + "learning_rate": 3.3996023856858855e-07, + "loss": 1.5587, + "step": 345 + }, + { + "epoch": 0.3405511811023622, + "grad_norm": 19.932144165039062, + "learning_rate": 3.409542743538768e-07, + "loss": 1.1403, + "step": 346 + }, + { + "epoch": 0.34153543307086615, + "grad_norm": 26.49030303955078, + "learning_rate": 3.4194831013916505e-07, + "loss": 1.9728, + "step": 347 + }, + { + "epoch": 0.3425196850393701, + "grad_norm": 18.05373191833496, + "learning_rate": 3.4294234592445336e-07, + "loss": 1.0818, + "step": 348 + }, + { + "epoch": 0.343503937007874, + "grad_norm": 19.15143585205078, + "learning_rate": 3.439363817097416e-07, + "loss": 1.2993, + "step": 349 + }, + { + "epoch": 0.34448818897637795, + "grad_norm": 21.032453536987305, + "learning_rate": 3.449304174950299e-07, + "loss": 1.7779, + "step": 350 + }, + { + "epoch": 0.3454724409448819, + "grad_norm": 20.775541305541992, + "learning_rate": 3.459244532803181e-07, + "loss": 1.319, + "step": 351 + }, + { + "epoch": 0.3464566929133858, + "grad_norm": 24.86508560180664, + "learning_rate": 3.469184890656064e-07, + "loss": 1.9236, + "step": 352 + }, + { + "epoch": 0.34744094488188976, + "grad_norm": 19.496232986450195, + "learning_rate": 3.4791252485089467e-07, + "loss": 1.3085, + "step": 353 + }, + { + "epoch": 0.3484251968503937, + "grad_norm": 24.63651466369629, + "learning_rate": 3.48906560636183e-07, + "loss": 2.2049, + "step": 354 + }, + { + "epoch": 0.3494094488188976, + "grad_norm": 16.268550872802734, + "learning_rate": 3.499005964214713e-07, + "loss": 1.3697, + "step": 355 + }, + { + "epoch": 0.35039370078740156, + "grad_norm": 17.859365463256836, + "learning_rate": 3.508946322067595e-07, + "loss": 1.5367, + "step": 356 + }, + { + "epoch": 0.3513779527559055, + "grad_norm": 22.554317474365234, + "learning_rate": 3.518886679920478e-07, + "loss": 1.2516, + "step": 357 + }, + { + "epoch": 0.35236220472440943, + "grad_norm": 20.86847496032715, + "learning_rate": 3.5288270377733604e-07, + "loss": 1.6497, + "step": 358 + }, + { + "epoch": 0.35334645669291337, + "grad_norm": 17.967374801635742, + "learning_rate": 3.5387673956262434e-07, + "loss": 1.2457, + "step": 359 + }, + { + "epoch": 0.3543307086614173, + "grad_norm": 20.069971084594727, + "learning_rate": 3.5487077534791254e-07, + "loss": 1.2733, + "step": 360 + }, + { + "epoch": 0.35531496062992124, + "grad_norm": 20.93694496154785, + "learning_rate": 3.5586481113320085e-07, + "loss": 1.4768, + "step": 361 + }, + { + "epoch": 0.3562992125984252, + "grad_norm": 19.617618560791016, + "learning_rate": 3.5685884691848915e-07, + "loss": 1.1363, + "step": 362 + }, + { + "epoch": 0.3572834645669291, + "grad_norm": 23.839365005493164, + "learning_rate": 3.578528827037774e-07, + "loss": 1.5731, + "step": 363 + }, + { + "epoch": 0.35826771653543305, + "grad_norm": 17.460094451904297, + "learning_rate": 3.5884691848906566e-07, + "loss": 1.0821, + "step": 364 + }, + { + "epoch": 0.359251968503937, + "grad_norm": 19.736909866333008, + "learning_rate": 3.598409542743539e-07, + "loss": 1.1563, + "step": 365 + }, + { + "epoch": 0.36023622047244097, + "grad_norm": 23.919334411621094, + "learning_rate": 3.608349900596422e-07, + "loss": 1.8843, + "step": 366 + }, + { + "epoch": 0.3612204724409449, + "grad_norm": 20.10567283630371, + "learning_rate": 3.6182902584493047e-07, + "loss": 1.2239, + "step": 367 + }, + { + "epoch": 0.36220472440944884, + "grad_norm": 20.869524002075195, + "learning_rate": 3.6282306163021877e-07, + "loss": 1.4411, + "step": 368 + }, + { + "epoch": 0.3631889763779528, + "grad_norm": 23.02889633178711, + "learning_rate": 3.63817097415507e-07, + "loss": 2.1003, + "step": 369 + }, + { + "epoch": 0.3641732283464567, + "grad_norm": 19.797988891601562, + "learning_rate": 3.648111332007953e-07, + "loss": 1.6558, + "step": 370 + }, + { + "epoch": 0.36515748031496065, + "grad_norm": 20.25324058532715, + "learning_rate": 3.658051689860836e-07, + "loss": 1.6502, + "step": 371 + }, + { + "epoch": 0.3661417322834646, + "grad_norm": 19.865921020507812, + "learning_rate": 3.6679920477137184e-07, + "loss": 1.7204, + "step": 372 + }, + { + "epoch": 0.3671259842519685, + "grad_norm": 20.396156311035156, + "learning_rate": 3.677932405566601e-07, + "loss": 1.7422, + "step": 373 + }, + { + "epoch": 0.36811023622047245, + "grad_norm": 20.016502380371094, + "learning_rate": 3.6878727634194834e-07, + "loss": 1.3859, + "step": 374 + }, + { + "epoch": 0.3690944881889764, + "grad_norm": 14.640628814697266, + "learning_rate": 3.6978131212723665e-07, + "loss": 0.8876, + "step": 375 + }, + { + "epoch": 0.3700787401574803, + "grad_norm": 20.027318954467773, + "learning_rate": 3.7077534791252495e-07, + "loss": 1.2399, + "step": 376 + }, + { + "epoch": 0.37106299212598426, + "grad_norm": 17.005905151367188, + "learning_rate": 3.7176938369781315e-07, + "loss": 1.1039, + "step": 377 + }, + { + "epoch": 0.3720472440944882, + "grad_norm": 21.551816940307617, + "learning_rate": 3.7276341948310145e-07, + "loss": 1.733, + "step": 378 + }, + { + "epoch": 0.37303149606299213, + "grad_norm": 23.17444610595703, + "learning_rate": 3.737574552683897e-07, + "loss": 1.6897, + "step": 379 + }, + { + "epoch": 0.37401574803149606, + "grad_norm": 22.414926528930664, + "learning_rate": 3.74751491053678e-07, + "loss": 2.0532, + "step": 380 + }, + { + "epoch": 0.375, + "grad_norm": 17.610034942626953, + "learning_rate": 3.7574552683896626e-07, + "loss": 1.0156, + "step": 381 + }, + { + "epoch": 0.37598425196850394, + "grad_norm": 15.841965675354004, + "learning_rate": 3.767395626242545e-07, + "loss": 0.8888, + "step": 382 + }, + { + "epoch": 0.37696850393700787, + "grad_norm": 20.21027946472168, + "learning_rate": 3.777335984095428e-07, + "loss": 1.322, + "step": 383 + }, + { + "epoch": 0.3779527559055118, + "grad_norm": 23.444580078125, + "learning_rate": 3.787276341948311e-07, + "loss": 1.6828, + "step": 384 + }, + { + "epoch": 0.37893700787401574, + "grad_norm": 18.69243812561035, + "learning_rate": 3.797216699801194e-07, + "loss": 1.1567, + "step": 385 + }, + { + "epoch": 0.3799212598425197, + "grad_norm": 20.93277931213379, + "learning_rate": 3.807157057654076e-07, + "loss": 1.6117, + "step": 386 + }, + { + "epoch": 0.3809055118110236, + "grad_norm": 17.97561264038086, + "learning_rate": 3.817097415506959e-07, + "loss": 1.1776, + "step": 387 + }, + { + "epoch": 0.38188976377952755, + "grad_norm": 23.98857307434082, + "learning_rate": 3.8270377733598414e-07, + "loss": 1.641, + "step": 388 + }, + { + "epoch": 0.3828740157480315, + "grad_norm": 18.457351684570312, + "learning_rate": 3.8369781312127244e-07, + "loss": 1.3454, + "step": 389 + }, + { + "epoch": 0.3838582677165354, + "grad_norm": 22.10936737060547, + "learning_rate": 3.8469184890656075e-07, + "loss": 1.4292, + "step": 390 + }, + { + "epoch": 0.38484251968503935, + "grad_norm": 18.214380264282227, + "learning_rate": 3.8568588469184895e-07, + "loss": 1.2256, + "step": 391 + }, + { + "epoch": 0.3858267716535433, + "grad_norm": 19.54240608215332, + "learning_rate": 3.8667992047713725e-07, + "loss": 1.08, + "step": 392 + }, + { + "epoch": 0.3868110236220472, + "grad_norm": 15.17556095123291, + "learning_rate": 3.876739562624255e-07, + "loss": 0.7436, + "step": 393 + }, + { + "epoch": 0.38779527559055116, + "grad_norm": 17.79245948791504, + "learning_rate": 3.886679920477138e-07, + "loss": 1.4112, + "step": 394 + }, + { + "epoch": 0.3887795275590551, + "grad_norm": 15.915721893310547, + "learning_rate": 3.89662027833002e-07, + "loss": 0.8917, + "step": 395 + }, + { + "epoch": 0.38976377952755903, + "grad_norm": 17.685527801513672, + "learning_rate": 3.906560636182903e-07, + "loss": 0.9955, + "step": 396 + }, + { + "epoch": 0.390748031496063, + "grad_norm": 18.575515747070312, + "learning_rate": 3.916500994035786e-07, + "loss": 1.2867, + "step": 397 + }, + { + "epoch": 0.39173228346456695, + "grad_norm": 19.448511123657227, + "learning_rate": 3.9264413518886687e-07, + "loss": 1.0683, + "step": 398 + }, + { + "epoch": 0.3927165354330709, + "grad_norm": 18.82157325744629, + "learning_rate": 3.936381709741551e-07, + "loss": 0.9355, + "step": 399 + }, + { + "epoch": 0.3937007874015748, + "grad_norm": 17.697452545166016, + "learning_rate": 3.946322067594434e-07, + "loss": 1.1153, + "step": 400 + }, + { + "epoch": 0.39468503937007876, + "grad_norm": 18.46881866455078, + "learning_rate": 3.956262425447317e-07, + "loss": 1.1724, + "step": 401 + }, + { + "epoch": 0.3956692913385827, + "grad_norm": 24.634397506713867, + "learning_rate": 3.9662027833001993e-07, + "loss": 1.4069, + "step": 402 + }, + { + "epoch": 0.39665354330708663, + "grad_norm": 20.34600257873535, + "learning_rate": 3.9761431411530824e-07, + "loss": 1.2546, + "step": 403 + }, + { + "epoch": 0.39763779527559057, + "grad_norm": 26.437278747558594, + "learning_rate": 3.986083499005965e-07, + "loss": 2.2862, + "step": 404 + }, + { + "epoch": 0.3986220472440945, + "grad_norm": 17.79900360107422, + "learning_rate": 3.9960238568588474e-07, + "loss": 1.2316, + "step": 405 + }, + { + "epoch": 0.39960629921259844, + "grad_norm": 24.866077423095703, + "learning_rate": 4.0059642147117305e-07, + "loss": 1.7876, + "step": 406 + }, + { + "epoch": 0.4005905511811024, + "grad_norm": 14.236503601074219, + "learning_rate": 4.015904572564613e-07, + "loss": 0.6936, + "step": 407 + }, + { + "epoch": 0.4015748031496063, + "grad_norm": 21.97986602783203, + "learning_rate": 4.0258449304174955e-07, + "loss": 1.3852, + "step": 408 + }, + { + "epoch": 0.40255905511811024, + "grad_norm": 19.722715377807617, + "learning_rate": 4.035785288270378e-07, + "loss": 1.9046, + "step": 409 + }, + { + "epoch": 0.4035433070866142, + "grad_norm": 18.966148376464844, + "learning_rate": 4.045725646123261e-07, + "loss": 1.4972, + "step": 410 + }, + { + "epoch": 0.4045275590551181, + "grad_norm": 13.10688304901123, + "learning_rate": 4.055666003976144e-07, + "loss": 0.5531, + "step": 411 + }, + { + "epoch": 0.40551181102362205, + "grad_norm": 16.64056968688965, + "learning_rate": 4.065606361829026e-07, + "loss": 1.3685, + "step": 412 + }, + { + "epoch": 0.406496062992126, + "grad_norm": 17.951778411865234, + "learning_rate": 4.075546719681909e-07, + "loss": 1.1367, + "step": 413 + }, + { + "epoch": 0.4074803149606299, + "grad_norm": 18.79546356201172, + "learning_rate": 4.085487077534792e-07, + "loss": 1.1304, + "step": 414 + }, + { + "epoch": 0.40846456692913385, + "grad_norm": 23.208480834960938, + "learning_rate": 4.095427435387675e-07, + "loss": 1.5953, + "step": 415 + }, + { + "epoch": 0.4094488188976378, + "grad_norm": 24.188060760498047, + "learning_rate": 4.1053677932405573e-07, + "loss": 2.0308, + "step": 416 + }, + { + "epoch": 0.4104330708661417, + "grad_norm": 24.551694869995117, + "learning_rate": 4.11530815109344e-07, + "loss": 1.7275, + "step": 417 + }, + { + "epoch": 0.41141732283464566, + "grad_norm": 17.98610496520996, + "learning_rate": 4.125248508946323e-07, + "loss": 0.9921, + "step": 418 + }, + { + "epoch": 0.4124015748031496, + "grad_norm": 18.42479705810547, + "learning_rate": 4.1351888667992054e-07, + "loss": 1.3418, + "step": 419 + }, + { + "epoch": 0.41338582677165353, + "grad_norm": 18.93707275390625, + "learning_rate": 4.1451292246520885e-07, + "loss": 1.108, + "step": 420 + }, + { + "epoch": 0.41437007874015747, + "grad_norm": 17.49553680419922, + "learning_rate": 4.1550695825049705e-07, + "loss": 1.4359, + "step": 421 + }, + { + "epoch": 0.4153543307086614, + "grad_norm": 18.783565521240234, + "learning_rate": 4.1650099403578535e-07, + "loss": 1.4537, + "step": 422 + }, + { + "epoch": 0.41633858267716534, + "grad_norm": 16.676990509033203, + "learning_rate": 4.174950298210736e-07, + "loss": 0.8416, + "step": 423 + }, + { + "epoch": 0.41732283464566927, + "grad_norm": 15.969378471374512, + "learning_rate": 4.184890656063619e-07, + "loss": 0.8904, + "step": 424 + }, + { + "epoch": 0.4183070866141732, + "grad_norm": 15.156847953796387, + "learning_rate": 4.194831013916502e-07, + "loss": 0.7937, + "step": 425 + }, + { + "epoch": 0.41929133858267714, + "grad_norm": 16.689002990722656, + "learning_rate": 4.204771371769384e-07, + "loss": 0.9105, + "step": 426 + }, + { + "epoch": 0.4202755905511811, + "grad_norm": 18.2681884765625, + "learning_rate": 4.214711729622267e-07, + "loss": 1.1661, + "step": 427 + }, + { + "epoch": 0.421259842519685, + "grad_norm": 17.457895278930664, + "learning_rate": 4.2246520874751497e-07, + "loss": 0.7751, + "step": 428 + }, + { + "epoch": 0.422244094488189, + "grad_norm": 16.101627349853516, + "learning_rate": 4.234592445328033e-07, + "loss": 0.9039, + "step": 429 + }, + { + "epoch": 0.42322834645669294, + "grad_norm": 21.5301513671875, + "learning_rate": 4.244532803180915e-07, + "loss": 1.2651, + "step": 430 + }, + { + "epoch": 0.4242125984251969, + "grad_norm": 24.272584915161133, + "learning_rate": 4.254473161033798e-07, + "loss": 1.44, + "step": 431 + }, + { + "epoch": 0.4251968503937008, + "grad_norm": 18.749507904052734, + "learning_rate": 4.264413518886681e-07, + "loss": 0.9795, + "step": 432 + }, + { + "epoch": 0.42618110236220474, + "grad_norm": 29.54834747314453, + "learning_rate": 4.2743538767395634e-07, + "loss": 2.1892, + "step": 433 + }, + { + "epoch": 0.4271653543307087, + "grad_norm": 21.851760864257812, + "learning_rate": 4.284294234592446e-07, + "loss": 1.214, + "step": 434 + }, + { + "epoch": 0.4281496062992126, + "grad_norm": 21.678668975830078, + "learning_rate": 4.2942345924453284e-07, + "loss": 1.185, + "step": 435 + }, + { + "epoch": 0.42913385826771655, + "grad_norm": 20.97591781616211, + "learning_rate": 4.3041749502982115e-07, + "loss": 1.2501, + "step": 436 + }, + { + "epoch": 0.4301181102362205, + "grad_norm": 20.558670043945312, + "learning_rate": 4.3141153081510945e-07, + "loss": 1.6432, + "step": 437 + }, + { + "epoch": 0.4311023622047244, + "grad_norm": 18.290483474731445, + "learning_rate": 4.324055666003977e-07, + "loss": 1.0203, + "step": 438 + }, + { + "epoch": 0.43208661417322836, + "grad_norm": 21.71299171447754, + "learning_rate": 4.3339960238568596e-07, + "loss": 1.5179, + "step": 439 + }, + { + "epoch": 0.4330708661417323, + "grad_norm": 19.34381866455078, + "learning_rate": 4.343936381709742e-07, + "loss": 1.1445, + "step": 440 + }, + { + "epoch": 0.4340551181102362, + "grad_norm": 17.65093994140625, + "learning_rate": 4.353876739562625e-07, + "loss": 1.3099, + "step": 441 + }, + { + "epoch": 0.43503937007874016, + "grad_norm": 16.550668716430664, + "learning_rate": 4.3638170974155077e-07, + "loss": 0.8856, + "step": 442 + }, + { + "epoch": 0.4360236220472441, + "grad_norm": 11.908206939697266, + "learning_rate": 4.37375745526839e-07, + "loss": 0.5869, + "step": 443 + }, + { + "epoch": 0.43700787401574803, + "grad_norm": 23.472492218017578, + "learning_rate": 4.383697813121273e-07, + "loss": 1.6335, + "step": 444 + }, + { + "epoch": 0.43799212598425197, + "grad_norm": 21.598403930664062, + "learning_rate": 4.393638170974156e-07, + "loss": 1.4134, + "step": 445 + }, + { + "epoch": 0.4389763779527559, + "grad_norm": 19.63410186767578, + "learning_rate": 4.403578528827039e-07, + "loss": 1.0244, + "step": 446 + }, + { + "epoch": 0.43996062992125984, + "grad_norm": 18.266443252563477, + "learning_rate": 4.413518886679921e-07, + "loss": 1.103, + "step": 447 + }, + { + "epoch": 0.4409448818897638, + "grad_norm": 21.08526611328125, + "learning_rate": 4.423459244532804e-07, + "loss": 0.9848, + "step": 448 + }, + { + "epoch": 0.4419291338582677, + "grad_norm": 24.54269027709961, + "learning_rate": 4.4333996023856864e-07, + "loss": 1.5089, + "step": 449 + }, + { + "epoch": 0.44291338582677164, + "grad_norm": 19.617361068725586, + "learning_rate": 4.4433399602385695e-07, + "loss": 1.0422, + "step": 450 + }, + { + "epoch": 0.4438976377952756, + "grad_norm": 20.60758399963379, + "learning_rate": 4.4532803180914525e-07, + "loss": 1.0462, + "step": 451 + }, + { + "epoch": 0.4448818897637795, + "grad_norm": 20.939077377319336, + "learning_rate": 4.4632206759443345e-07, + "loss": 1.2857, + "step": 452 + }, + { + "epoch": 0.44586614173228345, + "grad_norm": 19.295501708984375, + "learning_rate": 4.4731610337972176e-07, + "loss": 1.4132, + "step": 453 + }, + { + "epoch": 0.4468503937007874, + "grad_norm": 22.07086753845215, + "learning_rate": 4.4831013916501e-07, + "loss": 1.3061, + "step": 454 + }, + { + "epoch": 0.4478346456692913, + "grad_norm": 23.644939422607422, + "learning_rate": 4.493041749502983e-07, + "loss": 1.3977, + "step": 455 + }, + { + "epoch": 0.44881889763779526, + "grad_norm": 21.088239669799805, + "learning_rate": 4.502982107355865e-07, + "loss": 1.3557, + "step": 456 + }, + { + "epoch": 0.4498031496062992, + "grad_norm": 23.44856834411621, + "learning_rate": 4.512922465208748e-07, + "loss": 1.3595, + "step": 457 + }, + { + "epoch": 0.4507874015748031, + "grad_norm": 16.646135330200195, + "learning_rate": 4.522862823061631e-07, + "loss": 0.8647, + "step": 458 + }, + { + "epoch": 0.45177165354330706, + "grad_norm": 21.29861831665039, + "learning_rate": 4.532803180914514e-07, + "loss": 1.3905, + "step": 459 + }, + { + "epoch": 0.45177165354330706, + "eval_Qnli-dev_cosine_accuracy": 0.677734375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7905216217041016, + "eval_Qnli-dev_cosine_ap": 0.7092754004450792, + "eval_Qnli-dev_cosine_f1": 0.7005444646098005, + "eval_Qnli-dev_cosine_f1_threshold": 0.7738929986953735, + "eval_Qnli-dev_cosine_precision": 0.6126984126984127, + "eval_Qnli-dev_cosine_recall": 0.8177966101694916, + "eval_Qnli-dev_dot_accuracy": 0.677734375, + "eval_Qnli-dev_dot_accuracy_threshold": 607.5902099609375, + "eval_Qnli-dev_dot_ap": 0.7093564593718997, + "eval_Qnli-dev_dot_f1": 0.7005444646098005, + "eval_Qnli-dev_dot_f1_threshold": 594.82958984375, + "eval_Qnli-dev_dot_precision": 0.6126984126984127, + "eval_Qnli-dev_dot_recall": 0.8177966101694916, + "eval_Qnli-dev_euclidean_accuracy": 0.677734375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.944761276245117, + "eval_Qnli-dev_euclidean_ap": 0.7092822066603943, + "eval_Qnli-dev_euclidean_f1": 0.7005444646098005, + "eval_Qnli-dev_euclidean_f1_threshold": 18.64351463317871, + "eval_Qnli-dev_euclidean_precision": 0.6126984126984127, + "eval_Qnli-dev_euclidean_recall": 0.8177966101694916, + "eval_Qnli-dev_manhattan_accuracy": 0.677734375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 403.15380859375, + "eval_Qnli-dev_manhattan_ap": 0.7081773921924046, + "eval_Qnli-dev_manhattan_f1": 0.696113074204947, + "eval_Qnli-dev_manhattan_f1_threshold": 417.12298583984375, + "eval_Qnli-dev_manhattan_precision": 0.5969696969696969, + "eval_Qnli-dev_manhattan_recall": 0.8347457627118644, + "eval_Qnli-dev_max_accuracy": 0.677734375, + "eval_Qnli-dev_max_accuracy_threshold": 607.5902099609375, + "eval_Qnli-dev_max_ap": 0.7093564593718997, + "eval_Qnli-dev_max_f1": 0.7005444646098005, + "eval_Qnli-dev_max_f1_threshold": 594.82958984375, + "eval_Qnli-dev_max_precision": 0.6126984126984127, + "eval_Qnli-dev_max_recall": 0.8347457627118644, + "eval_allNLI-dev_cosine_accuracy": 0.689453125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9186867475509644, + "eval_allNLI-dev_cosine_ap": 0.493528404484353, + "eval_allNLI-dev_cosine_f1": 0.5283018867924528, + "eval_allNLI-dev_cosine_f1_threshold": 0.8205500245094299, + "eval_allNLI-dev_cosine_precision": 0.4144736842105263, + "eval_allNLI-dev_cosine_recall": 0.7283236994219653, + "eval_allNLI-dev_dot_accuracy": 0.689453125, + "eval_allNLI-dev_dot_accuracy_threshold": 706.1145629882812, + "eval_allNLI-dev_dot_ap": 0.49366530598882574, + "eval_allNLI-dev_dot_f1": 0.5283018867924528, + "eval_allNLI-dev_dot_f1_threshold": 630.7012939453125, + "eval_allNLI-dev_dot_precision": 0.4144736842105263, + "eval_allNLI-dev_dot_recall": 0.7283236994219653, + "eval_allNLI-dev_euclidean_accuracy": 0.689453125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.180255889892578, + "eval_allNLI-dev_euclidean_ap": 0.49351830756144216, + "eval_allNLI-dev_euclidean_f1": 0.5283018867924528, + "eval_allNLI-dev_euclidean_f1_threshold": 16.60896873474121, + "eval_allNLI-dev_euclidean_precision": 0.4144736842105263, + "eval_allNLI-dev_euclidean_recall": 0.7283236994219653, + "eval_allNLI-dev_manhattan_accuracy": 0.69140625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 251.75421142578125, + "eval_allNLI-dev_manhattan_ap": 0.4930222958157673, + "eval_allNLI-dev_manhattan_f1": 0.5285412262156448, + "eval_allNLI-dev_manhattan_f1_threshold": 364.7342529296875, + "eval_allNLI-dev_manhattan_precision": 0.4166666666666667, + "eval_allNLI-dev_manhattan_recall": 0.7225433526011561, + "eval_allNLI-dev_max_accuracy": 0.69140625, + "eval_allNLI-dev_max_accuracy_threshold": 706.1145629882812, + "eval_allNLI-dev_max_ap": 0.49366530598882574, + "eval_allNLI-dev_max_f1": 0.5285412262156448, + "eval_allNLI-dev_max_f1_threshold": 630.7012939453125, + "eval_allNLI-dev_max_precision": 0.4166666666666667, + "eval_allNLI-dev_max_recall": 0.7283236994219653, + "eval_loss": 1.2969162464141846, + "eval_runtime": 50.0648, + "eval_samples_per_second": 33.237, + "eval_sequential_score": 0.7093564593718997, + "eval_steps_per_second": 0.14, + "eval_sts-test_pearson_cosine": 0.5415735956899457, + "eval_sts-test_pearson_dot": 0.5415314946904102, + "eval_sts-test_pearson_euclidean": 0.5463232738462487, + "eval_sts-test_pearson_manhattan": 0.5451734598739164, + "eval_sts-test_pearson_max": 0.5463232738462487, + "eval_sts-test_spearman_cosine": 0.5432558343516091, + "eval_sts-test_spearman_dot": 0.543190094748625, + "eval_sts-test_spearman_euclidean": 0.5433125789510533, + "eval_sts-test_spearman_manhattan": 0.5424313371117337, + "eval_sts-test_spearman_max": 0.5433125789510533, + "step": 459 + }, + { + "epoch": 0.452755905511811, + "grad_norm": 19.235441207885742, + "learning_rate": 4.542743538767397e-07, + "loss": 0.9467, + "step": 460 + }, + { + "epoch": 0.453740157480315, + "grad_norm": 26.713695526123047, + "learning_rate": 4.552683896620279e-07, + "loss": 1.9372, + "step": 461 + }, + { + "epoch": 0.4547244094488189, + "grad_norm": 13.988560676574707, + "learning_rate": 4.562624254473162e-07, + "loss": 0.871, + "step": 462 + }, + { + "epoch": 0.45570866141732286, + "grad_norm": 20.346981048583984, + "learning_rate": 4.5725646123260444e-07, + "loss": 1.2282, + "step": 463 + }, + { + "epoch": 0.4566929133858268, + "grad_norm": 18.093189239501953, + "learning_rate": 4.5825049701789274e-07, + "loss": 1.3845, + "step": 464 + }, + { + "epoch": 0.45767716535433073, + "grad_norm": 20.26774024963379, + "learning_rate": 4.59244532803181e-07, + "loss": 1.2571, + "step": 465 + }, + { + "epoch": 0.45866141732283466, + "grad_norm": 22.511838912963867, + "learning_rate": 4.6023856858846925e-07, + "loss": 1.2288, + "step": 466 + }, + { + "epoch": 0.4596456692913386, + "grad_norm": 15.89570140838623, + "learning_rate": 4.6123260437375755e-07, + "loss": 1.1165, + "step": 467 + }, + { + "epoch": 0.46062992125984253, + "grad_norm": 21.215049743652344, + "learning_rate": 4.622266401590458e-07, + "loss": 1.8463, + "step": 468 + }, + { + "epoch": 0.46161417322834647, + "grad_norm": 18.364500045776367, + "learning_rate": 4.6322067594433406e-07, + "loss": 0.9158, + "step": 469 + }, + { + "epoch": 0.4625984251968504, + "grad_norm": 18.762876510620117, + "learning_rate": 4.642147117296223e-07, + "loss": 0.8711, + "step": 470 + }, + { + "epoch": 0.46358267716535434, + "grad_norm": 19.44731330871582, + "learning_rate": 4.652087475149106e-07, + "loss": 1.4741, + "step": 471 + }, + { + "epoch": 0.4645669291338583, + "grad_norm": 15.906646728515625, + "learning_rate": 4.662027833001989e-07, + "loss": 0.914, + "step": 472 + }, + { + "epoch": 0.4655511811023622, + "grad_norm": 15.942811012268066, + "learning_rate": 4.671968190854872e-07, + "loss": 0.9435, + "step": 473 + }, + { + "epoch": 0.46653543307086615, + "grad_norm": 17.3239803314209, + "learning_rate": 4.681908548707754e-07, + "loss": 1.0876, + "step": 474 + }, + { + "epoch": 0.4675196850393701, + "grad_norm": 22.522132873535156, + "learning_rate": 4.691848906560637e-07, + "loss": 1.2365, + "step": 475 + }, + { + "epoch": 0.468503937007874, + "grad_norm": 18.67418098449707, + "learning_rate": 4.70178926441352e-07, + "loss": 1.1237, + "step": 476 + }, + { + "epoch": 0.46948818897637795, + "grad_norm": 22.97576332092285, + "learning_rate": 4.7117296222664024e-07, + "loss": 1.0097, + "step": 477 + }, + { + "epoch": 0.4704724409448819, + "grad_norm": 18.241655349731445, + "learning_rate": 4.721669980119285e-07, + "loss": 1.1548, + "step": 478 + }, + { + "epoch": 0.4714566929133858, + "grad_norm": 23.206424713134766, + "learning_rate": 4.731610337972168e-07, + "loss": 1.3203, + "step": 479 + }, + { + "epoch": 0.47244094488188976, + "grad_norm": 20.233009338378906, + "learning_rate": 4.7415506958250505e-07, + "loss": 1.2533, + "step": 480 + }, + { + "epoch": 0.4734251968503937, + "grad_norm": 20.482505798339844, + "learning_rate": 4.7514910536779335e-07, + "loss": 1.093, + "step": 481 + }, + { + "epoch": 0.4744094488188976, + "grad_norm": 19.419912338256836, + "learning_rate": 4.7614314115308155e-07, + "loss": 1.2591, + "step": 482 + }, + { + "epoch": 0.47539370078740156, + "grad_norm": 15.261035919189453, + "learning_rate": 4.771371769383698e-07, + "loss": 0.6764, + "step": 483 + }, + { + "epoch": 0.4763779527559055, + "grad_norm": 15.785579681396484, + "learning_rate": 4.781312127236581e-07, + "loss": 0.8922, + "step": 484 + }, + { + "epoch": 0.47736220472440943, + "grad_norm": 15.987274169921875, + "learning_rate": 4.791252485089464e-07, + "loss": 0.8524, + "step": 485 + }, + { + "epoch": 0.47834645669291337, + "grad_norm": 21.142431259155273, + "learning_rate": 4.801192842942347e-07, + "loss": 1.2777, + "step": 486 + }, + { + "epoch": 0.4793307086614173, + "grad_norm": 18.106870651245117, + "learning_rate": 4.811133200795229e-07, + "loss": 1.1682, + "step": 487 + }, + { + "epoch": 0.48031496062992124, + "grad_norm": 19.057527542114258, + "learning_rate": 4.821073558648112e-07, + "loss": 0.8617, + "step": 488 + }, + { + "epoch": 0.4812992125984252, + "grad_norm": 18.410564422607422, + "learning_rate": 4.831013916500995e-07, + "loss": 1.0303, + "step": 489 + }, + { + "epoch": 0.4822834645669291, + "grad_norm": 20.72304916381836, + "learning_rate": 4.840954274353877e-07, + "loss": 0.9843, + "step": 490 + }, + { + "epoch": 0.48326771653543305, + "grad_norm": 20.666017532348633, + "learning_rate": 4.85089463220676e-07, + "loss": 1.2951, + "step": 491 + }, + { + "epoch": 0.484251968503937, + "grad_norm": 23.45387077331543, + "learning_rate": 4.860834990059642e-07, + "loss": 1.7889, + "step": 492 + }, + { + "epoch": 0.48523622047244097, + "grad_norm": 18.90536117553711, + "learning_rate": 4.870775347912525e-07, + "loss": 1.118, + "step": 493 + }, + { + "epoch": 0.4862204724409449, + "grad_norm": 14.932930946350098, + "learning_rate": 4.880715705765408e-07, + "loss": 0.6772, + "step": 494 + }, + { + "epoch": 0.48720472440944884, + "grad_norm": 21.98814582824707, + "learning_rate": 4.890656063618291e-07, + "loss": 1.5058, + "step": 495 + }, + { + "epoch": 0.4881889763779528, + "grad_norm": 15.60319995880127, + "learning_rate": 4.900596421471173e-07, + "loss": 1.0068, + "step": 496 + }, + { + "epoch": 0.4891732283464567, + "grad_norm": 15.000826835632324, + "learning_rate": 4.910536779324057e-07, + "loss": 0.9024, + "step": 497 + }, + { + "epoch": 0.49015748031496065, + "grad_norm": 20.488412857055664, + "learning_rate": 4.92047713717694e-07, + "loss": 1.4816, + "step": 498 + }, + { + "epoch": 0.4911417322834646, + "grad_norm": 16.296037673950195, + "learning_rate": 4.930417495029823e-07, + "loss": 0.894, + "step": 499 + }, + { + "epoch": 0.4921259842519685, + "grad_norm": 16.42416000366211, + "learning_rate": 4.940357852882705e-07, + "loss": 1.1582, + "step": 500 + }, + { + "epoch": 0.49311023622047245, + "grad_norm": 19.995149612426758, + "learning_rate": 4.950298210735588e-07, + "loss": 1.4804, + "step": 501 + }, + { + "epoch": 0.4940944881889764, + "grad_norm": 16.41611671447754, + "learning_rate": 4.96023856858847e-07, + "loss": 1.2636, + "step": 502 + }, + { + "epoch": 0.4950787401574803, + "grad_norm": 16.084943771362305, + "learning_rate": 4.970178926441353e-07, + "loss": 1.0094, + "step": 503 + }, + { + "epoch": 0.49606299212598426, + "grad_norm": 14.031723022460938, + "learning_rate": 4.980119284294235e-07, + "loss": 0.7594, + "step": 504 + }, + { + "epoch": 0.4970472440944882, + "grad_norm": 18.90411949157715, + "learning_rate": 4.990059642147118e-07, + "loss": 1.2898, + "step": 505 + }, + { + "epoch": 0.49803149606299213, + "grad_norm": 19.5909423828125, + "learning_rate": 5.000000000000001e-07, + "loss": 1.3565, + "step": 506 + }, + { + "epoch": 0.49901574803149606, + "grad_norm": 16.011884689331055, + "learning_rate": 5.009940357852883e-07, + "loss": 1.0325, + "step": 507 + }, + { + "epoch": 0.5, + "grad_norm": 16.250404357910156, + "learning_rate": 5.019880715705767e-07, + "loss": 1.0519, + "step": 508 + }, + { + "epoch": 0.5009842519685039, + "grad_norm": 16.943971633911133, + "learning_rate": 5.029821073558649e-07, + "loss": 0.9802, + "step": 509 + }, + { + "epoch": 0.5019685039370079, + "grad_norm": 17.802446365356445, + "learning_rate": 5.039761431411532e-07, + "loss": 1.1117, + "step": 510 + }, + { + "epoch": 0.5029527559055118, + "grad_norm": 18.555269241333008, + "learning_rate": 5.049701789264414e-07, + "loss": 1.3585, + "step": 511 + }, + { + "epoch": 0.5039370078740157, + "grad_norm": 15.964585304260254, + "learning_rate": 5.059642147117297e-07, + "loss": 1.0381, + "step": 512 + }, + { + "epoch": 0.5049212598425197, + "grad_norm": 14.011932373046875, + "learning_rate": 5.06958250497018e-07, + "loss": 1.0171, + "step": 513 + }, + { + "epoch": 0.5059055118110236, + "grad_norm": 12.088032722473145, + "learning_rate": 5.079522862823062e-07, + "loss": 0.5678, + "step": 514 + }, + { + "epoch": 0.5068897637795275, + "grad_norm": 17.39454460144043, + "learning_rate": 5.089463220675945e-07, + "loss": 0.9347, + "step": 515 + }, + { + "epoch": 0.5078740157480315, + "grad_norm": 14.170563697814941, + "learning_rate": 5.099403578528828e-07, + "loss": 0.6305, + "step": 516 + }, + { + "epoch": 0.5088582677165354, + "grad_norm": 12.88972282409668, + "learning_rate": 5.10934393638171e-07, + "loss": 0.7072, + "step": 517 + }, + { + "epoch": 0.5098425196850394, + "grad_norm": 16.725210189819336, + "learning_rate": 5.119284294234593e-07, + "loss": 0.9746, + "step": 518 + }, + { + "epoch": 0.5108267716535433, + "grad_norm": 19.107908248901367, + "learning_rate": 5.129224652087476e-07, + "loss": 1.1782, + "step": 519 + }, + { + "epoch": 0.5118110236220472, + "grad_norm": 19.989774703979492, + "learning_rate": 5.139165009940358e-07, + "loss": 1.1354, + "step": 520 + }, + { + "epoch": 0.5127952755905512, + "grad_norm": 21.850637435913086, + "learning_rate": 5.149105367793241e-07, + "loss": 1.5752, + "step": 521 + }, + { + "epoch": 0.5137795275590551, + "grad_norm": 14.340449333190918, + "learning_rate": 5.159045725646124e-07, + "loss": 0.5952, + "step": 522 + }, + { + "epoch": 0.514763779527559, + "grad_norm": 18.608619689941406, + "learning_rate": 5.168986083499007e-07, + "loss": 1.1171, + "step": 523 + }, + { + "epoch": 0.515748031496063, + "grad_norm": 18.41301918029785, + "learning_rate": 5.178926441351889e-07, + "loss": 0.8234, + "step": 524 + }, + { + "epoch": 0.5167322834645669, + "grad_norm": 25.120445251464844, + "learning_rate": 5.188866799204771e-07, + "loss": 1.6701, + "step": 525 + }, + { + "epoch": 0.5177165354330708, + "grad_norm": 18.166357040405273, + "learning_rate": 5.198807157057656e-07, + "loss": 1.2111, + "step": 526 + }, + { + "epoch": 0.5187007874015748, + "grad_norm": 16.0269718170166, + "learning_rate": 5.208747514910538e-07, + "loss": 0.8299, + "step": 527 + }, + { + "epoch": 0.5196850393700787, + "grad_norm": 22.35710334777832, + "learning_rate": 5.21868787276342e-07, + "loss": 1.5734, + "step": 528 + }, + { + "epoch": 0.5206692913385826, + "grad_norm": 14.45018196105957, + "learning_rate": 5.228628230616304e-07, + "loss": 0.9172, + "step": 529 + }, + { + "epoch": 0.5216535433070866, + "grad_norm": 15.022900581359863, + "learning_rate": 5.238568588469186e-07, + "loss": 0.8025, + "step": 530 + }, + { + "epoch": 0.5226377952755905, + "grad_norm": 18.771406173706055, + "learning_rate": 5.248508946322069e-07, + "loss": 1.1499, + "step": 531 + }, + { + "epoch": 0.5236220472440944, + "grad_norm": 19.911235809326172, + "learning_rate": 5.258449304174951e-07, + "loss": 1.0328, + "step": 532 + }, + { + "epoch": 0.5246062992125984, + "grad_norm": 19.31245231628418, + "learning_rate": 5.268389662027834e-07, + "loss": 1.1305, + "step": 533 + }, + { + "epoch": 0.5255905511811023, + "grad_norm": 14.231760025024414, + "learning_rate": 5.278330019880717e-07, + "loss": 0.6715, + "step": 534 + }, + { + "epoch": 0.5265748031496063, + "grad_norm": 16.54633331298828, + "learning_rate": 5.288270377733599e-07, + "loss": 1.1361, + "step": 535 + }, + { + "epoch": 0.5275590551181102, + "grad_norm": 19.317935943603516, + "learning_rate": 5.298210735586483e-07, + "loss": 0.9132, + "step": 536 + }, + { + "epoch": 0.5285433070866141, + "grad_norm": 20.749347686767578, + "learning_rate": 5.308151093439365e-07, + "loss": 1.2195, + "step": 537 + }, + { + "epoch": 0.5295275590551181, + "grad_norm": 12.56061840057373, + "learning_rate": 5.318091451292247e-07, + "loss": 0.3731, + "step": 538 + }, + { + "epoch": 0.530511811023622, + "grad_norm": 17.94748878479004, + "learning_rate": 5.32803180914513e-07, + "loss": 1.0005, + "step": 539 + }, + { + "epoch": 0.531496062992126, + "grad_norm": 15.667882919311523, + "learning_rate": 5.337972166998013e-07, + "loss": 0.5519, + "step": 540 + }, + { + "epoch": 0.53248031496063, + "grad_norm": 18.331357955932617, + "learning_rate": 5.347912524850895e-07, + "loss": 0.7529, + "step": 541 + }, + { + "epoch": 0.5334645669291339, + "grad_norm": 28.6558895111084, + "learning_rate": 5.357852882703778e-07, + "loss": 1.7004, + "step": 542 + }, + { + "epoch": 0.5344488188976378, + "grad_norm": 25.073156356811523, + "learning_rate": 5.367793240556661e-07, + "loss": 1.4667, + "step": 543 + }, + { + "epoch": 0.5354330708661418, + "grad_norm": 19.546232223510742, + "learning_rate": 5.377733598409544e-07, + "loss": 0.8349, + "step": 544 + }, + { + "epoch": 0.5364173228346457, + "grad_norm": 24.613113403320312, + "learning_rate": 5.387673956262426e-07, + "loss": 1.5575, + "step": 545 + }, + { + "epoch": 0.5374015748031497, + "grad_norm": 23.96712875366211, + "learning_rate": 5.397614314115308e-07, + "loss": 1.1703, + "step": 546 + }, + { + "epoch": 0.5383858267716536, + "grad_norm": 18.8927001953125, + "learning_rate": 5.407554671968192e-07, + "loss": 1.01, + "step": 547 + }, + { + "epoch": 0.5393700787401575, + "grad_norm": 19.96955108642578, + "learning_rate": 5.417495029821074e-07, + "loss": 1.1114, + "step": 548 + }, + { + "epoch": 0.5403543307086615, + "grad_norm": 11.816988945007324, + "learning_rate": 5.427435387673957e-07, + "loss": 0.516, + "step": 549 + }, + { + "epoch": 0.5413385826771654, + "grad_norm": 18.070484161376953, + "learning_rate": 5.43737574552684e-07, + "loss": 1.0422, + "step": 550 + }, + { + "epoch": 0.5423228346456693, + "grad_norm": 16.6149845123291, + "learning_rate": 5.447316103379722e-07, + "loss": 1.078, + "step": 551 + }, + { + "epoch": 0.5433070866141733, + "grad_norm": 17.799449920654297, + "learning_rate": 5.457256461232605e-07, + "loss": 1.0573, + "step": 552 + }, + { + "epoch": 0.5442913385826772, + "grad_norm": 19.17674446105957, + "learning_rate": 5.467196819085487e-07, + "loss": 0.9754, + "step": 553 + }, + { + "epoch": 0.5452755905511811, + "grad_norm": 19.11102867126465, + "learning_rate": 5.477137176938371e-07, + "loss": 0.9227, + "step": 554 + }, + { + "epoch": 0.5462598425196851, + "grad_norm": 24.212202072143555, + "learning_rate": 5.487077534791253e-07, + "loss": 1.5012, + "step": 555 + }, + { + "epoch": 0.547244094488189, + "grad_norm": 16.03239631652832, + "learning_rate": 5.497017892644135e-07, + "loss": 1.0697, + "step": 556 + }, + { + "epoch": 0.5482283464566929, + "grad_norm": 21.717912673950195, + "learning_rate": 5.50695825049702e-07, + "loss": 1.4437, + "step": 557 + }, + { + "epoch": 0.5492125984251969, + "grad_norm": 18.13492202758789, + "learning_rate": 5.516898608349902e-07, + "loss": 1.0697, + "step": 558 + }, + { + "epoch": 0.5501968503937008, + "grad_norm": 15.775596618652344, + "learning_rate": 5.526838966202784e-07, + "loss": 0.8346, + "step": 559 + }, + { + "epoch": 0.5511811023622047, + "grad_norm": 16.541046142578125, + "learning_rate": 5.536779324055667e-07, + "loss": 0.6421, + "step": 560 + }, + { + "epoch": 0.5521653543307087, + "grad_norm": 16.008865356445312, + "learning_rate": 5.54671968190855e-07, + "loss": 0.6687, + "step": 561 + }, + { + "epoch": 0.5531496062992126, + "grad_norm": 18.417842864990234, + "learning_rate": 5.556660039761433e-07, + "loss": 0.982, + "step": 562 + }, + { + "epoch": 0.5541338582677166, + "grad_norm": 17.81726837158203, + "learning_rate": 5.566600397614315e-07, + "loss": 0.9299, + "step": 563 + }, + { + "epoch": 0.5551181102362205, + "grad_norm": 20.61170768737793, + "learning_rate": 5.576540755467198e-07, + "loss": 1.5852, + "step": 564 + }, + { + "epoch": 0.5561023622047244, + "grad_norm": 18.6957950592041, + "learning_rate": 5.586481113320081e-07, + "loss": 1.2132, + "step": 565 + }, + { + "epoch": 0.5570866141732284, + "grad_norm": 15.698697090148926, + "learning_rate": 5.596421471172963e-07, + "loss": 0.8426, + "step": 566 + }, + { + "epoch": 0.5580708661417323, + "grad_norm": 17.419189453125, + "learning_rate": 5.606361829025846e-07, + "loss": 1.0496, + "step": 567 + }, + { + "epoch": 0.5590551181102362, + "grad_norm": 15.717952728271484, + "learning_rate": 5.616302186878729e-07, + "loss": 1.0436, + "step": 568 + }, + { + "epoch": 0.5600393700787402, + "grad_norm": 13.559417724609375, + "learning_rate": 5.626242544731611e-07, + "loss": 0.806, + "step": 569 + }, + { + "epoch": 0.5610236220472441, + "grad_norm": 12.456329345703125, + "learning_rate": 5.636182902584494e-07, + "loss": 0.6396, + "step": 570 + }, + { + "epoch": 0.562007874015748, + "grad_norm": 21.304899215698242, + "learning_rate": 5.646123260437377e-07, + "loss": 1.6315, + "step": 571 + }, + { + "epoch": 0.562992125984252, + "grad_norm": 16.898534774780273, + "learning_rate": 5.656063618290259e-07, + "loss": 1.3286, + "step": 572 + }, + { + "epoch": 0.5639763779527559, + "grad_norm": 14.442305564880371, + "learning_rate": 5.666003976143142e-07, + "loss": 0.7682, + "step": 573 + }, + { + "epoch": 0.5649606299212598, + "grad_norm": 14.211771011352539, + "learning_rate": 5.675944333996024e-07, + "loss": 0.7861, + "step": 574 + }, + { + "epoch": 0.5659448818897638, + "grad_norm": 16.277984619140625, + "learning_rate": 5.685884691848908e-07, + "loss": 1.0368, + "step": 575 + }, + { + "epoch": 0.5669291338582677, + "grad_norm": 18.00242805480957, + "learning_rate": 5.69582504970179e-07, + "loss": 1.1497, + "step": 576 + }, + { + "epoch": 0.5679133858267716, + "grad_norm": 17.694448471069336, + "learning_rate": 5.705765407554672e-07, + "loss": 0.9691, + "step": 577 + }, + { + "epoch": 0.5688976377952756, + "grad_norm": 15.09469223022461, + "learning_rate": 5.715705765407556e-07, + "loss": 0.7447, + "step": 578 + }, + { + "epoch": 0.5698818897637795, + "grad_norm": 21.568681716918945, + "learning_rate": 5.725646123260438e-07, + "loss": 1.3933, + "step": 579 + }, + { + "epoch": 0.5708661417322834, + "grad_norm": 17.880390167236328, + "learning_rate": 5.735586481113321e-07, + "loss": 1.0668, + "step": 580 + }, + { + "epoch": 0.5718503937007874, + "grad_norm": 14.031826972961426, + "learning_rate": 5.745526838966203e-07, + "loss": 0.6065, + "step": 581 + }, + { + "epoch": 0.5728346456692913, + "grad_norm": 16.969873428344727, + "learning_rate": 5.755467196819086e-07, + "loss": 0.9566, + "step": 582 + }, + { + "epoch": 0.5738188976377953, + "grad_norm": 15.315662384033203, + "learning_rate": 5.765407554671969e-07, + "loss": 0.7957, + "step": 583 + }, + { + "epoch": 0.5748031496062992, + "grad_norm": 18.925649642944336, + "learning_rate": 5.775347912524851e-07, + "loss": 1.0232, + "step": 584 + }, + { + "epoch": 0.5757874015748031, + "grad_norm": 19.93893814086914, + "learning_rate": 5.785288270377734e-07, + "loss": 1.4559, + "step": 585 + }, + { + "epoch": 0.5767716535433071, + "grad_norm": 14.928529739379883, + "learning_rate": 5.795228628230618e-07, + "loss": 0.8003, + "step": 586 + }, + { + "epoch": 0.577755905511811, + "grad_norm": 18.931825637817383, + "learning_rate": 5.8051689860835e-07, + "loss": 0.9504, + "step": 587 + }, + { + "epoch": 0.5787401574803149, + "grad_norm": 21.6677188873291, + "learning_rate": 5.815109343936383e-07, + "loss": 1.5257, + "step": 588 + }, + { + "epoch": 0.5797244094488189, + "grad_norm": 13.008130073547363, + "learning_rate": 5.825049701789266e-07, + "loss": 0.5798, + "step": 589 + }, + { + "epoch": 0.5807086614173228, + "grad_norm": 17.05565071105957, + "learning_rate": 5.834990059642148e-07, + "loss": 0.8169, + "step": 590 + }, + { + "epoch": 0.5816929133858267, + "grad_norm": 17.504405975341797, + "learning_rate": 5.844930417495031e-07, + "loss": 1.1131, + "step": 591 + }, + { + "epoch": 0.5826771653543307, + "grad_norm": 19.08473014831543, + "learning_rate": 5.854870775347914e-07, + "loss": 1.2498, + "step": 592 + }, + { + "epoch": 0.5836614173228346, + "grad_norm": 16.938589096069336, + "learning_rate": 5.864811133200797e-07, + "loss": 0.8541, + "step": 593 + }, + { + "epoch": 0.5846456692913385, + "grad_norm": 18.441364288330078, + "learning_rate": 5.874751491053679e-07, + "loss": 1.0848, + "step": 594 + }, + { + "epoch": 0.5856299212598425, + "grad_norm": 16.72596549987793, + "learning_rate": 5.884691848906561e-07, + "loss": 0.8909, + "step": 595 + }, + { + "epoch": 0.5866141732283464, + "grad_norm": 15.525846481323242, + "learning_rate": 5.894632206759445e-07, + "loss": 0.7572, + "step": 596 + }, + { + "epoch": 0.5875984251968503, + "grad_norm": 22.291608810424805, + "learning_rate": 5.904572564612327e-07, + "loss": 1.3636, + "step": 597 + }, + { + "epoch": 0.5885826771653543, + "grad_norm": 14.620648384094238, + "learning_rate": 5.914512922465209e-07, + "loss": 0.8493, + "step": 598 + }, + { + "epoch": 0.5895669291338582, + "grad_norm": 15.47690486907959, + "learning_rate": 5.924453280318093e-07, + "loss": 0.9594, + "step": 599 + }, + { + "epoch": 0.5905511811023622, + "grad_norm": 21.12307357788086, + "learning_rate": 5.934393638170975e-07, + "loss": 1.1143, + "step": 600 + }, + { + "epoch": 0.5915354330708661, + "grad_norm": 14.08637809753418, + "learning_rate": 5.944333996023858e-07, + "loss": 0.7093, + "step": 601 + }, + { + "epoch": 0.59251968503937, + "grad_norm": 15.423602104187012, + "learning_rate": 5.95427435387674e-07, + "loss": 1.0542, + "step": 602 + }, + { + "epoch": 0.593503937007874, + "grad_norm": 17.396760940551758, + "learning_rate": 5.964214711729623e-07, + "loss": 1.0621, + "step": 603 + }, + { + "epoch": 0.594488188976378, + "grad_norm": 12.562100410461426, + "learning_rate": 5.974155069582506e-07, + "loss": 0.6916, + "step": 604 + }, + { + "epoch": 0.5954724409448819, + "grad_norm": 20.98981285095215, + "learning_rate": 5.984095427435388e-07, + "loss": 1.0125, + "step": 605 + }, + { + "epoch": 0.5964566929133859, + "grad_norm": 16.202804565429688, + "learning_rate": 5.994035785288272e-07, + "loss": 0.8425, + "step": 606 + }, + { + "epoch": 0.5974409448818898, + "grad_norm": 18.76996421813965, + "learning_rate": 6.003976143141154e-07, + "loss": 1.2868, + "step": 607 + }, + { + "epoch": 0.5984251968503937, + "grad_norm": 17.652177810668945, + "learning_rate": 6.013916500994036e-07, + "loss": 1.3505, + "step": 608 + }, + { + "epoch": 0.5994094488188977, + "grad_norm": 18.34078598022461, + "learning_rate": 6.02385685884692e-07, + "loss": 1.2699, + "step": 609 + }, + { + "epoch": 0.6003937007874016, + "grad_norm": 18.553659439086914, + "learning_rate": 6.033797216699802e-07, + "loss": 1.1798, + "step": 610 + }, + { + "epoch": 0.6013779527559056, + "grad_norm": 17.771137237548828, + "learning_rate": 6.043737574552684e-07, + "loss": 1.3607, + "step": 611 + }, + { + "epoch": 0.6023622047244095, + "grad_norm": 16.680641174316406, + "learning_rate": 6.053677932405567e-07, + "loss": 1.0807, + "step": 612 + }, + { + "epoch": 0.6023622047244095, + "eval_Qnli-dev_cosine_accuracy": 0.689453125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8408171534538269, + "eval_Qnli-dev_cosine_ap": 0.7075487103878315, + "eval_Qnli-dev_cosine_f1": 0.6811352253756261, + "eval_Qnli-dev_cosine_f1_threshold": 0.7815157175064087, + "eval_Qnli-dev_cosine_precision": 0.5619834710743802, + "eval_Qnli-dev_cosine_recall": 0.864406779661017, + "eval_Qnli-dev_dot_accuracy": 0.689453125, + "eval_Qnli-dev_dot_accuracy_threshold": 646.57861328125, + "eval_Qnli-dev_dot_ap": 0.7075572006690205, + "eval_Qnli-dev_dot_f1": 0.6799999999999999, + "eval_Qnli-dev_dot_f1_threshold": 600.8392333984375, + "eval_Qnli-dev_dot_precision": 0.5604395604395604, + "eval_Qnli-dev_dot_recall": 0.864406779661017, + "eval_Qnli-dev_euclidean_accuracy": 0.689453125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.646707534790039, + "eval_Qnli-dev_euclidean_ap": 0.7075432765222807, + "eval_Qnli-dev_euclidean_f1": 0.6811352253756261, + "eval_Qnli-dev_euclidean_f1_threshold": 18.331012725830078, + "eval_Qnli-dev_euclidean_precision": 0.5619834710743802, + "eval_Qnli-dev_euclidean_recall": 0.864406779661017, + "eval_Qnli-dev_manhattan_accuracy": 0.69140625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 346.4283752441406, + "eval_Qnli-dev_manhattan_ap": 0.7067005115857219, + "eval_Qnli-dev_manhattan_f1": 0.6834170854271356, + "eval_Qnli-dev_manhattan_f1_threshold": 402.6258544921875, + "eval_Qnli-dev_manhattan_precision": 0.5650969529085873, + "eval_Qnli-dev_manhattan_recall": 0.864406779661017, + "eval_Qnli-dev_max_accuracy": 0.69140625, + "eval_Qnli-dev_max_accuracy_threshold": 646.57861328125, + "eval_Qnli-dev_max_ap": 0.7075572006690205, + "eval_Qnli-dev_max_f1": 0.6834170854271356, + "eval_Qnli-dev_max_f1_threshold": 600.8392333984375, + "eval_Qnli-dev_max_precision": 0.5650969529085873, + "eval_Qnli-dev_max_recall": 0.864406779661017, + "eval_allNLI-dev_cosine_accuracy": 0.701171875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.932601809501648, + "eval_allNLI-dev_cosine_ap": 0.5129639279245027, + "eval_allNLI-dev_cosine_f1": 0.5454545454545455, + "eval_allNLI-dev_cosine_f1_threshold": 0.8405359387397766, + "eval_allNLI-dev_cosine_precision": 0.4016393442622951, + "eval_allNLI-dev_cosine_recall": 0.8497109826589595, + "eval_allNLI-dev_dot_accuracy": 0.701171875, + "eval_allNLI-dev_dot_accuracy_threshold": 717.2557373046875, + "eval_allNLI-dev_dot_ap": 0.5127119301647263, + "eval_allNLI-dev_dot_f1": 0.5454545454545455, + "eval_allNLI-dev_dot_f1_threshold": 646.4378662109375, + "eval_allNLI-dev_dot_precision": 0.4016393442622951, + "eval_allNLI-dev_dot_recall": 0.8497109826589595, + "eval_allNLI-dev_euclidean_accuracy": 0.701171875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.181873321533203, + "eval_allNLI-dev_euclidean_ap": 0.5129706884494644, + "eval_allNLI-dev_euclidean_f1": 0.5454545454545455, + "eval_allNLI-dev_euclidean_f1_threshold": 15.661436080932617, + "eval_allNLI-dev_euclidean_precision": 0.4016393442622951, + "eval_allNLI-dev_euclidean_recall": 0.8497109826589595, + "eval_allNLI-dev_manhattan_accuracy": 0.697265625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 224.18515014648438, + "eval_allNLI-dev_manhattan_ap": 0.5143470215129051, + "eval_allNLI-dev_manhattan_f1": 0.5483870967741935, + "eval_allNLI-dev_manhattan_f1_threshold": 355.3348388671875, + "eval_allNLI-dev_manhattan_precision": 0.3974025974025974, + "eval_allNLI-dev_manhattan_recall": 0.884393063583815, + "eval_allNLI-dev_max_accuracy": 0.701171875, + "eval_allNLI-dev_max_accuracy_threshold": 717.2557373046875, + "eval_allNLI-dev_max_ap": 0.5143470215129051, + "eval_allNLI-dev_max_f1": 0.5483870967741935, + "eval_allNLI-dev_max_f1_threshold": 646.4378662109375, + "eval_allNLI-dev_max_precision": 0.4016393442622951, + "eval_allNLI-dev_max_recall": 0.884393063583815, + "eval_loss": 1.2166765928268433, + "eval_runtime": 49.6646, + "eval_samples_per_second": 33.505, + "eval_sequential_score": 0.7075572006690205, + "eval_steps_per_second": 0.141, + "eval_sts-test_pearson_cosine": 0.5814025455181653, + "eval_sts-test_pearson_dot": 0.5813644503290595, + "eval_sts-test_pearson_euclidean": 0.5887110213836103, + "eval_sts-test_pearson_manhattan": 0.588548583748776, + "eval_sts-test_pearson_max": 0.5887110213836103, + "eval_sts-test_spearman_cosine": 0.5878996927156691, + "eval_sts-test_spearman_dot": 0.5878656370057502, + "eval_sts-test_spearman_euclidean": 0.5879079269477651, + "eval_sts-test_spearman_manhattan": 0.5879298998171083, + "eval_sts-test_spearman_max": 0.5879298998171083, + "step": 612 + }, + { + "epoch": 0.6033464566929134, + "grad_norm": 17.835203170776367, + "learning_rate": 6.06361829025845e-07, + "loss": 1.4339, + "step": 613 + }, + { + "epoch": 0.6043307086614174, + "grad_norm": 14.815842628479004, + "learning_rate": 6.073558648111333e-07, + "loss": 1.1194, + "step": 614 + }, + { + "epoch": 0.6053149606299213, + "grad_norm": 15.13090705871582, + "learning_rate": 6.083499005964215e-07, + "loss": 1.0682, + "step": 615 + }, + { + "epoch": 0.6062992125984252, + "grad_norm": 14.048788070678711, + "learning_rate": 6.093439363817097e-07, + "loss": 1.0429, + "step": 616 + }, + { + "epoch": 0.6072834645669292, + "grad_norm": 15.525943756103516, + "learning_rate": 6.103379721669982e-07, + "loss": 1.2554, + "step": 617 + }, + { + "epoch": 0.6082677165354331, + "grad_norm": 17.508243560791016, + "learning_rate": 6.113320079522864e-07, + "loss": 1.2466, + "step": 618 + }, + { + "epoch": 0.609251968503937, + "grad_norm": 15.177538871765137, + "learning_rate": 6.123260437375747e-07, + "loss": 1.1207, + "step": 619 + }, + { + "epoch": 0.610236220472441, + "grad_norm": 15.74117660522461, + "learning_rate": 6.13320079522863e-07, + "loss": 0.9822, + "step": 620 + }, + { + "epoch": 0.6112204724409449, + "grad_norm": 18.58333969116211, + "learning_rate": 6.143141153081512e-07, + "loss": 1.7369, + "step": 621 + }, + { + "epoch": 0.6122047244094488, + "grad_norm": 21.2359561920166, + "learning_rate": 6.153081510934395e-07, + "loss": 1.3305, + "step": 622 + }, + { + "epoch": 0.6131889763779528, + "grad_norm": 17.642099380493164, + "learning_rate": 6.163021868787277e-07, + "loss": 0.9064, + "step": 623 + }, + { + "epoch": 0.6141732283464567, + "grad_norm": 14.812031745910645, + "learning_rate": 6.172962226640161e-07, + "loss": 0.7123, + "step": 624 + }, + { + "epoch": 0.6151574803149606, + "grad_norm": 15.9660062789917, + "learning_rate": 6.182902584493043e-07, + "loss": 0.7461, + "step": 625 + }, + { + "epoch": 0.6161417322834646, + "grad_norm": 17.61859130859375, + "learning_rate": 6.192842942345925e-07, + "loss": 0.8082, + "step": 626 + }, + { + "epoch": 0.6171259842519685, + "grad_norm": 16.456764221191406, + "learning_rate": 6.202783300198809e-07, + "loss": 1.0113, + "step": 627 + }, + { + "epoch": 0.6181102362204725, + "grad_norm": 16.365720748901367, + "learning_rate": 6.212723658051691e-07, + "loss": 0.9483, + "step": 628 + }, + { + "epoch": 0.6190944881889764, + "grad_norm": 18.512392044067383, + "learning_rate": 6.222664015904573e-07, + "loss": 0.9269, + "step": 629 + }, + { + "epoch": 0.6200787401574803, + "grad_norm": 21.816940307617188, + "learning_rate": 6.232604373757457e-07, + "loss": 1.3134, + "step": 630 + }, + { + "epoch": 0.6210629921259843, + "grad_norm": 18.167619705200195, + "learning_rate": 6.242544731610339e-07, + "loss": 0.7253, + "step": 631 + }, + { + "epoch": 0.6220472440944882, + "grad_norm": 14.523139953613281, + "learning_rate": 6.252485089463222e-07, + "loss": 0.809, + "step": 632 + }, + { + "epoch": 0.6230314960629921, + "grad_norm": 18.33440399169922, + "learning_rate": 6.262425447316104e-07, + "loss": 1.2514, + "step": 633 + }, + { + "epoch": 0.6240157480314961, + "grad_norm": 13.027684211730957, + "learning_rate": 6.272365805168987e-07, + "loss": 0.6718, + "step": 634 + }, + { + "epoch": 0.625, + "grad_norm": 15.819523811340332, + "learning_rate": 6.28230616302187e-07, + "loss": 0.6658, + "step": 635 + }, + { + "epoch": 0.6259842519685039, + "grad_norm": 21.912124633789062, + "learning_rate": 6.292246520874752e-07, + "loss": 1.3988, + "step": 636 + }, + { + "epoch": 0.6269685039370079, + "grad_norm": 12.767960548400879, + "learning_rate": 6.302186878727636e-07, + "loss": 0.7358, + "step": 637 + }, + { + "epoch": 0.6279527559055118, + "grad_norm": 16.3560848236084, + "learning_rate": 6.312127236580518e-07, + "loss": 0.7797, + "step": 638 + }, + { + "epoch": 0.6289370078740157, + "grad_norm": 16.61944007873535, + "learning_rate": 6.3220675944334e-07, + "loss": 1.048, + "step": 639 + }, + { + "epoch": 0.6299212598425197, + "grad_norm": 14.68630599975586, + "learning_rate": 6.332007952286283e-07, + "loss": 0.9559, + "step": 640 + }, + { + "epoch": 0.6309055118110236, + "grad_norm": 10.76396369934082, + "learning_rate": 6.341948310139166e-07, + "loss": 0.4561, + "step": 641 + }, + { + "epoch": 0.6318897637795275, + "grad_norm": 15.87311840057373, + "learning_rate": 6.351888667992048e-07, + "loss": 1.1078, + "step": 642 + }, + { + "epoch": 0.6328740157480315, + "grad_norm": 17.113353729248047, + "learning_rate": 6.361829025844931e-07, + "loss": 0.9724, + "step": 643 + }, + { + "epoch": 0.6338582677165354, + "grad_norm": 15.505386352539062, + "learning_rate": 6.371769383697813e-07, + "loss": 1.0702, + "step": 644 + }, + { + "epoch": 0.6348425196850394, + "grad_norm": 14.685168266296387, + "learning_rate": 6.381709741550698e-07, + "loss": 1.0911, + "step": 645 + }, + { + "epoch": 0.6358267716535433, + "grad_norm": 19.55362319946289, + "learning_rate": 6.39165009940358e-07, + "loss": 1.1584, + "step": 646 + }, + { + "epoch": 0.6368110236220472, + "grad_norm": 19.717212677001953, + "learning_rate": 6.401590457256462e-07, + "loss": 0.9063, + "step": 647 + }, + { + "epoch": 0.6377952755905512, + "grad_norm": 11.899023056030273, + "learning_rate": 6.411530815109346e-07, + "loss": 0.5036, + "step": 648 + }, + { + "epoch": 0.6387795275590551, + "grad_norm": 16.72358512878418, + "learning_rate": 6.421471172962228e-07, + "loss": 0.8331, + "step": 649 + }, + { + "epoch": 0.639763779527559, + "grad_norm": 16.15726661682129, + "learning_rate": 6.431411530815111e-07, + "loss": 1.0772, + "step": 650 + }, + { + "epoch": 0.640748031496063, + "grad_norm": 17.20926284790039, + "learning_rate": 6.441351888667994e-07, + "loss": 0.7466, + "step": 651 + }, + { + "epoch": 0.6417322834645669, + "grad_norm": 18.99711036682129, + "learning_rate": 6.451292246520876e-07, + "loss": 1.1614, + "step": 652 + }, + { + "epoch": 0.6427165354330708, + "grad_norm": 13.12380313873291, + "learning_rate": 6.461232604373759e-07, + "loss": 0.6319, + "step": 653 + }, + { + "epoch": 0.6437007874015748, + "grad_norm": 14.912959098815918, + "learning_rate": 6.471172962226641e-07, + "loss": 0.7519, + "step": 654 + }, + { + "epoch": 0.6446850393700787, + "grad_norm": 18.82720184326172, + "learning_rate": 6.481113320079524e-07, + "loss": 1.1067, + "step": 655 + }, + { + "epoch": 0.6456692913385826, + "grad_norm": 20.76983070373535, + "learning_rate": 6.491053677932407e-07, + "loss": 1.2561, + "step": 656 + }, + { + "epoch": 0.6466535433070866, + "grad_norm": 14.362175941467285, + "learning_rate": 6.500994035785289e-07, + "loss": 0.6509, + "step": 657 + }, + { + "epoch": 0.6476377952755905, + "grad_norm": 17.300748825073242, + "learning_rate": 6.510934393638173e-07, + "loss": 1.0201, + "step": 658 + }, + { + "epoch": 0.6486220472440944, + "grad_norm": 23.880773544311523, + "learning_rate": 6.520874751491055e-07, + "loss": 1.6782, + "step": 659 + }, + { + "epoch": 0.6496062992125984, + "grad_norm": 22.65629005432129, + "learning_rate": 6.530815109343937e-07, + "loss": 1.3718, + "step": 660 + }, + { + "epoch": 0.6505905511811023, + "grad_norm": 16.830120086669922, + "learning_rate": 6.54075546719682e-07, + "loss": 0.6883, + "step": 661 + }, + { + "epoch": 0.6515748031496063, + "grad_norm": 18.652036666870117, + "learning_rate": 6.550695825049703e-07, + "loss": 1.0951, + "step": 662 + }, + { + "epoch": 0.6525590551181102, + "grad_norm": 20.634252548217773, + "learning_rate": 6.560636182902586e-07, + "loss": 1.2543, + "step": 663 + }, + { + "epoch": 0.6535433070866141, + "grad_norm": 17.44988250732422, + "learning_rate": 6.570576540755468e-07, + "loss": 1.2208, + "step": 664 + }, + { + "epoch": 0.6545275590551181, + "grad_norm": 14.21064567565918, + "learning_rate": 6.580516898608351e-07, + "loss": 0.6009, + "step": 665 + }, + { + "epoch": 0.655511811023622, + "grad_norm": 17.813508987426758, + "learning_rate": 6.590457256461234e-07, + "loss": 1.1146, + "step": 666 + }, + { + "epoch": 0.656496062992126, + "grad_norm": 16.94171142578125, + "learning_rate": 6.600397614314116e-07, + "loss": 1.0411, + "step": 667 + }, + { + "epoch": 0.65748031496063, + "grad_norm": 14.979653358459473, + "learning_rate": 6.610337972166998e-07, + "loss": 0.6938, + "step": 668 + }, + { + "epoch": 0.6584645669291339, + "grad_norm": 20.234113693237305, + "learning_rate": 6.620278330019882e-07, + "loss": 1.0415, + "step": 669 + }, + { + "epoch": 0.6594488188976378, + "grad_norm": 12.613083839416504, + "learning_rate": 6.630218687872764e-07, + "loss": 0.4991, + "step": 670 + }, + { + "epoch": 0.6604330708661418, + "grad_norm": 20.836713790893555, + "learning_rate": 6.640159045725647e-07, + "loss": 1.4716, + "step": 671 + }, + { + "epoch": 0.6614173228346457, + "grad_norm": 16.035064697265625, + "learning_rate": 6.65009940357853e-07, + "loss": 0.745, + "step": 672 + }, + { + "epoch": 0.6624015748031497, + "grad_norm": 21.638883590698242, + "learning_rate": 6.660039761431412e-07, + "loss": 1.5687, + "step": 673 + }, + { + "epoch": 0.6633858267716536, + "grad_norm": 16.04642105102539, + "learning_rate": 6.669980119284295e-07, + "loss": 0.7606, + "step": 674 + }, + { + "epoch": 0.6643700787401575, + "grad_norm": 11.20046615600586, + "learning_rate": 6.679920477137177e-07, + "loss": 0.2446, + "step": 675 + }, + { + "epoch": 0.6653543307086615, + "grad_norm": 17.451993942260742, + "learning_rate": 6.689860834990062e-07, + "loss": 0.4829, + "step": 676 + }, + { + "epoch": 0.6663385826771654, + "grad_norm": 20.657482147216797, + "learning_rate": 6.699801192842944e-07, + "loss": 1.0112, + "step": 677 + }, + { + "epoch": 0.6673228346456693, + "grad_norm": 20.848297119140625, + "learning_rate": 6.709741550695826e-07, + "loss": 1.3718, + "step": 678 + }, + { + "epoch": 0.6683070866141733, + "grad_norm": 22.662883758544922, + "learning_rate": 6.71968190854871e-07, + "loss": 1.3441, + "step": 679 + }, + { + "epoch": 0.6692913385826772, + "grad_norm": 12.292403221130371, + "learning_rate": 6.729622266401592e-07, + "loss": 0.5089, + "step": 680 + }, + { + "epoch": 0.6702755905511811, + "grad_norm": 19.724782943725586, + "learning_rate": 6.739562624254474e-07, + "loss": 0.9052, + "step": 681 + }, + { + "epoch": 0.6712598425196851, + "grad_norm": 15.259573936462402, + "learning_rate": 6.749502982107357e-07, + "loss": 0.7006, + "step": 682 + }, + { + "epoch": 0.672244094488189, + "grad_norm": 21.092082977294922, + "learning_rate": 6.75944333996024e-07, + "loss": 1.2755, + "step": 683 + }, + { + "epoch": 0.6732283464566929, + "grad_norm": 16.612253189086914, + "learning_rate": 6.769383697813123e-07, + "loss": 0.8308, + "step": 684 + }, + { + "epoch": 0.6742125984251969, + "grad_norm": 13.980706214904785, + "learning_rate": 6.779324055666005e-07, + "loss": 0.797, + "step": 685 + }, + { + "epoch": 0.6751968503937008, + "grad_norm": 13.144355773925781, + "learning_rate": 6.789264413518888e-07, + "loss": 0.5807, + "step": 686 + }, + { + "epoch": 0.6761811023622047, + "grad_norm": 17.212087631225586, + "learning_rate": 6.799204771371771e-07, + "loss": 0.9666, + "step": 687 + }, + { + "epoch": 0.6771653543307087, + "grad_norm": 16.941543579101562, + "learning_rate": 6.809145129224653e-07, + "loss": 1.0587, + "step": 688 + }, + { + "epoch": 0.6781496062992126, + "grad_norm": 16.685169219970703, + "learning_rate": 6.819085487077536e-07, + "loss": 1.1675, + "step": 689 + }, + { + "epoch": 0.6791338582677166, + "grad_norm": 16.485889434814453, + "learning_rate": 6.829025844930419e-07, + "loss": 0.725, + "step": 690 + }, + { + "epoch": 0.6801181102362205, + "grad_norm": 15.698780059814453, + "learning_rate": 6.838966202783301e-07, + "loss": 0.9958, + "step": 691 + }, + { + "epoch": 0.6811023622047244, + "grad_norm": 20.153303146362305, + "learning_rate": 6.848906560636184e-07, + "loss": 1.13, + "step": 692 + }, + { + "epoch": 0.6820866141732284, + "grad_norm": 19.95404052734375, + "learning_rate": 6.858846918489067e-07, + "loss": 1.6021, + "step": 693 + }, + { + "epoch": 0.6830708661417323, + "grad_norm": 14.196332931518555, + "learning_rate": 6.86878727634195e-07, + "loss": 0.8968, + "step": 694 + }, + { + "epoch": 0.6840551181102362, + "grad_norm": 19.394855499267578, + "learning_rate": 6.878727634194832e-07, + "loss": 0.9741, + "step": 695 + }, + { + "epoch": 0.6850393700787402, + "grad_norm": 22.04231834411621, + "learning_rate": 6.888667992047714e-07, + "loss": 1.1929, + "step": 696 + }, + { + "epoch": 0.6860236220472441, + "grad_norm": 13.484358787536621, + "learning_rate": 6.898608349900598e-07, + "loss": 0.6117, + "step": 697 + }, + { + "epoch": 0.687007874015748, + "grad_norm": 18.666522979736328, + "learning_rate": 6.90854870775348e-07, + "loss": 0.9741, + "step": 698 + }, + { + "epoch": 0.687992125984252, + "grad_norm": 16.874784469604492, + "learning_rate": 6.918489065606362e-07, + "loss": 0.9963, + "step": 699 + }, + { + "epoch": 0.6889763779527559, + "grad_norm": 12.441784858703613, + "learning_rate": 6.928429423459246e-07, + "loss": 0.6098, + "step": 700 + }, + { + "epoch": 0.6899606299212598, + "grad_norm": 15.19897747039795, + "learning_rate": 6.938369781312128e-07, + "loss": 0.9233, + "step": 701 + }, + { + "epoch": 0.6909448818897638, + "grad_norm": 20.48088264465332, + "learning_rate": 6.948310139165011e-07, + "loss": 1.4652, + "step": 702 + }, + { + "epoch": 0.6919291338582677, + "grad_norm": 19.183616638183594, + "learning_rate": 6.958250497017893e-07, + "loss": 1.3325, + "step": 703 + }, + { + "epoch": 0.6929133858267716, + "grad_norm": 19.812631607055664, + "learning_rate": 6.968190854870776e-07, + "loss": 1.1559, + "step": 704 + }, + { + "epoch": 0.6938976377952756, + "grad_norm": 16.60248565673828, + "learning_rate": 6.97813121272366e-07, + "loss": 1.021, + "step": 705 + }, + { + "epoch": 0.6948818897637795, + "grad_norm": 19.515586853027344, + "learning_rate": 6.988071570576542e-07, + "loss": 1.1437, + "step": 706 + }, + { + "epoch": 0.6958661417322834, + "grad_norm": 22.46136474609375, + "learning_rate": 6.998011928429426e-07, + "loss": 1.5533, + "step": 707 + }, + { + "epoch": 0.6968503937007874, + "grad_norm": 12.519843101501465, + "learning_rate": 7.007952286282308e-07, + "loss": 0.4733, + "step": 708 + }, + { + "epoch": 0.6978346456692913, + "grad_norm": 18.812881469726562, + "learning_rate": 7.01789264413519e-07, + "loss": 1.4539, + "step": 709 + }, + { + "epoch": 0.6988188976377953, + "grad_norm": 19.017959594726562, + "learning_rate": 7.027833001988073e-07, + "loss": 1.132, + "step": 710 + }, + { + "epoch": 0.6998031496062992, + "grad_norm": 19.10792350769043, + "learning_rate": 7.037773359840956e-07, + "loss": 1.315, + "step": 711 + }, + { + "epoch": 0.7007874015748031, + "grad_norm": 14.288207054138184, + "learning_rate": 7.047713717693838e-07, + "loss": 0.6671, + "step": 712 + }, + { + "epoch": 0.7017716535433071, + "grad_norm": 16.61151695251465, + "learning_rate": 7.057654075546721e-07, + "loss": 1.0689, + "step": 713 + }, + { + "epoch": 0.702755905511811, + "grad_norm": 17.917814254760742, + "learning_rate": 7.067594433399604e-07, + "loss": 1.2344, + "step": 714 + }, + { + "epoch": 0.7037401574803149, + "grad_norm": 14.60793399810791, + "learning_rate": 7.077534791252487e-07, + "loss": 0.9918, + "step": 715 + }, + { + "epoch": 0.7047244094488189, + "grad_norm": 12.193852424621582, + "learning_rate": 7.087475149105369e-07, + "loss": 0.6537, + "step": 716 + }, + { + "epoch": 0.7057086614173228, + "grad_norm": 17.281579971313477, + "learning_rate": 7.097415506958251e-07, + "loss": 1.4362, + "step": 717 + }, + { + "epoch": 0.7066929133858267, + "grad_norm": 17.89337730407715, + "learning_rate": 7.107355864811135e-07, + "loss": 1.2486, + "step": 718 + }, + { + "epoch": 0.7076771653543307, + "grad_norm": 12.966585159301758, + "learning_rate": 7.117296222664017e-07, + "loss": 0.6777, + "step": 719 + }, + { + "epoch": 0.7086614173228346, + "grad_norm": 13.884028434753418, + "learning_rate": 7.1272365805169e-07, + "loss": 0.965, + "step": 720 + }, + { + "epoch": 0.7096456692913385, + "grad_norm": 17.88764762878418, + "learning_rate": 7.137176938369783e-07, + "loss": 1.1881, + "step": 721 + }, + { + "epoch": 0.7106299212598425, + "grad_norm": 17.052560806274414, + "learning_rate": 7.147117296222665e-07, + "loss": 1.2064, + "step": 722 + }, + { + "epoch": 0.7116141732283464, + "grad_norm": 12.48928165435791, + "learning_rate": 7.157057654075548e-07, + "loss": 0.5049, + "step": 723 + }, + { + "epoch": 0.7125984251968503, + "grad_norm": 13.620563507080078, + "learning_rate": 7.16699801192843e-07, + "loss": 0.7258, + "step": 724 + }, + { + "epoch": 0.7135826771653543, + "grad_norm": 11.101369857788086, + "learning_rate": 7.176938369781313e-07, + "loss": 0.458, + "step": 725 + }, + { + "epoch": 0.7145669291338582, + "grad_norm": 16.40080451965332, + "learning_rate": 7.186878727634196e-07, + "loss": 1.0756, + "step": 726 + }, + { + "epoch": 0.7155511811023622, + "grad_norm": 14.630434036254883, + "learning_rate": 7.196819085487078e-07, + "loss": 0.8171, + "step": 727 + }, + { + "epoch": 0.7165354330708661, + "grad_norm": 15.544898986816406, + "learning_rate": 7.206759443339962e-07, + "loss": 0.786, + "step": 728 + }, + { + "epoch": 0.71751968503937, + "grad_norm": 21.253652572631836, + "learning_rate": 7.216699801192844e-07, + "loss": 1.3556, + "step": 729 + }, + { + "epoch": 0.718503937007874, + "grad_norm": 19.556493759155273, + "learning_rate": 7.226640159045726e-07, + "loss": 1.181, + "step": 730 + }, + { + "epoch": 0.719488188976378, + "grad_norm": 19.362489700317383, + "learning_rate": 7.236580516898609e-07, + "loss": 1.0563, + "step": 731 + }, + { + "epoch": 0.7204724409448819, + "grad_norm": 10.839325904846191, + "learning_rate": 7.246520874751492e-07, + "loss": 0.5951, + "step": 732 + }, + { + "epoch": 0.7214566929133859, + "grad_norm": 18.89388084411621, + "learning_rate": 7.256461232604375e-07, + "loss": 0.8533, + "step": 733 + }, + { + "epoch": 0.7224409448818898, + "grad_norm": 15.119179725646973, + "learning_rate": 7.266401590457257e-07, + "loss": 0.6561, + "step": 734 + }, + { + "epoch": 0.7234251968503937, + "grad_norm": 17.914310455322266, + "learning_rate": 7.27634194831014e-07, + "loss": 1.1081, + "step": 735 + }, + { + "epoch": 0.7244094488188977, + "grad_norm": 12.161086082458496, + "learning_rate": 7.286282306163024e-07, + "loss": 0.6016, + "step": 736 + }, + { + "epoch": 0.7253937007874016, + "grad_norm": 12.749862670898438, + "learning_rate": 7.296222664015906e-07, + "loss": 0.6155, + "step": 737 + }, + { + "epoch": 0.7263779527559056, + "grad_norm": 6.194193363189697, + "learning_rate": 7.306163021868788e-07, + "loss": 0.2202, + "step": 738 + }, + { + "epoch": 0.7273622047244095, + "grad_norm": 17.684555053710938, + "learning_rate": 7.316103379721672e-07, + "loss": 1.1072, + "step": 739 + }, + { + "epoch": 0.7283464566929134, + "grad_norm": 15.796165466308594, + "learning_rate": 7.326043737574554e-07, + "loss": 1.0147, + "step": 740 + }, + { + "epoch": 0.7293307086614174, + "grad_norm": 7.441659450531006, + "learning_rate": 7.335984095427437e-07, + "loss": 0.2117, + "step": 741 + }, + { + "epoch": 0.7303149606299213, + "grad_norm": 23.53545570373535, + "learning_rate": 7.34592445328032e-07, + "loss": 1.3508, + "step": 742 + }, + { + "epoch": 0.7312992125984252, + "grad_norm": 17.080326080322266, + "learning_rate": 7.355864811133202e-07, + "loss": 0.7085, + "step": 743 + }, + { + "epoch": 0.7322834645669292, + "grad_norm": 15.710729598999023, + "learning_rate": 7.365805168986085e-07, + "loss": 0.7357, + "step": 744 + }, + { + "epoch": 0.7332677165354331, + "grad_norm": 18.150358200073242, + "learning_rate": 7.375745526838967e-07, + "loss": 1.0121, + "step": 745 + }, + { + "epoch": 0.734251968503937, + "grad_norm": 24.2531795501709, + "learning_rate": 7.385685884691851e-07, + "loss": 1.2527, + "step": 746 + }, + { + "epoch": 0.735236220472441, + "grad_norm": 22.69845962524414, + "learning_rate": 7.395626242544733e-07, + "loss": 1.5227, + "step": 747 + }, + { + "epoch": 0.7362204724409449, + "grad_norm": 18.442115783691406, + "learning_rate": 7.405566600397615e-07, + "loss": 1.2253, + "step": 748 + }, + { + "epoch": 0.7372047244094488, + "grad_norm": 17.732175827026367, + "learning_rate": 7.415506958250499e-07, + "loss": 0.8419, + "step": 749 + }, + { + "epoch": 0.7381889763779528, + "grad_norm": 12.847253799438477, + "learning_rate": 7.425447316103381e-07, + "loss": 0.5649, + "step": 750 + }, + { + "epoch": 0.7391732283464567, + "grad_norm": 17.360685348510742, + "learning_rate": 7.435387673956263e-07, + "loss": 1.3501, + "step": 751 + }, + { + "epoch": 0.7401574803149606, + "grad_norm": 13.992162704467773, + "learning_rate": 7.445328031809146e-07, + "loss": 1.042, + "step": 752 + }, + { + "epoch": 0.7411417322834646, + "grad_norm": 16.175880432128906, + "learning_rate": 7.455268389662029e-07, + "loss": 1.1964, + "step": 753 + }, + { + "epoch": 0.7421259842519685, + "grad_norm": 17.538881301879883, + "learning_rate": 7.465208747514912e-07, + "loss": 1.1352, + "step": 754 + }, + { + "epoch": 0.7431102362204725, + "grad_norm": 15.056358337402344, + "learning_rate": 7.475149105367794e-07, + "loss": 0.8928, + "step": 755 + }, + { + "epoch": 0.7440944881889764, + "grad_norm": 15.273807525634766, + "learning_rate": 7.485089463220677e-07, + "loss": 0.7438, + "step": 756 + }, + { + "epoch": 0.7450787401574803, + "grad_norm": 18.479049682617188, + "learning_rate": 7.49502982107356e-07, + "loss": 1.4773, + "step": 757 + }, + { + "epoch": 0.7460629921259843, + "grad_norm": 16.95153045654297, + "learning_rate": 7.504970178926442e-07, + "loss": 1.196, + "step": 758 + }, + { + "epoch": 0.7470472440944882, + "grad_norm": 18.301109313964844, + "learning_rate": 7.514910536779325e-07, + "loss": 1.1562, + "step": 759 + }, + { + "epoch": 0.7480314960629921, + "grad_norm": 13.751684188842773, + "learning_rate": 7.524850894632208e-07, + "loss": 0.8362, + "step": 760 + }, + { + "epoch": 0.7490157480314961, + "grad_norm": 14.711894035339355, + "learning_rate": 7.53479125248509e-07, + "loss": 0.904, + "step": 761 + }, + { + "epoch": 0.75, + "grad_norm": 15.363059043884277, + "learning_rate": 7.544731610337973e-07, + "loss": 0.855, + "step": 762 + }, + { + "epoch": 0.7509842519685039, + "grad_norm": 14.737297058105469, + "learning_rate": 7.554671968190856e-07, + "loss": 0.748, + "step": 763 + }, + { + "epoch": 0.7519685039370079, + "grad_norm": 13.840580940246582, + "learning_rate": 7.56461232604374e-07, + "loss": 0.6261, + "step": 764 + }, + { + "epoch": 0.7529527559055118, + "grad_norm": 17.710996627807617, + "learning_rate": 7.574552683896621e-07, + "loss": 1.1903, + "step": 765 + }, + { + "epoch": 0.7529527559055118, + "eval_Qnli-dev_cosine_accuracy": 0.689453125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8274233937263489, + "eval_Qnli-dev_cosine_ap": 0.7122261593973184, + "eval_Qnli-dev_cosine_f1": 0.6847826086956521, + "eval_Qnli-dev_cosine_f1_threshold": 0.7854544520378113, + "eval_Qnli-dev_cosine_precision": 0.5981012658227848, + "eval_Qnli-dev_cosine_recall": 0.8008474576271186, + "eval_Qnli-dev_dot_accuracy": 0.6875, + "eval_Qnli-dev_dot_accuracy_threshold": 636.4744262695312, + "eval_Qnli-dev_dot_ap": 0.7120508993892436, + "eval_Qnli-dev_dot_f1": 0.6847826086956521, + "eval_Qnli-dev_dot_f1_threshold": 604.311767578125, + "eval_Qnli-dev_dot_precision": 0.5981012658227848, + "eval_Qnli-dev_dot_recall": 0.8008474576271186, + "eval_Qnli-dev_euclidean_accuracy": 0.689453125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 16.29575538635254, + "eval_Qnli-dev_euclidean_ap": 0.7122614787233053, + "eval_Qnli-dev_euclidean_f1": 0.6847826086956521, + "eval_Qnli-dev_euclidean_f1_threshold": 18.169567108154297, + "eval_Qnli-dev_euclidean_precision": 0.5981012658227848, + "eval_Qnli-dev_euclidean_recall": 0.8008474576271186, + "eval_Qnli-dev_manhattan_accuracy": 0.685546875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 363.32275390625, + "eval_Qnli-dev_manhattan_ap": 0.7106099609248304, + "eval_Qnli-dev_manhattan_f1": 0.6798561151079136, + "eval_Qnli-dev_manhattan_f1_threshold": 403.3307800292969, + "eval_Qnli-dev_manhattan_precision": 0.590625, + "eval_Qnli-dev_manhattan_recall": 0.8008474576271186, + "eval_Qnli-dev_max_accuracy": 0.689453125, + "eval_Qnli-dev_max_accuracy_threshold": 636.4744262695312, + "eval_Qnli-dev_max_ap": 0.7122614787233053, + "eval_Qnli-dev_max_f1": 0.6847826086956521, + "eval_Qnli-dev_max_f1_threshold": 604.311767578125, + "eval_Qnli-dev_max_precision": 0.5981012658227848, + "eval_Qnli-dev_max_recall": 0.8008474576271186, + "eval_allNLI-dev_cosine_accuracy": 0.69921875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9163634777069092, + "eval_allNLI-dev_cosine_ap": 0.5203950730799954, + "eval_allNLI-dev_cosine_f1": 0.5397727272727272, + "eval_allNLI-dev_cosine_f1_threshold": 0.8912649154663086, + "eval_allNLI-dev_cosine_precision": 0.5307262569832403, + "eval_allNLI-dev_cosine_recall": 0.5491329479768786, + "eval_allNLI-dev_dot_accuracy": 0.69921875, + "eval_allNLI-dev_dot_accuracy_threshold": 704.9437255859375, + "eval_allNLI-dev_dot_ap": 0.5204314509187654, + "eval_allNLI-dev_dot_f1": 0.5397727272727272, + "eval_allNLI-dev_dot_f1_threshold": 685.6298217773438, + "eval_allNLI-dev_dot_precision": 0.5307262569832403, + "eval_allNLI-dev_dot_recall": 0.5491329479768786, + "eval_allNLI-dev_euclidean_accuracy": 0.69921875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.343721389770508, + "eval_allNLI-dev_euclidean_ap": 0.5204045464403957, + "eval_allNLI-dev_euclidean_f1": 0.5397727272727272, + "eval_allNLI-dev_euclidean_f1_threshold": 12.934247970581055, + "eval_allNLI-dev_euclidean_precision": 0.5307262569832403, + "eval_allNLI-dev_euclidean_recall": 0.5491329479768786, + "eval_allNLI-dev_manhattan_accuracy": 0.69921875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 250.848388671875, + "eval_allNLI-dev_manhattan_ap": 0.5196278189093784, + "eval_allNLI-dev_manhattan_f1": 0.538888888888889, + "eval_allNLI-dev_manhattan_f1_threshold": 287.8966064453125, + "eval_allNLI-dev_manhattan_precision": 0.5187165775401069, + "eval_allNLI-dev_manhattan_recall": 0.5606936416184971, + "eval_allNLI-dev_max_accuracy": 0.69921875, + "eval_allNLI-dev_max_accuracy_threshold": 704.9437255859375, + "eval_allNLI-dev_max_ap": 0.5204314509187654, + "eval_allNLI-dev_max_f1": 0.5397727272727272, + "eval_allNLI-dev_max_f1_threshold": 685.6298217773438, + "eval_allNLI-dev_max_precision": 0.5307262569832403, + "eval_allNLI-dev_max_recall": 0.5606936416184971, + "eval_loss": 1.1806925535202026, + "eval_runtime": 50.0974, + "eval_samples_per_second": 33.215, + "eval_sequential_score": 0.7122614787233053, + "eval_steps_per_second": 0.14, + "eval_sts-test_pearson_cosine": 0.5663924244809233, + "eval_sts-test_pearson_dot": 0.5664900588871681, + "eval_sts-test_pearson_euclidean": 0.5785108788501522, + "eval_sts-test_pearson_manhattan": 0.579538083337237, + "eval_sts-test_pearson_max": 0.579538083337237, + "eval_sts-test_spearman_cosine": 0.5774005992806329, + "eval_sts-test_spearman_dot": 0.577422795906283, + "eval_sts-test_spearman_euclidean": 0.5773859208668966, + "eval_sts-test_spearman_manhattan": 0.5777711397249536, + "eval_sts-test_spearman_max": 0.5777711397249536, + "step": 765 + }, + { + "epoch": 0.7539370078740157, + "grad_norm": 15.706062316894531, + "learning_rate": 7.584493041749503e-07, + "loss": 0.8415, + "step": 766 + }, + { + "epoch": 0.7549212598425197, + "grad_norm": 14.501514434814453, + "learning_rate": 7.594433399602388e-07, + "loss": 0.712, + "step": 767 + }, + { + "epoch": 0.7559055118110236, + "grad_norm": 18.915172576904297, + "learning_rate": 7.60437375745527e-07, + "loss": 1.4149, + "step": 768 + }, + { + "epoch": 0.7568897637795275, + "grad_norm": 13.904799461364746, + "learning_rate": 7.614314115308152e-07, + "loss": 0.844, + "step": 769 + }, + { + "epoch": 0.7578740157480315, + "grad_norm": 14.56535816192627, + "learning_rate": 7.624254473161036e-07, + "loss": 0.9184, + "step": 770 + }, + { + "epoch": 0.7588582677165354, + "grad_norm": 12.913688659667969, + "learning_rate": 7.634194831013918e-07, + "loss": 0.9229, + "step": 771 + }, + { + "epoch": 0.7598425196850394, + "grad_norm": 19.055513381958008, + "learning_rate": 7.644135188866801e-07, + "loss": 1.3872, + "step": 772 + }, + { + "epoch": 0.7608267716535433, + "grad_norm": 14.286948204040527, + "learning_rate": 7.654075546719683e-07, + "loss": 0.7914, + "step": 773 + }, + { + "epoch": 0.7618110236220472, + "grad_norm": 13.263849258422852, + "learning_rate": 7.664015904572566e-07, + "loss": 0.8064, + "step": 774 + }, + { + "epoch": 0.7627952755905512, + "grad_norm": 15.496315002441406, + "learning_rate": 7.673956262425449e-07, + "loss": 1.0489, + "step": 775 + }, + { + "epoch": 0.7637795275590551, + "grad_norm": 12.534154891967773, + "learning_rate": 7.683896620278331e-07, + "loss": 1.0517, + "step": 776 + }, + { + "epoch": 0.764763779527559, + "grad_norm": 13.477228164672852, + "learning_rate": 7.693836978131215e-07, + "loss": 0.9025, + "step": 777 + }, + { + "epoch": 0.765748031496063, + "grad_norm": 13.898954391479492, + "learning_rate": 7.703777335984097e-07, + "loss": 0.7241, + "step": 778 + }, + { + "epoch": 0.7667322834645669, + "grad_norm": 14.8298978805542, + "learning_rate": 7.713717693836979e-07, + "loss": 1.0115, + "step": 779 + }, + { + "epoch": 0.7677165354330708, + "grad_norm": 15.819334030151367, + "learning_rate": 7.723658051689862e-07, + "loss": 1.1583, + "step": 780 + }, + { + "epoch": 0.7687007874015748, + "grad_norm": 14.541418075561523, + "learning_rate": 7.733598409542745e-07, + "loss": 1.0957, + "step": 781 + }, + { + "epoch": 0.7696850393700787, + "grad_norm": 14.33630084991455, + "learning_rate": 7.743538767395627e-07, + "loss": 0.8654, + "step": 782 + }, + { + "epoch": 0.7706692913385826, + "grad_norm": 17.25847816467285, + "learning_rate": 7.75347912524851e-07, + "loss": 1.1943, + "step": 783 + }, + { + "epoch": 0.7716535433070866, + "grad_norm": 15.107831954956055, + "learning_rate": 7.763419483101393e-07, + "loss": 0.9565, + "step": 784 + }, + { + "epoch": 0.7726377952755905, + "grad_norm": 17.720556259155273, + "learning_rate": 7.773359840954276e-07, + "loss": 1.0079, + "step": 785 + }, + { + "epoch": 0.7736220472440944, + "grad_norm": 20.499629974365234, + "learning_rate": 7.783300198807158e-07, + "loss": 1.3196, + "step": 786 + }, + { + "epoch": 0.7746062992125984, + "grad_norm": 15.560935974121094, + "learning_rate": 7.79324055666004e-07, + "loss": 0.8066, + "step": 787 + }, + { + "epoch": 0.7755905511811023, + "grad_norm": 17.144840240478516, + "learning_rate": 7.803180914512924e-07, + "loss": 1.1875, + "step": 788 + }, + { + "epoch": 0.7765748031496063, + "grad_norm": 13.65082836151123, + "learning_rate": 7.813121272365806e-07, + "loss": 0.9068, + "step": 789 + }, + { + "epoch": 0.7775590551181102, + "grad_norm": 13.954313278198242, + "learning_rate": 7.823061630218689e-07, + "loss": 0.9388, + "step": 790 + }, + { + "epoch": 0.7785433070866141, + "grad_norm": 17.345203399658203, + "learning_rate": 7.833001988071572e-07, + "loss": 1.5462, + "step": 791 + }, + { + "epoch": 0.7795275590551181, + "grad_norm": 14.942344665527344, + "learning_rate": 7.842942345924454e-07, + "loss": 0.9369, + "step": 792 + }, + { + "epoch": 0.780511811023622, + "grad_norm": 16.15380096435547, + "learning_rate": 7.852882703777337e-07, + "loss": 1.6793, + "step": 793 + }, + { + "epoch": 0.781496062992126, + "grad_norm": 16.013246536254883, + "learning_rate": 7.862823061630219e-07, + "loss": 1.0793, + "step": 794 + }, + { + "epoch": 0.78248031496063, + "grad_norm": 11.829020500183105, + "learning_rate": 7.872763419483102e-07, + "loss": 0.7758, + "step": 795 + }, + { + "epoch": 0.7834645669291339, + "grad_norm": 12.535819053649902, + "learning_rate": 7.882703777335986e-07, + "loss": 0.6, + "step": 796 + }, + { + "epoch": 0.7844488188976378, + "grad_norm": 12.625960350036621, + "learning_rate": 7.892644135188868e-07, + "loss": 0.7136, + "step": 797 + }, + { + "epoch": 0.7854330708661418, + "grad_norm": 12.130288124084473, + "learning_rate": 7.902584493041752e-07, + "loss": 0.813, + "step": 798 + }, + { + "epoch": 0.7864173228346457, + "grad_norm": 14.265738487243652, + "learning_rate": 7.912524850894634e-07, + "loss": 0.8777, + "step": 799 + }, + { + "epoch": 0.7874015748031497, + "grad_norm": 16.624202728271484, + "learning_rate": 7.922465208747516e-07, + "loss": 1.119, + "step": 800 + }, + { + "epoch": 0.7883858267716536, + "grad_norm": 12.201735496520996, + "learning_rate": 7.932405566600399e-07, + "loss": 0.5711, + "step": 801 + }, + { + "epoch": 0.7893700787401575, + "grad_norm": 15.760323524475098, + "learning_rate": 7.942345924453282e-07, + "loss": 0.6798, + "step": 802 + }, + { + "epoch": 0.7903543307086615, + "grad_norm": 15.578442573547363, + "learning_rate": 7.952286282306165e-07, + "loss": 0.8154, + "step": 803 + }, + { + "epoch": 0.7913385826771654, + "grad_norm": 9.82174301147461, + "learning_rate": 7.962226640159047e-07, + "loss": 0.3272, + "step": 804 + }, + { + "epoch": 0.7923228346456693, + "grad_norm": 18.133094787597656, + "learning_rate": 7.97216699801193e-07, + "loss": 0.9906, + "step": 805 + }, + { + "epoch": 0.7933070866141733, + "grad_norm": 18.373262405395508, + "learning_rate": 7.982107355864813e-07, + "loss": 1.0634, + "step": 806 + }, + { + "epoch": 0.7942913385826772, + "grad_norm": 17.529993057250977, + "learning_rate": 7.992047713717695e-07, + "loss": 0.9913, + "step": 807 + }, + { + "epoch": 0.7952755905511811, + "grad_norm": 16.67307472229004, + "learning_rate": 8.001988071570577e-07, + "loss": 1.0392, + "step": 808 + }, + { + "epoch": 0.7962598425196851, + "grad_norm": 18.125699996948242, + "learning_rate": 8.011928429423461e-07, + "loss": 0.7832, + "step": 809 + }, + { + "epoch": 0.797244094488189, + "grad_norm": 12.720748901367188, + "learning_rate": 8.021868787276343e-07, + "loss": 0.4475, + "step": 810 + }, + { + "epoch": 0.7982283464566929, + "grad_norm": 13.569591522216797, + "learning_rate": 8.031809145129226e-07, + "loss": 0.708, + "step": 811 + }, + { + "epoch": 0.7992125984251969, + "grad_norm": 15.576632499694824, + "learning_rate": 8.041749502982109e-07, + "loss": 0.8815, + "step": 812 + }, + { + "epoch": 0.8001968503937008, + "grad_norm": 20.23661994934082, + "learning_rate": 8.051689860834991e-07, + "loss": 1.3039, + "step": 813 + }, + { + "epoch": 0.8011811023622047, + "grad_norm": 17.46402359008789, + "learning_rate": 8.061630218687874e-07, + "loss": 1.3863, + "step": 814 + }, + { + "epoch": 0.8021653543307087, + "grad_norm": 14.91606330871582, + "learning_rate": 8.071570576540756e-07, + "loss": 1.0562, + "step": 815 + }, + { + "epoch": 0.8031496062992126, + "grad_norm": 16.14639663696289, + "learning_rate": 8.08151093439364e-07, + "loss": 0.7251, + "step": 816 + }, + { + "epoch": 0.8041338582677166, + "grad_norm": 14.937382698059082, + "learning_rate": 8.091451292246522e-07, + "loss": 0.6901, + "step": 817 + }, + { + "epoch": 0.8051181102362205, + "grad_norm": 13.609417915344238, + "learning_rate": 8.101391650099404e-07, + "loss": 0.7074, + "step": 818 + }, + { + "epoch": 0.8061023622047244, + "grad_norm": 10.547968864440918, + "learning_rate": 8.111332007952288e-07, + "loss": 0.5985, + "step": 819 + }, + { + "epoch": 0.8070866141732284, + "grad_norm": 14.27464771270752, + "learning_rate": 8.12127236580517e-07, + "loss": 0.674, + "step": 820 + }, + { + "epoch": 0.8080708661417323, + "grad_norm": 12.711366653442383, + "learning_rate": 8.131212723658052e-07, + "loss": 0.6977, + "step": 821 + }, + { + "epoch": 0.8090551181102362, + "grad_norm": 14.525792121887207, + "learning_rate": 8.141153081510935e-07, + "loss": 0.6939, + "step": 822 + }, + { + "epoch": 0.8100393700787402, + "grad_norm": 13.829584121704102, + "learning_rate": 8.151093439363818e-07, + "loss": 0.7825, + "step": 823 + }, + { + "epoch": 0.8110236220472441, + "grad_norm": 13.423285484313965, + "learning_rate": 8.161033797216701e-07, + "loss": 0.9403, + "step": 824 + }, + { + "epoch": 0.812007874015748, + "grad_norm": 11.258942604064941, + "learning_rate": 8.170974155069583e-07, + "loss": 0.5739, + "step": 825 + }, + { + "epoch": 0.812992125984252, + "grad_norm": 17.79983901977539, + "learning_rate": 8.180914512922467e-07, + "loss": 1.2775, + "step": 826 + }, + { + "epoch": 0.8139763779527559, + "grad_norm": 15.341512680053711, + "learning_rate": 8.19085487077535e-07, + "loss": 0.7558, + "step": 827 + }, + { + "epoch": 0.8149606299212598, + "grad_norm": 19.59752082824707, + "learning_rate": 8.200795228628232e-07, + "loss": 0.9289, + "step": 828 + }, + { + "epoch": 0.8159448818897638, + "grad_norm": 15.072026252746582, + "learning_rate": 8.210735586481115e-07, + "loss": 0.7306, + "step": 829 + }, + { + "epoch": 0.8169291338582677, + "grad_norm": 16.29781150817871, + "learning_rate": 8.220675944333998e-07, + "loss": 0.8876, + "step": 830 + }, + { + "epoch": 0.8179133858267716, + "grad_norm": 16.281766891479492, + "learning_rate": 8.23061630218688e-07, + "loss": 0.9344, + "step": 831 + }, + { + "epoch": 0.8188976377952756, + "grad_norm": 12.69700813293457, + "learning_rate": 8.240556660039763e-07, + "loss": 0.8379, + "step": 832 + }, + { + "epoch": 0.8198818897637795, + "grad_norm": 10.936923027038574, + "learning_rate": 8.250497017892646e-07, + "loss": 0.3775, + "step": 833 + }, + { + "epoch": 0.8208661417322834, + "grad_norm": 9.851780891418457, + "learning_rate": 8.260437375745529e-07, + "loss": 0.4071, + "step": 834 + }, + { + "epoch": 0.8218503937007874, + "grad_norm": 13.968542098999023, + "learning_rate": 8.270377733598411e-07, + "loss": 0.5419, + "step": 835 + }, + { + "epoch": 0.8228346456692913, + "grad_norm": 14.457850456237793, + "learning_rate": 8.280318091451293e-07, + "loss": 0.7428, + "step": 836 + }, + { + "epoch": 0.8238188976377953, + "grad_norm": 16.458881378173828, + "learning_rate": 8.290258449304177e-07, + "loss": 0.905, + "step": 837 + }, + { + "epoch": 0.8248031496062992, + "grad_norm": 15.484702110290527, + "learning_rate": 8.300198807157059e-07, + "loss": 0.605, + "step": 838 + }, + { + "epoch": 0.8257874015748031, + "grad_norm": 21.334392547607422, + "learning_rate": 8.310139165009941e-07, + "loss": 1.6087, + "step": 839 + }, + { + "epoch": 0.8267716535433071, + "grad_norm": 13.859783172607422, + "learning_rate": 8.320079522862825e-07, + "loss": 0.5758, + "step": 840 + }, + { + "epoch": 0.827755905511811, + "grad_norm": 16.148588180541992, + "learning_rate": 8.330019880715707e-07, + "loss": 0.9991, + "step": 841 + }, + { + "epoch": 0.8287401574803149, + "grad_norm": 21.421781539916992, + "learning_rate": 8.33996023856859e-07, + "loss": 1.3015, + "step": 842 + }, + { + "epoch": 0.8297244094488189, + "grad_norm": 16.3444766998291, + "learning_rate": 8.349900596421472e-07, + "loss": 0.8529, + "step": 843 + }, + { + "epoch": 0.8307086614173228, + "grad_norm": 19.439611434936523, + "learning_rate": 8.359840954274355e-07, + "loss": 0.8257, + "step": 844 + }, + { + "epoch": 0.8316929133858267, + "grad_norm": 16.935365676879883, + "learning_rate": 8.369781312127238e-07, + "loss": 0.8513, + "step": 845 + }, + { + "epoch": 0.8326771653543307, + "grad_norm": 13.878691673278809, + "learning_rate": 8.37972166998012e-07, + "loss": 0.9995, + "step": 846 + }, + { + "epoch": 0.8336614173228346, + "grad_norm": 14.36600112915039, + "learning_rate": 8.389662027833004e-07, + "loss": 1.0182, + "step": 847 + }, + { + "epoch": 0.8346456692913385, + "grad_norm": 13.805697441101074, + "learning_rate": 8.399602385685886e-07, + "loss": 0.6523, + "step": 848 + }, + { + "epoch": 0.8356299212598425, + "grad_norm": 13.495059967041016, + "learning_rate": 8.409542743538768e-07, + "loss": 0.8436, + "step": 849 + }, + { + "epoch": 0.8366141732283464, + "grad_norm": 17.51573944091797, + "learning_rate": 8.419483101391651e-07, + "loss": 1.4555, + "step": 850 + }, + { + "epoch": 0.8375984251968503, + "grad_norm": 11.79956340789795, + "learning_rate": 8.429423459244534e-07, + "loss": 0.6176, + "step": 851 + }, + { + "epoch": 0.8385826771653543, + "grad_norm": 15.17784595489502, + "learning_rate": 8.439363817097416e-07, + "loss": 1.1224, + "step": 852 + }, + { + "epoch": 0.8395669291338582, + "grad_norm": 11.504303932189941, + "learning_rate": 8.449304174950299e-07, + "loss": 0.5743, + "step": 853 + }, + { + "epoch": 0.8405511811023622, + "grad_norm": 12.635910034179688, + "learning_rate": 8.459244532803182e-07, + "loss": 0.6488, + "step": 854 + }, + { + "epoch": 0.8415354330708661, + "grad_norm": 11.405928611755371, + "learning_rate": 8.469184890656066e-07, + "loss": 0.6553, + "step": 855 + }, + { + "epoch": 0.84251968503937, + "grad_norm": 15.8025541305542, + "learning_rate": 8.479125248508948e-07, + "loss": 1.0901, + "step": 856 + }, + { + "epoch": 0.843503937007874, + "grad_norm": 20.577259063720703, + "learning_rate": 8.48906560636183e-07, + "loss": 1.2568, + "step": 857 + }, + { + "epoch": 0.844488188976378, + "grad_norm": 15.693262100219727, + "learning_rate": 8.499005964214714e-07, + "loss": 0.7643, + "step": 858 + }, + { + "epoch": 0.8454724409448819, + "grad_norm": 12.100361824035645, + "learning_rate": 8.508946322067596e-07, + "loss": 0.3966, + "step": 859 + }, + { + "epoch": 0.8464566929133859, + "grad_norm": 15.864554405212402, + "learning_rate": 8.518886679920479e-07, + "loss": 0.6586, + "step": 860 + }, + { + "epoch": 0.8474409448818898, + "grad_norm": 14.532626152038574, + "learning_rate": 8.528827037773362e-07, + "loss": 0.8597, + "step": 861 + }, + { + "epoch": 0.8484251968503937, + "grad_norm": 18.547473907470703, + "learning_rate": 8.538767395626244e-07, + "loss": 1.237, + "step": 862 + }, + { + "epoch": 0.8494094488188977, + "grad_norm": 20.073894500732422, + "learning_rate": 8.548707753479127e-07, + "loss": 0.9306, + "step": 863 + }, + { + "epoch": 0.8503937007874016, + "grad_norm": 16.489517211914062, + "learning_rate": 8.558648111332009e-07, + "loss": 0.7643, + "step": 864 + }, + { + "epoch": 0.8513779527559056, + "grad_norm": 14.844143867492676, + "learning_rate": 8.568588469184892e-07, + "loss": 0.7402, + "step": 865 + }, + { + "epoch": 0.8523622047244095, + "grad_norm": 19.942005157470703, + "learning_rate": 8.578528827037775e-07, + "loss": 0.9191, + "step": 866 + }, + { + "epoch": 0.8533464566929134, + "grad_norm": 15.869391441345215, + "learning_rate": 8.588469184890657e-07, + "loss": 0.9644, + "step": 867 + }, + { + "epoch": 0.8543307086614174, + "grad_norm": 14.878215789794922, + "learning_rate": 8.598409542743541e-07, + "loss": 0.7933, + "step": 868 + }, + { + "epoch": 0.8553149606299213, + "grad_norm": 20.548446655273438, + "learning_rate": 8.608349900596423e-07, + "loss": 1.5964, + "step": 869 + }, + { + "epoch": 0.8562992125984252, + "grad_norm": 17.03843116760254, + "learning_rate": 8.618290258449305e-07, + "loss": 0.8953, + "step": 870 + }, + { + "epoch": 0.8572834645669292, + "grad_norm": 16.03615951538086, + "learning_rate": 8.628230616302189e-07, + "loss": 1.0073, + "step": 871 + }, + { + "epoch": 0.8582677165354331, + "grad_norm": 12.262407302856445, + "learning_rate": 8.638170974155071e-07, + "loss": 0.517, + "step": 872 + }, + { + "epoch": 0.859251968503937, + "grad_norm": 17.016586303710938, + "learning_rate": 8.648111332007954e-07, + "loss": 0.8879, + "step": 873 + }, + { + "epoch": 0.860236220472441, + "grad_norm": 16.42693328857422, + "learning_rate": 8.658051689860836e-07, + "loss": 1.5371, + "step": 874 + }, + { + "epoch": 0.8612204724409449, + "grad_norm": 16.2298526763916, + "learning_rate": 8.667992047713719e-07, + "loss": 0.9743, + "step": 875 + }, + { + "epoch": 0.8622047244094488, + "grad_norm": 17.53359031677246, + "learning_rate": 8.677932405566602e-07, + "loss": 1.0717, + "step": 876 + }, + { + "epoch": 0.8631889763779528, + "grad_norm": 13.198174476623535, + "learning_rate": 8.687872763419484e-07, + "loss": 0.6625, + "step": 877 + }, + { + "epoch": 0.8641732283464567, + "grad_norm": 14.632850646972656, + "learning_rate": 8.697813121272366e-07, + "loss": 0.8521, + "step": 878 + }, + { + "epoch": 0.8651574803149606, + "grad_norm": 15.524723052978516, + "learning_rate": 8.70775347912525e-07, + "loss": 0.7955, + "step": 879 + }, + { + "epoch": 0.8661417322834646, + "grad_norm": 14.512434005737305, + "learning_rate": 8.717693836978132e-07, + "loss": 0.9416, + "step": 880 + }, + { + "epoch": 0.8671259842519685, + "grad_norm": 14.040345191955566, + "learning_rate": 8.727634194831015e-07, + "loss": 0.8257, + "step": 881 + }, + { + "epoch": 0.8681102362204725, + "grad_norm": 16.287595748901367, + "learning_rate": 8.737574552683898e-07, + "loss": 1.3879, + "step": 882 + }, + { + "epoch": 0.8690944881889764, + "grad_norm": 12.638394355773926, + "learning_rate": 8.74751491053678e-07, + "loss": 0.9457, + "step": 883 + }, + { + "epoch": 0.8700787401574803, + "grad_norm": 14.083216667175293, + "learning_rate": 8.757455268389663e-07, + "loss": 0.891, + "step": 884 + }, + { + "epoch": 0.8710629921259843, + "grad_norm": 16.631366729736328, + "learning_rate": 8.767395626242545e-07, + "loss": 0.9427, + "step": 885 + }, + { + "epoch": 0.8720472440944882, + "grad_norm": 13.598855018615723, + "learning_rate": 8.77733598409543e-07, + "loss": 0.8526, + "step": 886 + }, + { + "epoch": 0.8730314960629921, + "grad_norm": 17.016712188720703, + "learning_rate": 8.787276341948312e-07, + "loss": 1.2298, + "step": 887 + }, + { + "epoch": 0.8740157480314961, + "grad_norm": 11.86217975616455, + "learning_rate": 8.797216699801194e-07, + "loss": 0.6241, + "step": 888 + }, + { + "epoch": 0.875, + "grad_norm": 15.895013809204102, + "learning_rate": 8.807157057654078e-07, + "loss": 0.7055, + "step": 889 + }, + { + "epoch": 0.8759842519685039, + "grad_norm": 16.778099060058594, + "learning_rate": 8.81709741550696e-07, + "loss": 0.9713, + "step": 890 + }, + { + "epoch": 0.8769685039370079, + "grad_norm": 17.01609230041504, + "learning_rate": 8.827037773359842e-07, + "loss": 1.0591, + "step": 891 + }, + { + "epoch": 0.8779527559055118, + "grad_norm": 15.640320777893066, + "learning_rate": 8.836978131212726e-07, + "loss": 1.0597, + "step": 892 + }, + { + "epoch": 0.8789370078740157, + "grad_norm": 18.290407180786133, + "learning_rate": 8.846918489065608e-07, + "loss": 1.1631, + "step": 893 + }, + { + "epoch": 0.8799212598425197, + "grad_norm": 12.475908279418945, + "learning_rate": 8.856858846918491e-07, + "loss": 0.6098, + "step": 894 + }, + { + "epoch": 0.8809055118110236, + "grad_norm": 15.96957778930664, + "learning_rate": 8.866799204771373e-07, + "loss": 1.1498, + "step": 895 + }, + { + "epoch": 0.8818897637795275, + "grad_norm": 11.635284423828125, + "learning_rate": 8.876739562624256e-07, + "loss": 0.5379, + "step": 896 + }, + { + "epoch": 0.8828740157480315, + "grad_norm": 13.4318265914917, + "learning_rate": 8.886679920477139e-07, + "loss": 0.7921, + "step": 897 + }, + { + "epoch": 0.8838582677165354, + "grad_norm": 13.293816566467285, + "learning_rate": 8.896620278330021e-07, + "loss": 0.9092, + "step": 898 + }, + { + "epoch": 0.8848425196850394, + "grad_norm": 16.943866729736328, + "learning_rate": 8.906560636182905e-07, + "loss": 1.0348, + "step": 899 + }, + { + "epoch": 0.8858267716535433, + "grad_norm": 13.879645347595215, + "learning_rate": 8.916500994035787e-07, + "loss": 0.9087, + "step": 900 + }, + { + "epoch": 0.8868110236220472, + "grad_norm": 18.99857521057129, + "learning_rate": 8.926441351888669e-07, + "loss": 1.5328, + "step": 901 + }, + { + "epoch": 0.8877952755905512, + "grad_norm": 16.0674991607666, + "learning_rate": 8.936381709741552e-07, + "loss": 0.8664, + "step": 902 + }, + { + "epoch": 0.8887795275590551, + "grad_norm": 13.391024589538574, + "learning_rate": 8.946322067594435e-07, + "loss": 0.6873, + "step": 903 + }, + { + "epoch": 0.889763779527559, + "grad_norm": 17.474857330322266, + "learning_rate": 8.956262425447318e-07, + "loss": 1.1763, + "step": 904 + }, + { + "epoch": 0.890748031496063, + "grad_norm": 17.762374877929688, + "learning_rate": 8.9662027833002e-07, + "loss": 1.2853, + "step": 905 + }, + { + "epoch": 0.8917322834645669, + "grad_norm": 15.883246421813965, + "learning_rate": 8.976143141153082e-07, + "loss": 0.8163, + "step": 906 + }, + { + "epoch": 0.8927165354330708, + "grad_norm": 11.354119300842285, + "learning_rate": 8.986083499005966e-07, + "loss": 0.7383, + "step": 907 + }, + { + "epoch": 0.8937007874015748, + "grad_norm": 11.073287963867188, + "learning_rate": 8.996023856858848e-07, + "loss": 0.7833, + "step": 908 + }, + { + "epoch": 0.8946850393700787, + "grad_norm": 14.48930549621582, + "learning_rate": 9.00596421471173e-07, + "loss": 1.078, + "step": 909 + }, + { + "epoch": 0.8956692913385826, + "grad_norm": 12.383027076721191, + "learning_rate": 9.015904572564614e-07, + "loss": 0.6647, + "step": 910 + }, + { + "epoch": 0.8966535433070866, + "grad_norm": 15.81402587890625, + "learning_rate": 9.025844930417496e-07, + "loss": 1.0016, + "step": 911 + }, + { + "epoch": 0.8976377952755905, + "grad_norm": 13.038455963134766, + "learning_rate": 9.035785288270379e-07, + "loss": 0.8432, + "step": 912 + }, + { + "epoch": 0.8986220472440944, + "grad_norm": 13.19998550415039, + "learning_rate": 9.045725646123262e-07, + "loss": 0.9927, + "step": 913 + }, + { + "epoch": 0.8996062992125984, + "grad_norm": 10.537164688110352, + "learning_rate": 9.055666003976144e-07, + "loss": 0.4985, + "step": 914 + }, + { + "epoch": 0.9005905511811023, + "grad_norm": 6.800235748291016, + "learning_rate": 9.065606361829028e-07, + "loss": 0.1726, + "step": 915 + }, + { + "epoch": 0.9015748031496063, + "grad_norm": 17.059080123901367, + "learning_rate": 9.07554671968191e-07, + "loss": 0.9437, + "step": 916 + }, + { + "epoch": 0.9025590551181102, + "grad_norm": 9.479357719421387, + "learning_rate": 9.085487077534794e-07, + "loss": 0.2565, + "step": 917 + }, + { + "epoch": 0.9035433070866141, + "grad_norm": 12.993677139282227, + "learning_rate": 9.095427435387676e-07, + "loss": 0.6238, + "step": 918 + }, + { + "epoch": 0.9035433070866141, + "eval_Qnli-dev_cosine_accuracy": 0.685546875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7730953097343445, + "eval_Qnli-dev_cosine_ap": 0.6998329653700477, + "eval_Qnli-dev_cosine_f1": 0.6864564007421151, + "eval_Qnli-dev_cosine_f1_threshold": 0.7575435638427734, + "eval_Qnli-dev_cosine_precision": 0.6105610561056105, + "eval_Qnli-dev_cosine_recall": 0.7838983050847458, + "eval_Qnli-dev_dot_accuracy": 0.685546875, + "eval_Qnli-dev_dot_accuracy_threshold": 594.6486206054688, + "eval_Qnli-dev_dot_ap": 0.6993568496858908, + "eval_Qnli-dev_dot_f1": 0.6864564007421151, + "eval_Qnli-dev_dot_f1_threshold": 582.533203125, + "eval_Qnli-dev_dot_precision": 0.6105610561056105, + "eval_Qnli-dev_dot_recall": 0.7838983050847458, + "eval_Qnli-dev_euclidean_accuracy": 0.685546875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 18.682254791259766, + "eval_Qnli-dev_euclidean_ap": 0.6998558914512977, + "eval_Qnli-dev_euclidean_f1": 0.6864564007421151, + "eval_Qnli-dev_euclidean_f1_threshold": 19.310272216796875, + "eval_Qnli-dev_euclidean_precision": 0.6105610561056105, + "eval_Qnli-dev_euclidean_recall": 0.7838983050847458, + "eval_Qnli-dev_manhattan_accuracy": 0.677734375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 406.2527770996094, + "eval_Qnli-dev_manhattan_ap": 0.6984213903603985, + "eval_Qnli-dev_manhattan_f1": 0.6878504672897197, + "eval_Qnli-dev_manhattan_f1_threshold": 425.31719970703125, + "eval_Qnli-dev_manhattan_precision": 0.6153846153846154, + "eval_Qnli-dev_manhattan_recall": 0.7796610169491526, + "eval_Qnli-dev_max_accuracy": 0.685546875, + "eval_Qnli-dev_max_accuracy_threshold": 594.6486206054688, + "eval_Qnli-dev_max_ap": 0.6998558914512977, + "eval_Qnli-dev_max_f1": 0.6878504672897197, + "eval_Qnli-dev_max_f1_threshold": 582.533203125, + "eval_Qnli-dev_max_precision": 0.6153846153846154, + "eval_Qnli-dev_max_recall": 0.7838983050847458, + "eval_allNLI-dev_cosine_accuracy": 0.708984375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.8877366781234741, + "eval_allNLI-dev_cosine_ap": 0.546000327650908, + "eval_allNLI-dev_cosine_f1": 0.5522041763341067, + "eval_allNLI-dev_cosine_f1_threshold": 0.8286556005477905, + "eval_allNLI-dev_cosine_precision": 0.46124031007751937, + "eval_allNLI-dev_cosine_recall": 0.6878612716763006, + "eval_allNLI-dev_dot_accuracy": 0.7109375, + "eval_allNLI-dev_dot_accuracy_threshold": 680.6962890625, + "eval_allNLI-dev_dot_ap": 0.5459476454670233, + "eval_allNLI-dev_dot_f1": 0.5522041763341067, + "eval_allNLI-dev_dot_f1_threshold": 637.2539672851562, + "eval_allNLI-dev_dot_precision": 0.46124031007751937, + "eval_allNLI-dev_dot_recall": 0.6878612716763006, + "eval_allNLI-dev_euclidean_accuracy": 0.708984375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 13.14069938659668, + "eval_allNLI-dev_euclidean_ap": 0.5459041451400851, + "eval_allNLI-dev_euclidean_f1": 0.5522041763341067, + "eval_allNLI-dev_euclidean_f1_threshold": 16.234378814697266, + "eval_allNLI-dev_euclidean_precision": 0.46124031007751937, + "eval_allNLI-dev_euclidean_recall": 0.6878612716763006, + "eval_allNLI-dev_manhattan_accuracy": 0.712890625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 288.51678466796875, + "eval_allNLI-dev_manhattan_ap": 0.5481183507291686, + "eval_allNLI-dev_manhattan_f1": 0.5509259259259259, + "eval_allNLI-dev_manhattan_f1_threshold": 360.83740234375, + "eval_allNLI-dev_manhattan_precision": 0.4594594594594595, + "eval_allNLI-dev_manhattan_recall": 0.6878612716763006, + "eval_allNLI-dev_max_accuracy": 0.712890625, + "eval_allNLI-dev_max_accuracy_threshold": 680.6962890625, + "eval_allNLI-dev_max_ap": 0.5481183507291686, + "eval_allNLI-dev_max_f1": 0.5522041763341067, + "eval_allNLI-dev_max_f1_threshold": 637.2539672851562, + "eval_allNLI-dev_max_precision": 0.46124031007751937, + "eval_allNLI-dev_max_recall": 0.6878612716763006, + "eval_loss": 1.1640249490737915, + "eval_runtime": 50.4592, + "eval_samples_per_second": 32.977, + "eval_sequential_score": 0.6998558914512977, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.5946051127937603, + "eval_sts-test_pearson_dot": 0.5946033281863461, + "eval_sts-test_pearson_euclidean": 0.6072687591964337, + "eval_sts-test_pearson_manhattan": 0.6067220567514913, + "eval_sts-test_pearson_max": 0.6072687591964337, + "eval_sts-test_spearman_cosine": 0.6063628101304416, + "eval_sts-test_spearman_dot": 0.6064043393010128, + "eval_sts-test_spearman_euclidean": 0.606384917253569, + "eval_sts-test_spearman_manhattan": 0.6064764783343758, + "eval_sts-test_spearman_max": 0.6064764783343758, + "step": 918 + }, + { + "epoch": 0.9045275590551181, + "grad_norm": 20.241952896118164, + "learning_rate": 9.105367793240558e-07, + "loss": 0.9311, + "step": 919 + }, + { + "epoch": 0.905511811023622, + "grad_norm": 27.6416015625, + "learning_rate": 9.115308151093442e-07, + "loss": 1.6868, + "step": 920 + }, + { + "epoch": 0.906496062992126, + "grad_norm": 16.65912628173828, + "learning_rate": 9.125248508946324e-07, + "loss": 0.8606, + "step": 921 + }, + { + "epoch": 0.90748031496063, + "grad_norm": 10.079182624816895, + "learning_rate": 9.135188866799206e-07, + "loss": 0.4508, + "step": 922 + }, + { + "epoch": 0.9084645669291339, + "grad_norm": 12.931349754333496, + "learning_rate": 9.145129224652089e-07, + "loss": 0.6556, + "step": 923 + }, + { + "epoch": 0.9094488188976378, + "grad_norm": 12.358150482177734, + "learning_rate": 9.155069582504972e-07, + "loss": 0.5244, + "step": 924 + }, + { + "epoch": 0.9104330708661418, + "grad_norm": 16.49843406677246, + "learning_rate": 9.165009940357855e-07, + "loss": 0.6512, + "step": 925 + }, + { + "epoch": 0.9114173228346457, + "grad_norm": 16.631345748901367, + "learning_rate": 9.174950298210737e-07, + "loss": 0.6594, + "step": 926 + }, + { + "epoch": 0.9124015748031497, + "grad_norm": 16.298473358154297, + "learning_rate": 9.18489065606362e-07, + "loss": 1.0091, + "step": 927 + }, + { + "epoch": 0.9133858267716536, + "grad_norm": 15.244587898254395, + "learning_rate": 9.194831013916503e-07, + "loss": 0.6465, + "step": 928 + }, + { + "epoch": 0.9143700787401575, + "grad_norm": 15.43215274810791, + "learning_rate": 9.204771371769385e-07, + "loss": 0.798, + "step": 929 + }, + { + "epoch": 0.9153543307086615, + "grad_norm": 17.460987091064453, + "learning_rate": 9.214711729622268e-07, + "loss": 1.0069, + "step": 930 + }, + { + "epoch": 0.9163385826771654, + "grad_norm": 13.587342262268066, + "learning_rate": 9.224652087475151e-07, + "loss": 0.7407, + "step": 931 + }, + { + "epoch": 0.9173228346456693, + "grad_norm": 14.738330841064453, + "learning_rate": 9.234592445328033e-07, + "loss": 0.7979, + "step": 932 + }, + { + "epoch": 0.9183070866141733, + "grad_norm": 18.48650360107422, + "learning_rate": 9.244532803180916e-07, + "loss": 1.688, + "step": 933 + }, + { + "epoch": 0.9192913385826772, + "grad_norm": 15.389516830444336, + "learning_rate": 9.254473161033799e-07, + "loss": 0.8505, + "step": 934 + }, + { + "epoch": 0.9202755905511811, + "grad_norm": 16.759868621826172, + "learning_rate": 9.264413518886681e-07, + "loss": 0.8101, + "step": 935 + }, + { + "epoch": 0.9212598425196851, + "grad_norm": 13.334268569946289, + "learning_rate": 9.274353876739564e-07, + "loss": 0.8542, + "step": 936 + }, + { + "epoch": 0.922244094488189, + "grad_norm": 14.320956230163574, + "learning_rate": 9.284294234592446e-07, + "loss": 0.956, + "step": 937 + }, + { + "epoch": 0.9232283464566929, + "grad_norm": 13.361112594604492, + "learning_rate": 9.29423459244533e-07, + "loss": 0.7072, + "step": 938 + }, + { + "epoch": 0.9242125984251969, + "grad_norm": 11.964262962341309, + "learning_rate": 9.304174950298212e-07, + "loss": 0.6316, + "step": 939 + }, + { + "epoch": 0.9251968503937008, + "grad_norm": 17.206703186035156, + "learning_rate": 9.314115308151094e-07, + "loss": 1.3868, + "step": 940 + }, + { + "epoch": 0.9261811023622047, + "grad_norm": 8.030574798583984, + "learning_rate": 9.324055666003978e-07, + "loss": 0.3854, + "step": 941 + }, + { + "epoch": 0.9271653543307087, + "grad_norm": 18.32826805114746, + "learning_rate": 9.33399602385686e-07, + "loss": 1.3662, + "step": 942 + }, + { + "epoch": 0.9281496062992126, + "grad_norm": 15.768746376037598, + "learning_rate": 9.343936381709743e-07, + "loss": 1.0133, + "step": 943 + }, + { + "epoch": 0.9291338582677166, + "grad_norm": 15.841397285461426, + "learning_rate": 9.353876739562625e-07, + "loss": 0.8543, + "step": 944 + }, + { + "epoch": 0.9301181102362205, + "grad_norm": 16.678314208984375, + "learning_rate": 9.363817097415509e-07, + "loss": 0.9697, + "step": 945 + }, + { + "epoch": 0.9311023622047244, + "grad_norm": 11.96019458770752, + "learning_rate": 9.373757455268392e-07, + "loss": 0.7126, + "step": 946 + }, + { + "epoch": 0.9320866141732284, + "grad_norm": 16.553489685058594, + "learning_rate": 9.383697813121274e-07, + "loss": 1.0513, + "step": 947 + }, + { + "epoch": 0.9330708661417323, + "grad_norm": 12.496065139770508, + "learning_rate": 9.393638170974157e-07, + "loss": 0.6503, + "step": 948 + }, + { + "epoch": 0.9340551181102362, + "grad_norm": 14.360799789428711, + "learning_rate": 9.40357852882704e-07, + "loss": 0.894, + "step": 949 + }, + { + "epoch": 0.9350393700787402, + "grad_norm": 12.230989456176758, + "learning_rate": 9.413518886679922e-07, + "loss": 0.4608, + "step": 950 + }, + { + "epoch": 0.9360236220472441, + "grad_norm": 14.994596481323242, + "learning_rate": 9.423459244532805e-07, + "loss": 0.8966, + "step": 951 + }, + { + "epoch": 0.937007874015748, + "grad_norm": 13.53376293182373, + "learning_rate": 9.433399602385688e-07, + "loss": 0.7412, + "step": 952 + }, + { + "epoch": 0.937992125984252, + "grad_norm": 14.893637657165527, + "learning_rate": 9.44333996023857e-07, + "loss": 0.9692, + "step": 953 + }, + { + "epoch": 0.9389763779527559, + "grad_norm": 12.472609519958496, + "learning_rate": 9.453280318091453e-07, + "loss": 0.5728, + "step": 954 + }, + { + "epoch": 0.9399606299212598, + "grad_norm": 12.355321884155273, + "learning_rate": 9.463220675944336e-07, + "loss": 0.6812, + "step": 955 + }, + { + "epoch": 0.9409448818897638, + "grad_norm": 11.577045440673828, + "learning_rate": 9.473161033797219e-07, + "loss": 0.5401, + "step": 956 + }, + { + "epoch": 0.9419291338582677, + "grad_norm": 15.942525863647461, + "learning_rate": 9.483101391650101e-07, + "loss": 0.9315, + "step": 957 + }, + { + "epoch": 0.9429133858267716, + "grad_norm": 15.413877487182617, + "learning_rate": 9.493041749502983e-07, + "loss": 0.6438, + "step": 958 + }, + { + "epoch": 0.9438976377952756, + "grad_norm": 15.510786056518555, + "learning_rate": 9.502982107355867e-07, + "loss": 1.0856, + "step": 959 + }, + { + "epoch": 0.9448818897637795, + "grad_norm": 13.81240177154541, + "learning_rate": 9.512922465208749e-07, + "loss": 0.7523, + "step": 960 + }, + { + "epoch": 0.9458661417322834, + "grad_norm": 13.06954574584961, + "learning_rate": 9.522862823061631e-07, + "loss": 0.5826, + "step": 961 + }, + { + "epoch": 0.9468503937007874, + "grad_norm": 21.632646560668945, + "learning_rate": 9.532803180914515e-07, + "loss": 0.9469, + "step": 962 + }, + { + "epoch": 0.9478346456692913, + "grad_norm": 14.201640129089355, + "learning_rate": 9.542743538767396e-07, + "loss": 0.647, + "step": 963 + }, + { + "epoch": 0.9488188976377953, + "grad_norm": 22.599220275878906, + "learning_rate": 9.55268389662028e-07, + "loss": 1.8012, + "step": 964 + }, + { + "epoch": 0.9498031496062992, + "grad_norm": 13.054848670959473, + "learning_rate": 9.562624254473162e-07, + "loss": 0.6264, + "step": 965 + }, + { + "epoch": 0.9507874015748031, + "grad_norm": 22.16948890686035, + "learning_rate": 9.572564612326044e-07, + "loss": 1.2779, + "step": 966 + }, + { + "epoch": 0.9517716535433071, + "grad_norm": 13.487398147583008, + "learning_rate": 9.582504970178928e-07, + "loss": 0.8312, + "step": 967 + }, + { + "epoch": 0.952755905511811, + "grad_norm": 16.275707244873047, + "learning_rate": 9.59244532803181e-07, + "loss": 0.6442, + "step": 968 + }, + { + "epoch": 0.9537401574803149, + "grad_norm": 12.01401424407959, + "learning_rate": 9.602385685884694e-07, + "loss": 0.5953, + "step": 969 + }, + { + "epoch": 0.9547244094488189, + "grad_norm": 9.097768783569336, + "learning_rate": 9.612326043737576e-07, + "loss": 0.4099, + "step": 970 + }, + { + "epoch": 0.9557086614173228, + "grad_norm": 14.573530197143555, + "learning_rate": 9.622266401590458e-07, + "loss": 0.9008, + "step": 971 + }, + { + "epoch": 0.9566929133858267, + "grad_norm": 15.452958106994629, + "learning_rate": 9.632206759443342e-07, + "loss": 1.4286, + "step": 972 + }, + { + "epoch": 0.9576771653543307, + "grad_norm": 13.133222579956055, + "learning_rate": 9.642147117296224e-07, + "loss": 0.9222, + "step": 973 + }, + { + "epoch": 0.9586614173228346, + "grad_norm": 12.39025592803955, + "learning_rate": 9.652087475149109e-07, + "loss": 0.8414, + "step": 974 + }, + { + "epoch": 0.9596456692913385, + "grad_norm": 14.773319244384766, + "learning_rate": 9.66202783300199e-07, + "loss": 0.7063, + "step": 975 + }, + { + "epoch": 0.9606299212598425, + "grad_norm": 13.327868461608887, + "learning_rate": 9.671968190854873e-07, + "loss": 0.6207, + "step": 976 + }, + { + "epoch": 0.9616141732283464, + "grad_norm": 17.18140411376953, + "learning_rate": 9.681908548707755e-07, + "loss": 0.8273, + "step": 977 + }, + { + "epoch": 0.9625984251968503, + "grad_norm": 15.404573440551758, + "learning_rate": 9.691848906560639e-07, + "loss": 0.8464, + "step": 978 + }, + { + "epoch": 0.9635826771653543, + "grad_norm": 13.830384254455566, + "learning_rate": 9.70178926441352e-07, + "loss": 0.8247, + "step": 979 + }, + { + "epoch": 0.9645669291338582, + "grad_norm": 14.349933624267578, + "learning_rate": 9.711729622266403e-07, + "loss": 0.7133, + "step": 980 + }, + { + "epoch": 0.9655511811023622, + "grad_norm": 17.402868270874023, + "learning_rate": 9.721669980119285e-07, + "loss": 0.7903, + "step": 981 + }, + { + "epoch": 0.9665354330708661, + "grad_norm": 19.871767044067383, + "learning_rate": 9.731610337972169e-07, + "loss": 1.0951, + "step": 982 + }, + { + "epoch": 0.96751968503937, + "grad_norm": 22.14597511291504, + "learning_rate": 9.74155069582505e-07, + "loss": 1.099, + "step": 983 + }, + { + "epoch": 0.968503937007874, + "grad_norm": 16.688730239868164, + "learning_rate": 9.751491053677933e-07, + "loss": 0.5865, + "step": 984 + }, + { + "epoch": 0.969488188976378, + "grad_norm": 18.638708114624023, + "learning_rate": 9.761431411530817e-07, + "loss": 0.8717, + "step": 985 + }, + { + "epoch": 0.9704724409448819, + "grad_norm": 19.519134521484375, + "learning_rate": 9.771371769383699e-07, + "loss": 0.9441, + "step": 986 + }, + { + "epoch": 0.9714566929133859, + "grad_norm": 18.85181999206543, + "learning_rate": 9.781312127236583e-07, + "loss": 0.9627, + "step": 987 + }, + { + "epoch": 0.9724409448818898, + "grad_norm": 14.521993637084961, + "learning_rate": 9.791252485089465e-07, + "loss": 1.039, + "step": 988 + }, + { + "epoch": 0.9734251968503937, + "grad_norm": 19.123416900634766, + "learning_rate": 9.801192842942347e-07, + "loss": 1.2139, + "step": 989 + }, + { + "epoch": 0.9744094488188977, + "grad_norm": 13.49136734008789, + "learning_rate": 9.811133200795231e-07, + "loss": 0.6284, + "step": 990 + }, + { + "epoch": 0.9753937007874016, + "grad_norm": 13.103900909423828, + "learning_rate": 9.821073558648113e-07, + "loss": 0.914, + "step": 991 + }, + { + "epoch": 0.9763779527559056, + "grad_norm": 9.744898796081543, + "learning_rate": 9.831013916500995e-07, + "loss": 0.4021, + "step": 992 + }, + { + "epoch": 0.9773622047244095, + "grad_norm": 9.621757507324219, + "learning_rate": 9.84095427435388e-07, + "loss": 0.5822, + "step": 993 + }, + { + "epoch": 0.9783464566929134, + "grad_norm": 13.71444034576416, + "learning_rate": 9.850894632206761e-07, + "loss": 0.5666, + "step": 994 + }, + { + "epoch": 0.9793307086614174, + "grad_norm": 15.207693099975586, + "learning_rate": 9.860834990059645e-07, + "loss": 1.3606, + "step": 995 + }, + { + "epoch": 0.9803149606299213, + "grad_norm": 13.993303298950195, + "learning_rate": 9.870775347912527e-07, + "loss": 0.7811, + "step": 996 + }, + { + "epoch": 0.9812992125984252, + "grad_norm": 11.10175609588623, + "learning_rate": 9.88071570576541e-07, + "loss": 0.3886, + "step": 997 + }, + { + "epoch": 0.9822834645669292, + "grad_norm": 21.290891647338867, + "learning_rate": 9.890656063618291e-07, + "loss": 1.5192, + "step": 998 + }, + { + "epoch": 0.9832677165354331, + "grad_norm": 9.359945297241211, + "learning_rate": 9.900596421471175e-07, + "loss": 0.3559, + "step": 999 + }, + { + "epoch": 0.984251968503937, + "grad_norm": 18.680240631103516, + "learning_rate": 9.910536779324057e-07, + "loss": 1.1638, + "step": 1000 + }, + { + "epoch": 0.985236220472441, + "grad_norm": 11.718449592590332, + "learning_rate": 9.92047713717694e-07, + "loss": 0.5369, + "step": 1001 + }, + { + "epoch": 0.9862204724409449, + "grad_norm": 14.164551734924316, + "learning_rate": 9.930417495029821e-07, + "loss": 0.6804, + "step": 1002 + }, + { + "epoch": 0.9872047244094488, + "grad_norm": 13.837848663330078, + "learning_rate": 9.940357852882705e-07, + "loss": 0.5193, + "step": 1003 + }, + { + "epoch": 0.9881889763779528, + "grad_norm": 20.806840896606445, + "learning_rate": 9.950298210735587e-07, + "loss": 1.1559, + "step": 1004 + }, + { + "epoch": 0.9891732283464567, + "grad_norm": 13.79966926574707, + "learning_rate": 9.96023856858847e-07, + "loss": 0.7566, + "step": 1005 + }, + { + "epoch": 0.9901574803149606, + "grad_norm": 14.3021821975708, + "learning_rate": 9.970178926441354e-07, + "loss": 0.6815, + "step": 1006 + }, + { + "epoch": 0.9911417322834646, + "grad_norm": 17.469287872314453, + "learning_rate": 9.980119284294236e-07, + "loss": 1.0367, + "step": 1007 + }, + { + "epoch": 0.9921259842519685, + "grad_norm": 12.036974906921387, + "learning_rate": 9.99005964214712e-07, + "loss": 0.5514, + "step": 1008 + }, + { + "epoch": 0.9931102362204725, + "grad_norm": 16.500354766845703, + "learning_rate": 1.0000000000000002e-06, + "loss": 0.7299, + "step": 1009 + }, + { + "epoch": 0.9940944881889764, + "grad_norm": 18.891454696655273, + "learning_rate": 9.999994102361802e-07, + "loss": 1.1474, + "step": 1010 + }, + { + "epoch": 0.9950787401574803, + "grad_norm": 13.927214622497559, + "learning_rate": 9.999976409461164e-07, + "loss": 0.5969, + "step": 1011 + }, + { + "epoch": 0.9960629921259843, + "grad_norm": 11.425060272216797, + "learning_rate": 9.999946921339963e-07, + "loss": 0.575, + "step": 1012 + }, + { + "epoch": 0.9970472440944882, + "grad_norm": 10.967757225036621, + "learning_rate": 9.999905638068e-07, + "loss": 0.6357, + "step": 1013 + }, + { + "epoch": 0.9980314960629921, + "grad_norm": 12.922017097473145, + "learning_rate": 9.999852559742985e-07, + "loss": 0.8671, + "step": 1014 + }, + { + "epoch": 0.9990157480314961, + "grad_norm": 13.73558521270752, + "learning_rate": 9.999787686490555e-07, + "loss": 0.7021, + "step": 1015 + }, + { + "epoch": 1.0, + "grad_norm": 18.500980377197266, + "learning_rate": 9.99971101846426e-07, + "loss": 0.6608, + "step": 1016 + }, + { + "epoch": 1.000984251968504, + "grad_norm": 18.25552749633789, + "learning_rate": 9.999622555845573e-07, + "loss": 1.0985, + "step": 1017 + }, + { + "epoch": 1.0019685039370079, + "grad_norm": 20.82244110107422, + "learning_rate": 9.999522298843875e-07, + "loss": 1.7193, + "step": 1018 + }, + { + "epoch": 1.0029527559055118, + "grad_norm": 8.279940605163574, + "learning_rate": 9.99941024769647e-07, + "loss": 0.3447, + "step": 1019 + }, + { + "epoch": 1.0039370078740157, + "grad_norm": 13.146919250488281, + "learning_rate": 9.999286402668578e-07, + "loss": 0.6574, + "step": 1020 + }, + { + "epoch": 1.0049212598425197, + "grad_norm": 15.445282936096191, + "learning_rate": 9.999150764053334e-07, + "loss": 0.7307, + "step": 1021 + }, + { + "epoch": 1.0059055118110236, + "grad_norm": 13.378528594970703, + "learning_rate": 9.999003332171784e-07, + "loss": 0.605, + "step": 1022 + }, + { + "epoch": 1.0068897637795275, + "grad_norm": 16.779502868652344, + "learning_rate": 9.998844107372894e-07, + "loss": 0.9642, + "step": 1023 + }, + { + "epoch": 1.0078740157480315, + "grad_norm": 15.733071327209473, + "learning_rate": 9.99867309003354e-07, + "loss": 0.7899, + "step": 1024 + }, + { + "epoch": 1.0088582677165354, + "grad_norm": 12.460901260375977, + "learning_rate": 9.998490280558511e-07, + "loss": 0.6818, + "step": 1025 + }, + { + "epoch": 1.0098425196850394, + "grad_norm": 17.373077392578125, + "learning_rate": 9.998295679380506e-07, + "loss": 0.9047, + "step": 1026 + }, + { + "epoch": 1.0108267716535433, + "grad_norm": 16.287940979003906, + "learning_rate": 9.998089286960135e-07, + "loss": 1.0251, + "step": 1027 + }, + { + "epoch": 1.0118110236220472, + "grad_norm": 13.410775184631348, + "learning_rate": 9.99787110378592e-07, + "loss": 0.635, + "step": 1028 + }, + { + "epoch": 1.0127952755905512, + "grad_norm": 16.206838607788086, + "learning_rate": 9.997641130374284e-07, + "loss": 0.9896, + "step": 1029 + }, + { + "epoch": 1.013779527559055, + "grad_norm": 16.609249114990234, + "learning_rate": 9.997399367269565e-07, + "loss": 0.8988, + "step": 1030 + }, + { + "epoch": 1.014763779527559, + "grad_norm": 14.937371253967285, + "learning_rate": 9.997145815044003e-07, + "loss": 0.6257, + "step": 1031 + }, + { + "epoch": 1.015748031496063, + "grad_norm": 9.599014282226562, + "learning_rate": 9.99688047429774e-07, + "loss": 0.4369, + "step": 1032 + }, + { + "epoch": 1.016732283464567, + "grad_norm": 12.918846130371094, + "learning_rate": 9.996603345658824e-07, + "loss": 0.7827, + "step": 1033 + }, + { + "epoch": 1.0177165354330708, + "grad_norm": 15.849491119384766, + "learning_rate": 9.996314429783204e-07, + "loss": 0.9601, + "step": 1034 + }, + { + "epoch": 1.0187007874015748, + "grad_norm": 17.180509567260742, + "learning_rate": 9.99601372735473e-07, + "loss": 0.9565, + "step": 1035 + }, + { + "epoch": 1.0196850393700787, + "grad_norm": 14.358114242553711, + "learning_rate": 9.995701239085143e-07, + "loss": 0.6667, + "step": 1036 + }, + { + "epoch": 1.0206692913385826, + "grad_norm": 10.355342864990234, + "learning_rate": 9.99537696571409e-07, + "loss": 0.4217, + "step": 1037 + }, + { + "epoch": 1.0216535433070866, + "grad_norm": 16.553524017333984, + "learning_rate": 9.995040908009107e-07, + "loss": 0.7592, + "step": 1038 + }, + { + "epoch": 1.0226377952755905, + "grad_norm": 16.797199249267578, + "learning_rate": 9.994693066765622e-07, + "loss": 0.8667, + "step": 1039 + }, + { + "epoch": 1.0236220472440944, + "grad_norm": 15.846183776855469, + "learning_rate": 9.994333442806959e-07, + "loss": 0.7705, + "step": 1040 + }, + { + "epoch": 1.0246062992125984, + "grad_norm": 17.184005737304688, + "learning_rate": 9.993962036984325e-07, + "loss": 0.9951, + "step": 1041 + }, + { + "epoch": 1.0255905511811023, + "grad_norm": 18.776084899902344, + "learning_rate": 9.99357885017682e-07, + "loss": 1.1144, + "step": 1042 + }, + { + "epoch": 1.0265748031496063, + "grad_norm": 18.937816619873047, + "learning_rate": 9.993183883291425e-07, + "loss": 1.0319, + "step": 1043 + }, + { + "epoch": 1.0275590551181102, + "grad_norm": 19.087995529174805, + "learning_rate": 9.992777137263007e-07, + "loss": 1.1595, + "step": 1044 + }, + { + "epoch": 1.0285433070866141, + "grad_norm": 12.790840148925781, + "learning_rate": 9.99235861305431e-07, + "loss": 0.6343, + "step": 1045 + }, + { + "epoch": 1.029527559055118, + "grad_norm": 22.165191650390625, + "learning_rate": 9.991928311655956e-07, + "loss": 1.2074, + "step": 1046 + }, + { + "epoch": 1.030511811023622, + "grad_norm": 14.055591583251953, + "learning_rate": 9.991486234086447e-07, + "loss": 0.8404, + "step": 1047 + }, + { + "epoch": 1.031496062992126, + "grad_norm": 24.25628089904785, + "learning_rate": 9.991032381392158e-07, + "loss": 1.5037, + "step": 1048 + }, + { + "epoch": 1.0324803149606299, + "grad_norm": 11.641840934753418, + "learning_rate": 9.990566754647328e-07, + "loss": 0.4995, + "step": 1049 + }, + { + "epoch": 1.0334645669291338, + "grad_norm": 10.425209045410156, + "learning_rate": 9.990089354954073e-07, + "loss": 0.568, + "step": 1050 + }, + { + "epoch": 1.0344488188976377, + "grad_norm": 12.596050262451172, + "learning_rate": 9.989600183442373e-07, + "loss": 0.7489, + "step": 1051 + }, + { + "epoch": 1.0354330708661417, + "grad_norm": 13.544334411621094, + "learning_rate": 9.98909924127007e-07, + "loss": 0.7327, + "step": 1052 + }, + { + "epoch": 1.0364173228346456, + "grad_norm": 17.814496994018555, + "learning_rate": 9.988586529622864e-07, + "loss": 1.3957, + "step": 1053 + }, + { + "epoch": 1.0374015748031495, + "grad_norm": 15.446733474731445, + "learning_rate": 9.98806204971432e-07, + "loss": 1.0428, + "step": 1054 + }, + { + "epoch": 1.0383858267716535, + "grad_norm": 12.223246574401855, + "learning_rate": 9.987525802785847e-07, + "loss": 0.7656, + "step": 1055 + }, + { + "epoch": 1.0393700787401574, + "grad_norm": 12.582673072814941, + "learning_rate": 9.986977790106715e-07, + "loss": 1.1611, + "step": 1056 + }, + { + "epoch": 1.0403543307086613, + "grad_norm": 9.084281921386719, + "learning_rate": 9.986418012974042e-07, + "loss": 0.4786, + "step": 1057 + }, + { + "epoch": 1.0413385826771653, + "grad_norm": 12.558609962463379, + "learning_rate": 9.985846472712786e-07, + "loss": 0.5765, + "step": 1058 + }, + { + "epoch": 1.0423228346456692, + "grad_norm": 15.489008903503418, + "learning_rate": 9.985263170675756e-07, + "loss": 0.9421, + "step": 1059 + }, + { + "epoch": 1.0433070866141732, + "grad_norm": 11.702895164489746, + "learning_rate": 9.98466810824359e-07, + "loss": 0.7738, + "step": 1060 + }, + { + "epoch": 1.044291338582677, + "grad_norm": 12.038683891296387, + "learning_rate": 9.984061286824774e-07, + "loss": 0.7882, + "step": 1061 + }, + { + "epoch": 1.045275590551181, + "grad_norm": 15.36705207824707, + "learning_rate": 9.983442707855616e-07, + "loss": 0.9898, + "step": 1062 + }, + { + "epoch": 1.046259842519685, + "grad_norm": 13.331137657165527, + "learning_rate": 9.982812372800262e-07, + "loss": 0.7618, + "step": 1063 + }, + { + "epoch": 1.047244094488189, + "grad_norm": 11.37984848022461, + "learning_rate": 9.982170283150678e-07, + "loss": 0.5399, + "step": 1064 + }, + { + "epoch": 1.0482283464566928, + "grad_norm": 13.407549858093262, + "learning_rate": 9.981516440426659e-07, + "loss": 0.8189, + "step": 1065 + }, + { + "epoch": 1.0492125984251968, + "grad_norm": 12.853594779968262, + "learning_rate": 9.98085084617581e-07, + "loss": 0.4776, + "step": 1066 + }, + { + "epoch": 1.0501968503937007, + "grad_norm": 11.08696460723877, + "learning_rate": 9.980173501973558e-07, + "loss": 0.4333, + "step": 1067 + }, + { + "epoch": 1.0511811023622046, + "grad_norm": 15.92361068725586, + "learning_rate": 9.97948440942314e-07, + "loss": 0.4207, + "step": 1068 + }, + { + "epoch": 1.0521653543307086, + "grad_norm": 15.281506538391113, + "learning_rate": 9.978783570155597e-07, + "loss": 1.0206, + "step": 1069 + }, + { + "epoch": 1.0531496062992125, + "grad_norm": 15.706804275512695, + "learning_rate": 9.978070985829782e-07, + "loss": 0.4865, + "step": 1070 + }, + { + "epoch": 1.0541338582677164, + "grad_norm": 19.01339340209961, + "learning_rate": 9.97734665813234e-07, + "loss": 0.897, + "step": 1071 + }, + { + "epoch": 1.0541338582677164, + "eval_Qnli-dev_cosine_accuracy": 0.68359375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8038332462310791, + "eval_Qnli-dev_cosine_ap": 0.6915503966365267, + "eval_Qnli-dev_cosine_f1": 0.6764227642276421, + "eval_Qnli-dev_cosine_f1_threshold": 0.7276865839958191, + "eval_Qnli-dev_cosine_precision": 0.5488126649076517, + "eval_Qnli-dev_cosine_recall": 0.8813559322033898, + "eval_Qnli-dev_dot_accuracy": 0.68359375, + "eval_Qnli-dev_dot_accuracy_threshold": 617.9757690429688, + "eval_Qnli-dev_dot_ap": 0.6914604071082934, + "eval_Qnli-dev_dot_f1": 0.6775244299674267, + "eval_Qnli-dev_dot_f1_threshold": 559.7400512695312, + "eval_Qnli-dev_dot_precision": 0.5502645502645502, + "eval_Qnli-dev_dot_recall": 0.8813559322033898, + "eval_Qnli-dev_euclidean_accuracy": 0.68359375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.36817741394043, + "eval_Qnli-dev_euclidean_ap": 0.6915804106776542, + "eval_Qnli-dev_euclidean_f1": 0.6764227642276421, + "eval_Qnli-dev_euclidean_f1_threshold": 20.461692810058594, + "eval_Qnli-dev_euclidean_precision": 0.5488126649076517, + "eval_Qnli-dev_euclidean_recall": 0.8813559322033898, + "eval_Qnli-dev_manhattan_accuracy": 0.681640625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 384.64373779296875, + "eval_Qnli-dev_manhattan_ap": 0.6911560630995964, + "eval_Qnli-dev_manhattan_f1": 0.67430441898527, + "eval_Qnli-dev_manhattan_f1_threshold": 451.5675048828125, + "eval_Qnli-dev_manhattan_precision": 0.5493333333333333, + "eval_Qnli-dev_manhattan_recall": 0.8728813559322034, + "eval_Qnli-dev_max_accuracy": 0.68359375, + "eval_Qnli-dev_max_accuracy_threshold": 617.9757690429688, + "eval_Qnli-dev_max_ap": 0.6915804106776542, + "eval_Qnli-dev_max_f1": 0.6775244299674267, + "eval_Qnli-dev_max_f1_threshold": 559.7400512695312, + "eval_Qnli-dev_max_precision": 0.5502645502645502, + "eval_Qnli-dev_max_recall": 0.8813559322033898, + "eval_allNLI-dev_cosine_accuracy": 0.70703125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.908595085144043, + "eval_allNLI-dev_cosine_ap": 0.5430466488954966, + "eval_allNLI-dev_cosine_f1": 0.5522388059701493, + "eval_allNLI-dev_cosine_f1_threshold": 0.8629225492477417, + "eval_allNLI-dev_cosine_precision": 0.4847161572052402, + "eval_allNLI-dev_cosine_recall": 0.6416184971098265, + "eval_allNLI-dev_dot_accuracy": 0.708984375, + "eval_allNLI-dev_dot_accuracy_threshold": 698.3015747070312, + "eval_allNLI-dev_dot_ap": 0.5430436315328802, + "eval_allNLI-dev_dot_f1": 0.5522388059701493, + "eval_allNLI-dev_dot_f1_threshold": 663.3272705078125, + "eval_allNLI-dev_dot_precision": 0.4847161572052402, + "eval_allNLI-dev_dot_recall": 0.6416184971098265, + "eval_allNLI-dev_euclidean_accuracy": 0.70703125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.85396957397461, + "eval_allNLI-dev_euclidean_ap": 0.5430008642299873, + "eval_allNLI-dev_euclidean_f1": 0.5522388059701493, + "eval_allNLI-dev_euclidean_f1_threshold": 14.51696491241455, + "eval_allNLI-dev_euclidean_precision": 0.4847161572052402, + "eval_allNLI-dev_euclidean_recall": 0.6416184971098265, + "eval_allNLI-dev_manhattan_accuracy": 0.705078125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 263.4867858886719, + "eval_allNLI-dev_manhattan_ap": 0.5413936285926788, + "eval_allNLI-dev_manhattan_f1": 0.5520581113801454, + "eval_allNLI-dev_manhattan_f1_threshold": 326.1763000488281, + "eval_allNLI-dev_manhattan_precision": 0.475, + "eval_allNLI-dev_manhattan_recall": 0.6589595375722543, + "eval_allNLI-dev_max_accuracy": 0.708984375, + "eval_allNLI-dev_max_accuracy_threshold": 698.3015747070312, + "eval_allNLI-dev_max_ap": 0.5430466488954966, + "eval_allNLI-dev_max_f1": 0.5522388059701493, + "eval_allNLI-dev_max_f1_threshold": 663.3272705078125, + "eval_allNLI-dev_max_precision": 0.4847161572052402, + "eval_allNLI-dev_max_recall": 0.6589595375722543, + "eval_loss": 1.0709744691848755, + "eval_runtime": 50.1989, + "eval_samples_per_second": 33.148, + "eval_sequential_score": 0.6915804106776542, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.6368197640293093, + "eval_sts-test_pearson_dot": 0.6367905887877633, + "eval_sts-test_pearson_euclidean": 0.6470519111681319, + "eval_sts-test_pearson_manhattan": 0.6467215914161899, + "eval_sts-test_pearson_max": 0.6470519111681319, + "eval_sts-test_spearman_cosine": 0.6345637125214598, + "eval_sts-test_spearman_dot": 0.6346041676617576, + "eval_sts-test_spearman_euclidean": 0.6345603114255941, + "eval_sts-test_spearman_manhattan": 0.6336825601846632, + "eval_sts-test_spearman_max": 0.6346041676617576, + "step": 1071 + }, + { + "epoch": 1.0551181102362204, + "grad_norm": 15.656901359558105, + "learning_rate": 9.976610588777717e-07, + "loss": 0.8402, + "step": 1072 + }, + { + "epoch": 1.0561023622047243, + "grad_norm": 15.366618156433105, + "learning_rate": 9.975862779508149e-07, + "loss": 0.7688, + "step": 1073 + }, + { + "epoch": 1.0570866141732282, + "grad_norm": 9.356171607971191, + "learning_rate": 9.975103232093656e-07, + "loss": 0.2184, + "step": 1074 + }, + { + "epoch": 1.0580708661417322, + "grad_norm": 15.986027717590332, + "learning_rate": 9.974331948332048e-07, + "loss": 0.863, + "step": 1075 + }, + { + "epoch": 1.0590551181102361, + "grad_norm": 11.812033653259277, + "learning_rate": 9.97354893004891e-07, + "loss": 0.63, + "step": 1076 + }, + { + "epoch": 1.06003937007874, + "grad_norm": 14.299402236938477, + "learning_rate": 9.972754179097602e-07, + "loss": 0.6715, + "step": 1077 + }, + { + "epoch": 1.061023622047244, + "grad_norm": 11.965864181518555, + "learning_rate": 9.971947697359257e-07, + "loss": 0.5824, + "step": 1078 + }, + { + "epoch": 1.0620078740157481, + "grad_norm": 10.061518669128418, + "learning_rate": 9.971129486742777e-07, + "loss": 0.4253, + "step": 1079 + }, + { + "epoch": 1.0629921259842519, + "grad_norm": 13.379720687866211, + "learning_rate": 9.970299549184814e-07, + "loss": 0.7626, + "step": 1080 + }, + { + "epoch": 1.063976377952756, + "grad_norm": 12.801810264587402, + "learning_rate": 9.969457886649791e-07, + "loss": 0.6314, + "step": 1081 + }, + { + "epoch": 1.0649606299212597, + "grad_norm": 11.5104398727417, + "learning_rate": 9.968604501129877e-07, + "loss": 0.6581, + "step": 1082 + }, + { + "epoch": 1.0659448818897639, + "grad_norm": 12.517723083496094, + "learning_rate": 9.967739394644982e-07, + "loss": 0.4651, + "step": 1083 + }, + { + "epoch": 1.0669291338582678, + "grad_norm": 20.153242111206055, + "learning_rate": 9.966862569242773e-07, + "loss": 1.3387, + "step": 1084 + }, + { + "epoch": 1.0679133858267718, + "grad_norm": 17.8863525390625, + "learning_rate": 9.965974026998642e-07, + "loss": 0.8808, + "step": 1085 + }, + { + "epoch": 1.0688976377952757, + "grad_norm": 15.032797813415527, + "learning_rate": 9.965073770015722e-07, + "loss": 0.7236, + "step": 1086 + }, + { + "epoch": 1.0698818897637796, + "grad_norm": 15.820505142211914, + "learning_rate": 9.964161800424873e-07, + "loss": 0.7806, + "step": 1087 + }, + { + "epoch": 1.0708661417322836, + "grad_norm": 20.563907623291016, + "learning_rate": 9.963238120384677e-07, + "loss": 1.3413, + "step": 1088 + }, + { + "epoch": 1.0718503937007875, + "grad_norm": 14.543024063110352, + "learning_rate": 9.962302732081432e-07, + "loss": 0.4676, + "step": 1089 + }, + { + "epoch": 1.0728346456692914, + "grad_norm": 9.891223907470703, + "learning_rate": 9.961355637729153e-07, + "loss": 0.3322, + "step": 1090 + }, + { + "epoch": 1.0738188976377954, + "grad_norm": 11.006906509399414, + "learning_rate": 9.960396839569557e-07, + "loss": 0.3032, + "step": 1091 + }, + { + "epoch": 1.0748031496062993, + "grad_norm": 18.213397979736328, + "learning_rate": 9.95942633987207e-07, + "loss": 0.7566, + "step": 1092 + }, + { + "epoch": 1.0757874015748032, + "grad_norm": 21.541196823120117, + "learning_rate": 9.95844414093381e-07, + "loss": 1.2515, + "step": 1093 + }, + { + "epoch": 1.0767716535433072, + "grad_norm": 17.573991775512695, + "learning_rate": 9.957450245079588e-07, + "loss": 1.1035, + "step": 1094 + }, + { + "epoch": 1.077755905511811, + "grad_norm": 14.521368980407715, + "learning_rate": 9.9564446546619e-07, + "loss": 0.5504, + "step": 1095 + }, + { + "epoch": 1.078740157480315, + "grad_norm": 27.325939178466797, + "learning_rate": 9.955427372060928e-07, + "loss": 1.2568, + "step": 1096 + }, + { + "epoch": 1.079724409448819, + "grad_norm": 21.24674415588379, + "learning_rate": 9.954398399684515e-07, + "loss": 1.0059, + "step": 1097 + }, + { + "epoch": 1.080708661417323, + "grad_norm": 19.656803131103516, + "learning_rate": 9.953357739968192e-07, + "loss": 0.9695, + "step": 1098 + }, + { + "epoch": 1.0816929133858268, + "grad_norm": 14.839178085327148, + "learning_rate": 9.952305395375137e-07, + "loss": 0.5669, + "step": 1099 + }, + { + "epoch": 1.0826771653543308, + "grad_norm": 13.978408813476562, + "learning_rate": 9.951241368396194e-07, + "loss": 0.6268, + "step": 1100 + }, + { + "epoch": 1.0836614173228347, + "grad_norm": 15.542141914367676, + "learning_rate": 9.950165661549856e-07, + "loss": 1.013, + "step": 1101 + }, + { + "epoch": 1.0846456692913387, + "grad_norm": 19.537845611572266, + "learning_rate": 9.949078277382262e-07, + "loss": 1.5633, + "step": 1102 + }, + { + "epoch": 1.0856299212598426, + "grad_norm": 15.735995292663574, + "learning_rate": 9.947979218467193e-07, + "loss": 1.3625, + "step": 1103 + }, + { + "epoch": 1.0866141732283465, + "grad_norm": 15.879256248474121, + "learning_rate": 9.946868487406057e-07, + "loss": 0.7289, + "step": 1104 + }, + { + "epoch": 1.0875984251968505, + "grad_norm": 14.243614196777344, + "learning_rate": 9.945746086827896e-07, + "loss": 1.0045, + "step": 1105 + }, + { + "epoch": 1.0885826771653544, + "grad_norm": 14.811985969543457, + "learning_rate": 9.94461201938937e-07, + "loss": 1.2376, + "step": 1106 + }, + { + "epoch": 1.0895669291338583, + "grad_norm": 10.17866325378418, + "learning_rate": 9.943466287774756e-07, + "loss": 0.4695, + "step": 1107 + }, + { + "epoch": 1.0905511811023623, + "grad_norm": 15.04418659210205, + "learning_rate": 9.942308894695932e-07, + "loss": 1.1059, + "step": 1108 + }, + { + "epoch": 1.0915354330708662, + "grad_norm": 11.269502639770508, + "learning_rate": 9.941139842892391e-07, + "loss": 0.6343, + "step": 1109 + }, + { + "epoch": 1.0925196850393701, + "grad_norm": 12.981204986572266, + "learning_rate": 9.939959135131214e-07, + "loss": 0.7101, + "step": 1110 + }, + { + "epoch": 1.093503937007874, + "grad_norm": 10.896045684814453, + "learning_rate": 9.938766774207064e-07, + "loss": 0.6253, + "step": 1111 + }, + { + "epoch": 1.094488188976378, + "grad_norm": 15.882280349731445, + "learning_rate": 9.937562762942202e-07, + "loss": 1.1293, + "step": 1112 + }, + { + "epoch": 1.095472440944882, + "grad_norm": 10.046773910522461, + "learning_rate": 9.936347104186452e-07, + "loss": 0.5038, + "step": 1113 + }, + { + "epoch": 1.0964566929133859, + "grad_norm": 14.951889991760254, + "learning_rate": 9.935119800817215e-07, + "loss": 0.8907, + "step": 1114 + }, + { + "epoch": 1.0974409448818898, + "grad_norm": 12.613266944885254, + "learning_rate": 9.933880855739446e-07, + "loss": 0.553, + "step": 1115 + }, + { + "epoch": 1.0984251968503937, + "grad_norm": 13.19701099395752, + "learning_rate": 9.932630271885662e-07, + "loss": 0.8102, + "step": 1116 + }, + { + "epoch": 1.0994094488188977, + "grad_norm": 15.783533096313477, + "learning_rate": 9.93136805221593e-07, + "loss": 0.904, + "step": 1117 + }, + { + "epoch": 1.1003937007874016, + "grad_norm": 12.4768648147583, + "learning_rate": 9.930094199717848e-07, + "loss": 0.5524, + "step": 1118 + }, + { + "epoch": 1.1013779527559056, + "grad_norm": 18.58526039123535, + "learning_rate": 9.92880871740656e-07, + "loss": 1.1347, + "step": 1119 + }, + { + "epoch": 1.1023622047244095, + "grad_norm": 13.961736679077148, + "learning_rate": 9.927511608324732e-07, + "loss": 0.4371, + "step": 1120 + }, + { + "epoch": 1.1033464566929134, + "grad_norm": 19.944766998291016, + "learning_rate": 9.926202875542546e-07, + "loss": 0.875, + "step": 1121 + }, + { + "epoch": 1.1043307086614174, + "grad_norm": 20.518611907958984, + "learning_rate": 9.924882522157706e-07, + "loss": 1.3085, + "step": 1122 + }, + { + "epoch": 1.1053149606299213, + "grad_norm": 12.748093605041504, + "learning_rate": 9.923550551295411e-07, + "loss": 0.7923, + "step": 1123 + }, + { + "epoch": 1.1062992125984252, + "grad_norm": 12.471030235290527, + "learning_rate": 9.922206966108367e-07, + "loss": 0.5889, + "step": 1124 + }, + { + "epoch": 1.1072834645669292, + "grad_norm": 13.881964683532715, + "learning_rate": 9.920851769776761e-07, + "loss": 0.5114, + "step": 1125 + }, + { + "epoch": 1.108267716535433, + "grad_norm": 14.893072128295898, + "learning_rate": 9.919484965508276e-07, + "loss": 0.6616, + "step": 1126 + }, + { + "epoch": 1.109251968503937, + "grad_norm": 15.041365623474121, + "learning_rate": 9.918106556538057e-07, + "loss": 0.9752, + "step": 1127 + }, + { + "epoch": 1.110236220472441, + "grad_norm": 17.84661865234375, + "learning_rate": 9.91671654612872e-07, + "loss": 0.6389, + "step": 1128 + }, + { + "epoch": 1.111220472440945, + "grad_norm": 17.089975357055664, + "learning_rate": 9.91531493757035e-07, + "loss": 0.9866, + "step": 1129 + }, + { + "epoch": 1.1122047244094488, + "grad_norm": 10.975593566894531, + "learning_rate": 9.913901734180476e-07, + "loss": 0.365, + "step": 1130 + }, + { + "epoch": 1.1131889763779528, + "grad_norm": 14.723518371582031, + "learning_rate": 9.912476939304069e-07, + "loss": 0.6243, + "step": 1131 + }, + { + "epoch": 1.1141732283464567, + "grad_norm": 12.603139877319336, + "learning_rate": 9.911040556313543e-07, + "loss": 0.5302, + "step": 1132 + }, + { + "epoch": 1.1151574803149606, + "grad_norm": 11.720128059387207, + "learning_rate": 9.909592588608738e-07, + "loss": 0.5457, + "step": 1133 + }, + { + "epoch": 1.1161417322834646, + "grad_norm": 15.438112258911133, + "learning_rate": 9.908133039616911e-07, + "loss": 0.7722, + "step": 1134 + }, + { + "epoch": 1.1171259842519685, + "grad_norm": 20.828332901000977, + "learning_rate": 9.90666191279274e-07, + "loss": 1.2737, + "step": 1135 + }, + { + "epoch": 1.1181102362204725, + "grad_norm": 15.74884033203125, + "learning_rate": 9.905179211618296e-07, + "loss": 0.7274, + "step": 1136 + }, + { + "epoch": 1.1190944881889764, + "grad_norm": 16.951387405395508, + "learning_rate": 9.903684939603054e-07, + "loss": 0.9102, + "step": 1137 + }, + { + "epoch": 1.1200787401574803, + "grad_norm": 13.51593017578125, + "learning_rate": 9.902179100283875e-07, + "loss": 0.673, + "step": 1138 + }, + { + "epoch": 1.1210629921259843, + "grad_norm": 13.578470230102539, + "learning_rate": 9.900661697224995e-07, + "loss": 0.5895, + "step": 1139 + }, + { + "epoch": 1.1220472440944882, + "grad_norm": 12.887831687927246, + "learning_rate": 9.899132734018024e-07, + "loss": 0.6718, + "step": 1140 + }, + { + "epoch": 1.1230314960629921, + "grad_norm": 14.972274780273438, + "learning_rate": 9.897592214281936e-07, + "loss": 0.9109, + "step": 1141 + }, + { + "epoch": 1.124015748031496, + "grad_norm": 11.685628890991211, + "learning_rate": 9.896040141663054e-07, + "loss": 0.4086, + "step": 1142 + }, + { + "epoch": 1.125, + "grad_norm": 15.600942611694336, + "learning_rate": 9.894476519835052e-07, + "loss": 0.8096, + "step": 1143 + }, + { + "epoch": 1.125984251968504, + "grad_norm": 11.58348274230957, + "learning_rate": 9.892901352498932e-07, + "loss": 0.6197, + "step": 1144 + }, + { + "epoch": 1.1269685039370079, + "grad_norm": 15.484606742858887, + "learning_rate": 9.891314643383037e-07, + "loss": 0.6548, + "step": 1145 + }, + { + "epoch": 1.1279527559055118, + "grad_norm": 12.262903213500977, + "learning_rate": 9.88971639624301e-07, + "loss": 0.7632, + "step": 1146 + }, + { + "epoch": 1.1289370078740157, + "grad_norm": 10.261554718017578, + "learning_rate": 9.888106614861823e-07, + "loss": 0.3288, + "step": 1147 + }, + { + "epoch": 1.1299212598425197, + "grad_norm": 11.71741008758545, + "learning_rate": 9.886485303049734e-07, + "loss": 0.6114, + "step": 1148 + }, + { + "epoch": 1.1309055118110236, + "grad_norm": 11.544774055480957, + "learning_rate": 9.884852464644305e-07, + "loss": 0.4121, + "step": 1149 + }, + { + "epoch": 1.1318897637795275, + "grad_norm": 15.385348320007324, + "learning_rate": 9.883208103510367e-07, + "loss": 0.6772, + "step": 1150 + }, + { + "epoch": 1.1328740157480315, + "grad_norm": 15.743762969970703, + "learning_rate": 9.88155222354004e-07, + "loss": 0.9555, + "step": 1151 + }, + { + "epoch": 1.1338582677165354, + "grad_norm": 11.80981731414795, + "learning_rate": 9.879884828652698e-07, + "loss": 0.5712, + "step": 1152 + }, + { + "epoch": 1.1348425196850394, + "grad_norm": 13.315688133239746, + "learning_rate": 9.878205922794972e-07, + "loss": 0.8391, + "step": 1153 + }, + { + "epoch": 1.1358267716535433, + "grad_norm": 14.4414644241333, + "learning_rate": 9.876515509940742e-07, + "loss": 0.6745, + "step": 1154 + }, + { + "epoch": 1.1368110236220472, + "grad_norm": 12.528006553649902, + "learning_rate": 9.874813594091122e-07, + "loss": 0.5267, + "step": 1155 + }, + { + "epoch": 1.1377952755905512, + "grad_norm": 16.966638565063477, + "learning_rate": 9.873100179274453e-07, + "loss": 1.0252, + "step": 1156 + }, + { + "epoch": 1.138779527559055, + "grad_norm": 10.528403282165527, + "learning_rate": 9.871375269546292e-07, + "loss": 0.4004, + "step": 1157 + }, + { + "epoch": 1.139763779527559, + "grad_norm": 17.006420135498047, + "learning_rate": 9.869638868989407e-07, + "loss": 0.925, + "step": 1158 + }, + { + "epoch": 1.140748031496063, + "grad_norm": 13.599428176879883, + "learning_rate": 9.867890981713764e-07, + "loss": 0.6741, + "step": 1159 + }, + { + "epoch": 1.141732283464567, + "grad_norm": 11.754899978637695, + "learning_rate": 9.866131611856514e-07, + "loss": 0.5167, + "step": 1160 + }, + { + "epoch": 1.1427165354330708, + "grad_norm": 13.304365158081055, + "learning_rate": 9.864360763581994e-07, + "loss": 0.6953, + "step": 1161 + }, + { + "epoch": 1.1437007874015748, + "grad_norm": 10.479991912841797, + "learning_rate": 9.862578441081699e-07, + "loss": 0.5611, + "step": 1162 + }, + { + "epoch": 1.1446850393700787, + "grad_norm": 19.750497817993164, + "learning_rate": 9.860784648574288e-07, + "loss": 1.0161, + "step": 1163 + }, + { + "epoch": 1.1456692913385826, + "grad_norm": 16.4873104095459, + "learning_rate": 9.858979390305573e-07, + "loss": 1.3154, + "step": 1164 + }, + { + "epoch": 1.1466535433070866, + "grad_norm": 15.80156135559082, + "learning_rate": 9.857162670548502e-07, + "loss": 0.6765, + "step": 1165 + }, + { + "epoch": 1.1476377952755905, + "grad_norm": 13.030824661254883, + "learning_rate": 9.855334493603147e-07, + "loss": 0.8017, + "step": 1166 + }, + { + "epoch": 1.1486220472440944, + "grad_norm": 14.154465675354004, + "learning_rate": 9.853494863796705e-07, + "loss": 0.8971, + "step": 1167 + }, + { + "epoch": 1.1496062992125984, + "grad_norm": 13.809678077697754, + "learning_rate": 9.851643785483478e-07, + "loss": 0.4928, + "step": 1168 + }, + { + "epoch": 1.1505905511811023, + "grad_norm": 12.12923526763916, + "learning_rate": 9.849781263044867e-07, + "loss": 0.6463, + "step": 1169 + }, + { + "epoch": 1.1515748031496063, + "grad_norm": 15.836526870727539, + "learning_rate": 9.847907300889358e-07, + "loss": 1.1188, + "step": 1170 + }, + { + "epoch": 1.1525590551181102, + "grad_norm": 15.554471015930176, + "learning_rate": 9.846021903452519e-07, + "loss": 0.7682, + "step": 1171 + }, + { + "epoch": 1.1535433070866141, + "grad_norm": 10.269407272338867, + "learning_rate": 9.844125075196983e-07, + "loss": 0.4076, + "step": 1172 + }, + { + "epoch": 1.154527559055118, + "grad_norm": 14.624336242675781, + "learning_rate": 9.842216820612438e-07, + "loss": 0.6429, + "step": 1173 + }, + { + "epoch": 1.155511811023622, + "grad_norm": 16.956457138061523, + "learning_rate": 9.840297144215615e-07, + "loss": 1.1348, + "step": 1174 + }, + { + "epoch": 1.156496062992126, + "grad_norm": 16.757246017456055, + "learning_rate": 9.838366050550287e-07, + "loss": 0.4246, + "step": 1175 + }, + { + "epoch": 1.1574803149606299, + "grad_norm": 15.031197547912598, + "learning_rate": 9.836423544187245e-07, + "loss": 0.8091, + "step": 1176 + }, + { + "epoch": 1.1584645669291338, + "grad_norm": 10.670025825500488, + "learning_rate": 9.834469629724293e-07, + "loss": 0.3452, + "step": 1177 + }, + { + "epoch": 1.1594488188976377, + "grad_norm": 13.752747535705566, + "learning_rate": 9.832504311786238e-07, + "loss": 0.7898, + "step": 1178 + }, + { + "epoch": 1.1604330708661417, + "grad_norm": 14.414322853088379, + "learning_rate": 9.830527595024885e-07, + "loss": 0.5909, + "step": 1179 + }, + { + "epoch": 1.1614173228346456, + "grad_norm": 17.380176544189453, + "learning_rate": 9.82853948411901e-07, + "loss": 1.0561, + "step": 1180 + }, + { + "epoch": 1.1624015748031495, + "grad_norm": 18.22872543334961, + "learning_rate": 9.826539983774365e-07, + "loss": 1.0296, + "step": 1181 + }, + { + "epoch": 1.1633858267716535, + "grad_norm": 11.854442596435547, + "learning_rate": 9.824529098723657e-07, + "loss": 0.5792, + "step": 1182 + }, + { + "epoch": 1.1643700787401574, + "grad_norm": 13.676430702209473, + "learning_rate": 9.822506833726537e-07, + "loss": 0.5314, + "step": 1183 + }, + { + "epoch": 1.1653543307086613, + "grad_norm": 17.245376586914062, + "learning_rate": 9.8204731935696e-07, + "loss": 0.8981, + "step": 1184 + }, + { + "epoch": 1.1663385826771653, + "grad_norm": 16.515378952026367, + "learning_rate": 9.818428183066356e-07, + "loss": 0.8561, + "step": 1185 + }, + { + "epoch": 1.1673228346456692, + "grad_norm": 10.962270736694336, + "learning_rate": 9.816371807057237e-07, + "loss": 0.6095, + "step": 1186 + }, + { + "epoch": 1.1683070866141732, + "grad_norm": 14.608488082885742, + "learning_rate": 9.81430407040957e-07, + "loss": 0.9399, + "step": 1187 + }, + { + "epoch": 1.169291338582677, + "grad_norm": 16.92159080505371, + "learning_rate": 9.812224978017573e-07, + "loss": 1.1345, + "step": 1188 + }, + { + "epoch": 1.170275590551181, + "grad_norm": 13.052497863769531, + "learning_rate": 9.810134534802346e-07, + "loss": 0.4627, + "step": 1189 + }, + { + "epoch": 1.171259842519685, + "grad_norm": 10.044450759887695, + "learning_rate": 9.80803274571185e-07, + "loss": 0.6207, + "step": 1190 + }, + { + "epoch": 1.172244094488189, + "grad_norm": 13.393180847167969, + "learning_rate": 9.80591961572091e-07, + "loss": 0.6967, + "step": 1191 + }, + { + "epoch": 1.1732283464566928, + "grad_norm": 10.935124397277832, + "learning_rate": 9.803795149831183e-07, + "loss": 0.498, + "step": 1192 + }, + { + "epoch": 1.1742125984251968, + "grad_norm": 12.005045890808105, + "learning_rate": 9.801659353071165e-07, + "loss": 0.7233, + "step": 1193 + }, + { + "epoch": 1.1751968503937007, + "grad_norm": 9.3803129196167, + "learning_rate": 9.799512230496172e-07, + "loss": 0.443, + "step": 1194 + }, + { + "epoch": 1.1761811023622046, + "grad_norm": 9.84170913696289, + "learning_rate": 9.79735378718832e-07, + "loss": 0.6022, + "step": 1195 + }, + { + "epoch": 1.1771653543307086, + "grad_norm": 12.042092323303223, + "learning_rate": 9.795184028256534e-07, + "loss": 0.5702, + "step": 1196 + }, + { + "epoch": 1.1781496062992125, + "grad_norm": 14.674236297607422, + "learning_rate": 9.793002958836506e-07, + "loss": 0.8733, + "step": 1197 + }, + { + "epoch": 1.1791338582677164, + "grad_norm": 11.526865005493164, + "learning_rate": 9.790810584090712e-07, + "loss": 0.432, + "step": 1198 + }, + { + "epoch": 1.1801181102362204, + "grad_norm": 12.801411628723145, + "learning_rate": 9.78860690920838e-07, + "loss": 0.6508, + "step": 1199 + }, + { + "epoch": 1.1811023622047245, + "grad_norm": 18.136734008789062, + "learning_rate": 9.78639193940549e-07, + "loss": 0.8595, + "step": 1200 + }, + { + "epoch": 1.1820866141732282, + "grad_norm": 13.083518981933594, + "learning_rate": 9.784165679924752e-07, + "loss": 0.6948, + "step": 1201 + }, + { + "epoch": 1.1830708661417324, + "grad_norm": 14.338788986206055, + "learning_rate": 9.781928136035604e-07, + "loss": 0.6306, + "step": 1202 + }, + { + "epoch": 1.1840551181102361, + "grad_norm": 12.982590675354004, + "learning_rate": 9.779679313034182e-07, + "loss": 0.9615, + "step": 1203 + }, + { + "epoch": 1.1850393700787403, + "grad_norm": 11.232832908630371, + "learning_rate": 9.777419216243333e-07, + "loss": 0.5652, + "step": 1204 + }, + { + "epoch": 1.186023622047244, + "grad_norm": 12.041215896606445, + "learning_rate": 9.77514785101258e-07, + "loss": 0.4482, + "step": 1205 + }, + { + "epoch": 1.1870078740157481, + "grad_norm": 16.642314910888672, + "learning_rate": 9.77286522271812e-07, + "loss": 0.8112, + "step": 1206 + }, + { + "epoch": 1.1879921259842519, + "grad_norm": 15.4851713180542, + "learning_rate": 9.77057133676281e-07, + "loss": 0.6432, + "step": 1207 + }, + { + "epoch": 1.188976377952756, + "grad_norm": 12.808008193969727, + "learning_rate": 9.768266198576147e-07, + "loss": 0.6797, + "step": 1208 + }, + { + "epoch": 1.1899606299212597, + "grad_norm": 11.812533378601074, + "learning_rate": 9.765949813614273e-07, + "loss": 0.4737, + "step": 1209 + }, + { + "epoch": 1.1909448818897639, + "grad_norm": 11.782402992248535, + "learning_rate": 9.763622187359942e-07, + "loss": 0.5752, + "step": 1210 + }, + { + "epoch": 1.1919291338582676, + "grad_norm": 11.302077293395996, + "learning_rate": 9.761283325322515e-07, + "loss": 0.4858, + "step": 1211 + }, + { + "epoch": 1.1929133858267718, + "grad_norm": 8.972127914428711, + "learning_rate": 9.758933233037952e-07, + "loss": 0.4213, + "step": 1212 + }, + { + "epoch": 1.1938976377952755, + "grad_norm": 10.73824691772461, + "learning_rate": 9.756571916068793e-07, + "loss": 0.3251, + "step": 1213 + }, + { + "epoch": 1.1948818897637796, + "grad_norm": 16.07195472717285, + "learning_rate": 9.754199380004145e-07, + "loss": 0.8442, + "step": 1214 + }, + { + "epoch": 1.1958661417322836, + "grad_norm": 11.785378456115723, + "learning_rate": 9.75181563045967e-07, + "loss": 0.4813, + "step": 1215 + }, + { + "epoch": 1.1968503937007875, + "grad_norm": 11.342597007751465, + "learning_rate": 9.749420673077574e-07, + "loss": 0.4635, + "step": 1216 + }, + { + "epoch": 1.1978346456692914, + "grad_norm": 11.755318641662598, + "learning_rate": 9.747014513526588e-07, + "loss": 0.4121, + "step": 1217 + }, + { + "epoch": 1.1988188976377954, + "grad_norm": 16.061979293823242, + "learning_rate": 9.744597157501957e-07, + "loss": 0.8145, + "step": 1218 + }, + { + "epoch": 1.1998031496062993, + "grad_norm": 20.689319610595703, + "learning_rate": 9.742168610725437e-07, + "loss": 1.7243, + "step": 1219 + }, + { + "epoch": 1.2007874015748032, + "grad_norm": 18.057886123657227, + "learning_rate": 9.73972887894526e-07, + "loss": 1.0789, + "step": 1220 + }, + { + "epoch": 1.2017716535433072, + "grad_norm": 15.676920890808105, + "learning_rate": 9.737277967936137e-07, + "loss": 0.9369, + "step": 1221 + }, + { + "epoch": 1.202755905511811, + "grad_norm": 11.642635345458984, + "learning_rate": 9.734815883499241e-07, + "loss": 0.3984, + "step": 1222 + }, + { + "epoch": 1.203740157480315, + "grad_norm": 13.799805641174316, + "learning_rate": 9.73234263146219e-07, + "loss": 0.9133, + "step": 1223 + }, + { + "epoch": 1.204724409448819, + "grad_norm": 15.339109420776367, + "learning_rate": 9.72985821767904e-07, + "loss": 0.9049, + "step": 1224 + }, + { + "epoch": 1.204724409448819, + "eval_Qnli-dev_cosine_accuracy": 0.693359375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8382962346076965, + "eval_Qnli-dev_cosine_ap": 0.7133253547180689, + "eval_Qnli-dev_cosine_f1": 0.7109515260323159, + "eval_Qnli-dev_cosine_f1_threshold": 0.8057848215103149, + "eval_Qnli-dev_cosine_precision": 0.616822429906542, + "eval_Qnli-dev_cosine_recall": 0.8389830508474576, + "eval_Qnli-dev_dot_accuracy": 0.693359375, + "eval_Qnli-dev_dot_accuracy_threshold": 643.9669189453125, + "eval_Qnli-dev_dot_ap": 0.7134047131919685, + "eval_Qnli-dev_dot_f1": 0.7109515260323159, + "eval_Qnli-dev_dot_f1_threshold": 619.1158447265625, + "eval_Qnli-dev_dot_precision": 0.616822429906542, + "eval_Qnli-dev_dot_recall": 0.8389830508474576, + "eval_Qnli-dev_euclidean_accuracy": 0.693359375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.761869430541992, + "eval_Qnli-dev_euclidean_ap": 0.7135035257803053, + "eval_Qnli-dev_euclidean_f1": 0.7096774193548387, + "eval_Qnli-dev_euclidean_f1_threshold": 17.31216049194336, + "eval_Qnli-dev_euclidean_precision": 0.6149068322981367, + "eval_Qnli-dev_euclidean_recall": 0.8389830508474576, + "eval_Qnli-dev_manhattan_accuracy": 0.693359375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 344.0986328125, + "eval_Qnli-dev_manhattan_ap": 0.714158789767608, + "eval_Qnli-dev_manhattan_f1": 0.7094474153297684, + "eval_Qnli-dev_manhattan_f1_threshold": 386.4290771484375, + "eval_Qnli-dev_manhattan_precision": 0.6123076923076923, + "eval_Qnli-dev_manhattan_recall": 0.8432203389830508, + "eval_Qnli-dev_max_accuracy": 0.693359375, + "eval_Qnli-dev_max_accuracy_threshold": 643.9669189453125, + "eval_Qnli-dev_max_ap": 0.714158789767608, + "eval_Qnli-dev_max_f1": 0.7109515260323159, + "eval_Qnli-dev_max_f1_threshold": 619.1158447265625, + "eval_Qnli-dev_max_precision": 0.616822429906542, + "eval_Qnli-dev_max_recall": 0.8432203389830508, + "eval_allNLI-dev_cosine_accuracy": 0.69140625, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9378935098648071, + "eval_allNLI-dev_cosine_ap": 0.5137236413286866, + "eval_allNLI-dev_cosine_f1": 0.5607476635514018, + "eval_allNLI-dev_cosine_f1_threshold": 0.844283401966095, + "eval_allNLI-dev_cosine_precision": 0.4143646408839779, + "eval_allNLI-dev_cosine_recall": 0.8670520231213873, + "eval_allNLI-dev_dot_accuracy": 0.69140625, + "eval_allNLI-dev_dot_accuracy_threshold": 721.7477416992188, + "eval_allNLI-dev_dot_ap": 0.5133162898135143, + "eval_allNLI-dev_dot_f1": 0.5607476635514018, + "eval_allNLI-dev_dot_f1_threshold": 648.6195068359375, + "eval_allNLI-dev_dot_precision": 0.4143646408839779, + "eval_allNLI-dev_dot_recall": 0.8670520231213873, + "eval_allNLI-dev_euclidean_accuracy": 0.69140625, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.768742561340332, + "eval_allNLI-dev_euclidean_ap": 0.5137088223264205, + "eval_allNLI-dev_euclidean_f1": 0.5607476635514018, + "eval_allNLI-dev_euclidean_f1_threshold": 15.467927932739258, + "eval_allNLI-dev_euclidean_precision": 0.4143646408839779, + "eval_allNLI-dev_euclidean_recall": 0.8670520231213873, + "eval_allNLI-dev_manhattan_accuracy": 0.69140625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 213.49217224121094, + "eval_allNLI-dev_manhattan_ap": 0.5139763222738567, + "eval_allNLI-dev_manhattan_f1": 0.5612052730696798, + "eval_allNLI-dev_manhattan_f1_threshold": 336.9559631347656, + "eval_allNLI-dev_manhattan_precision": 0.41620111731843573, + "eval_allNLI-dev_manhattan_recall": 0.861271676300578, + "eval_allNLI-dev_max_accuracy": 0.69140625, + "eval_allNLI-dev_max_accuracy_threshold": 721.7477416992188, + "eval_allNLI-dev_max_ap": 0.5139763222738567, + "eval_allNLI-dev_max_f1": 0.5612052730696798, + "eval_allNLI-dev_max_f1_threshold": 648.6195068359375, + "eval_allNLI-dev_max_precision": 0.41620111731843573, + "eval_allNLI-dev_max_recall": 0.8670520231213873, + "eval_loss": 0.9805020093917847, + "eval_runtime": 50.2249, + "eval_samples_per_second": 33.131, + "eval_sequential_score": 0.714158789767608, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.6277143846962348, + "eval_sts-test_pearson_dot": 0.6275456185076753, + "eval_sts-test_pearson_euclidean": 0.6547007840503307, + "eval_sts-test_pearson_manhattan": 0.6560218429796432, + "eval_sts-test_pearson_max": 0.6560218429796432, + "eval_sts-test_spearman_cosine": 0.6501180426970528, + "eval_sts-test_spearman_dot": 0.6498629605071217, + "eval_sts-test_spearman_euclidean": 0.6501079289119782, + "eval_sts-test_spearman_manhattan": 0.650492521252376, + "eval_sts-test_spearman_max": 0.650492521252376, + "step": 1224 + }, + { + "epoch": 1.205708661417323, + "grad_norm": 18.853090286254883, + "learning_rate": 9.727362648030254e-07, + "loss": 1.5579, + "step": 1225 + }, + { + "epoch": 1.2066929133858268, + "grad_norm": 11.721993446350098, + "learning_rate": 9.724855928422714e-07, + "loss": 0.5835, + "step": 1226 + }, + { + "epoch": 1.2076771653543308, + "grad_norm": 13.666594505310059, + "learning_rate": 9.722338064789688e-07, + "loss": 0.9229, + "step": 1227 + }, + { + "epoch": 1.2086614173228347, + "grad_norm": 11.524203300476074, + "learning_rate": 9.719809063090818e-07, + "loss": 0.657, + "step": 1228 + }, + { + "epoch": 1.2096456692913387, + "grad_norm": 12.093695640563965, + "learning_rate": 9.717268929312112e-07, + "loss": 0.8755, + "step": 1229 + }, + { + "epoch": 1.2106299212598426, + "grad_norm": 14.301728248596191, + "learning_rate": 9.714717669465932e-07, + "loss": 0.7818, + "step": 1230 + }, + { + "epoch": 1.2116141732283465, + "grad_norm": 10.158224105834961, + "learning_rate": 9.712155289590964e-07, + "loss": 0.4695, + "step": 1231 + }, + { + "epoch": 1.2125984251968505, + "grad_norm": 10.989998817443848, + "learning_rate": 9.709581795752225e-07, + "loss": 0.4954, + "step": 1232 + }, + { + "epoch": 1.2135826771653544, + "grad_norm": 11.121186256408691, + "learning_rate": 9.70699719404103e-07, + "loss": 0.6049, + "step": 1233 + }, + { + "epoch": 1.2145669291338583, + "grad_norm": 10.738130569458008, + "learning_rate": 9.704401490574993e-07, + "loss": 0.5966, + "step": 1234 + }, + { + "epoch": 1.2155511811023623, + "grad_norm": 13.900117874145508, + "learning_rate": 9.701794691497997e-07, + "loss": 0.7618, + "step": 1235 + }, + { + "epoch": 1.2165354330708662, + "grad_norm": 14.386923789978027, + "learning_rate": 9.699176802980199e-07, + "loss": 0.7156, + "step": 1236 + }, + { + "epoch": 1.2175196850393701, + "grad_norm": 14.443093299865723, + "learning_rate": 9.69654783121799e-07, + "loss": 0.951, + "step": 1237 + }, + { + "epoch": 1.218503937007874, + "grad_norm": 11.733538627624512, + "learning_rate": 9.693907782434008e-07, + "loss": 0.7208, + "step": 1238 + }, + { + "epoch": 1.219488188976378, + "grad_norm": 8.966593742370605, + "learning_rate": 9.691256662877099e-07, + "loss": 0.4287, + "step": 1239 + }, + { + "epoch": 1.220472440944882, + "grad_norm": 17.972139358520508, + "learning_rate": 9.688594478822317e-07, + "loss": 1.2094, + "step": 1240 + }, + { + "epoch": 1.2214566929133859, + "grad_norm": 12.227025032043457, + "learning_rate": 9.685921236570914e-07, + "loss": 0.5808, + "step": 1241 + }, + { + "epoch": 1.2224409448818898, + "grad_norm": 17.6931095123291, + "learning_rate": 9.683236942450294e-07, + "loss": 0.932, + "step": 1242 + }, + { + "epoch": 1.2234251968503937, + "grad_norm": 12.494466781616211, + "learning_rate": 9.680541602814047e-07, + "loss": 0.6208, + "step": 1243 + }, + { + "epoch": 1.2244094488188977, + "grad_norm": 13.920388221740723, + "learning_rate": 9.677835224041885e-07, + "loss": 1.0067, + "step": 1244 + }, + { + "epoch": 1.2253937007874016, + "grad_norm": 14.761850357055664, + "learning_rate": 9.67511781253966e-07, + "loss": 1.0854, + "step": 1245 + }, + { + "epoch": 1.2263779527559056, + "grad_norm": 10.390329360961914, + "learning_rate": 9.67238937473934e-07, + "loss": 0.3263, + "step": 1246 + }, + { + "epoch": 1.2273622047244095, + "grad_norm": 11.141219139099121, + "learning_rate": 9.669649917098983e-07, + "loss": 0.5505, + "step": 1247 + }, + { + "epoch": 1.2283464566929134, + "grad_norm": 15.410104751586914, + "learning_rate": 9.666899446102736e-07, + "loss": 0.8948, + "step": 1248 + }, + { + "epoch": 1.2293307086614174, + "grad_norm": 13.195194244384766, + "learning_rate": 9.664137968260814e-07, + "loss": 0.7472, + "step": 1249 + }, + { + "epoch": 1.2303149606299213, + "grad_norm": 15.664710998535156, + "learning_rate": 9.661365490109482e-07, + "loss": 1.1227, + "step": 1250 + }, + { + "epoch": 1.2312992125984252, + "grad_norm": 9.55164909362793, + "learning_rate": 9.658582018211045e-07, + "loss": 0.289, + "step": 1251 + }, + { + "epoch": 1.2322834645669292, + "grad_norm": 13.60750675201416, + "learning_rate": 9.655787559153824e-07, + "loss": 0.941, + "step": 1252 + }, + { + "epoch": 1.233267716535433, + "grad_norm": 10.023324012756348, + "learning_rate": 9.652982119552154e-07, + "loss": 0.4583, + "step": 1253 + }, + { + "epoch": 1.234251968503937, + "grad_norm": 13.188741683959961, + "learning_rate": 9.650165706046359e-07, + "loss": 0.6132, + "step": 1254 + }, + { + "epoch": 1.235236220472441, + "grad_norm": 14.773223876953125, + "learning_rate": 9.64733832530273e-07, + "loss": 0.688, + "step": 1255 + }, + { + "epoch": 1.236220472440945, + "grad_norm": 10.513198852539062, + "learning_rate": 9.644499984013525e-07, + "loss": 0.403, + "step": 1256 + }, + { + "epoch": 1.2372047244094488, + "grad_norm": 12.736245155334473, + "learning_rate": 9.64165068889694e-07, + "loss": 0.6865, + "step": 1257 + }, + { + "epoch": 1.2381889763779528, + "grad_norm": 15.945908546447754, + "learning_rate": 9.6387904466971e-07, + "loss": 1.0758, + "step": 1258 + }, + { + "epoch": 1.2391732283464567, + "grad_norm": 11.326615333557129, + "learning_rate": 9.635919264184043e-07, + "loss": 0.4865, + "step": 1259 + }, + { + "epoch": 1.2401574803149606, + "grad_norm": 15.17660140991211, + "learning_rate": 9.6330371481537e-07, + "loss": 0.6469, + "step": 1260 + }, + { + "epoch": 1.2411417322834646, + "grad_norm": 9.221441268920898, + "learning_rate": 9.630144105427883e-07, + "loss": 0.2623, + "step": 1261 + }, + { + "epoch": 1.2421259842519685, + "grad_norm": 16.187929153442383, + "learning_rate": 9.627240142854262e-07, + "loss": 1.0156, + "step": 1262 + }, + { + "epoch": 1.2431102362204725, + "grad_norm": 14.63491153717041, + "learning_rate": 9.624325267306357e-07, + "loss": 0.7141, + "step": 1263 + }, + { + "epoch": 1.2440944881889764, + "grad_norm": 15.21778678894043, + "learning_rate": 9.621399485683523e-07, + "loss": 0.6317, + "step": 1264 + }, + { + "epoch": 1.2450787401574803, + "grad_norm": 15.638778686523438, + "learning_rate": 9.618462804910917e-07, + "loss": 0.5897, + "step": 1265 + }, + { + "epoch": 1.2460629921259843, + "grad_norm": 17.20949935913086, + "learning_rate": 9.61551523193951e-07, + "loss": 0.8692, + "step": 1266 + }, + { + "epoch": 1.2470472440944882, + "grad_norm": 22.7119083404541, + "learning_rate": 9.612556773746038e-07, + "loss": 1.1336, + "step": 1267 + }, + { + "epoch": 1.2480314960629921, + "grad_norm": 9.964015007019043, + "learning_rate": 9.609587437333011e-07, + "loss": 0.2962, + "step": 1268 + }, + { + "epoch": 1.249015748031496, + "grad_norm": 13.330865859985352, + "learning_rate": 9.606607229728688e-07, + "loss": 0.6667, + "step": 1269 + }, + { + "epoch": 1.25, + "grad_norm": 14.649953842163086, + "learning_rate": 9.603616157987054e-07, + "loss": 0.4759, + "step": 1270 + }, + { + "epoch": 1.250984251968504, + "grad_norm": 9.555685997009277, + "learning_rate": 9.600614229187812e-07, + "loss": 0.3171, + "step": 1271 + }, + { + "epoch": 1.2519685039370079, + "grad_norm": 10.980401039123535, + "learning_rate": 9.597601450436364e-07, + "loss": 0.3023, + "step": 1272 + }, + { + "epoch": 1.2529527559055118, + "grad_norm": 19.766407012939453, + "learning_rate": 9.594577828863794e-07, + "loss": 1.6589, + "step": 1273 + }, + { + "epoch": 1.2539370078740157, + "grad_norm": 11.739914894104004, + "learning_rate": 9.591543371626842e-07, + "loss": 0.5558, + "step": 1274 + }, + { + "epoch": 1.2549212598425197, + "grad_norm": 15.003819465637207, + "learning_rate": 9.588498085907905e-07, + "loss": 0.6133, + "step": 1275 + }, + { + "epoch": 1.2559055118110236, + "grad_norm": 17.301162719726562, + "learning_rate": 9.585441978915008e-07, + "loss": 1.0877, + "step": 1276 + }, + { + "epoch": 1.2568897637795275, + "grad_norm": 9.885178565979004, + "learning_rate": 9.582375057881785e-07, + "loss": 0.4351, + "step": 1277 + }, + { + "epoch": 1.2578740157480315, + "grad_norm": 9.31080150604248, + "learning_rate": 9.579297330067476e-07, + "loss": 0.5071, + "step": 1278 + }, + { + "epoch": 1.2588582677165354, + "grad_norm": 12.793647766113281, + "learning_rate": 9.576208802756888e-07, + "loss": 0.706, + "step": 1279 + }, + { + "epoch": 1.2598425196850394, + "grad_norm": 10.657276153564453, + "learning_rate": 9.5731094832604e-07, + "loss": 0.5484, + "step": 1280 + }, + { + "epoch": 1.2608267716535433, + "grad_norm": 13.734872817993164, + "learning_rate": 9.569999378913925e-07, + "loss": 0.7437, + "step": 1281 + }, + { + "epoch": 1.2618110236220472, + "grad_norm": 8.950067520141602, + "learning_rate": 9.566878497078914e-07, + "loss": 0.3304, + "step": 1282 + }, + { + "epoch": 1.2627952755905512, + "grad_norm": 13.826399803161621, + "learning_rate": 9.563746845142326e-07, + "loss": 0.6643, + "step": 1283 + }, + { + "epoch": 1.263779527559055, + "grad_norm": 17.823461532592773, + "learning_rate": 9.5606044305166e-07, + "loss": 0.671, + "step": 1284 + }, + { + "epoch": 1.264763779527559, + "grad_norm": 14.35014820098877, + "learning_rate": 9.557451260639665e-07, + "loss": 0.5944, + "step": 1285 + }, + { + "epoch": 1.265748031496063, + "grad_norm": 13.089896202087402, + "learning_rate": 9.5542873429749e-07, + "loss": 0.4285, + "step": 1286 + }, + { + "epoch": 1.266732283464567, + "grad_norm": 15.769146919250488, + "learning_rate": 9.551112685011127e-07, + "loss": 1.1699, + "step": 1287 + }, + { + "epoch": 1.2677165354330708, + "grad_norm": 14.079538345336914, + "learning_rate": 9.547927294262583e-07, + "loss": 0.852, + "step": 1288 + }, + { + "epoch": 1.2687007874015748, + "grad_norm": 11.764381408691406, + "learning_rate": 9.544731178268913e-07, + "loss": 0.7029, + "step": 1289 + }, + { + "epoch": 1.2696850393700787, + "grad_norm": 13.370294570922852, + "learning_rate": 9.54152434459515e-07, + "loss": 0.6146, + "step": 1290 + }, + { + "epoch": 1.2706692913385826, + "grad_norm": 13.992321968078613, + "learning_rate": 9.538306800831695e-07, + "loss": 0.8509, + "step": 1291 + }, + { + "epoch": 1.2716535433070866, + "grad_norm": 9.824807167053223, + "learning_rate": 9.535078554594293e-07, + "loss": 0.3799, + "step": 1292 + }, + { + "epoch": 1.2726377952755905, + "grad_norm": 11.534995079040527, + "learning_rate": 9.531839613524029e-07, + "loss": 0.6748, + "step": 1293 + }, + { + "epoch": 1.2736220472440944, + "grad_norm": 10.216886520385742, + "learning_rate": 9.528589985287297e-07, + "loss": 0.5403, + "step": 1294 + }, + { + "epoch": 1.2746062992125984, + "grad_norm": 14.406759262084961, + "learning_rate": 9.525329677575787e-07, + "loss": 0.9597, + "step": 1295 + }, + { + "epoch": 1.2755905511811023, + "grad_norm": 11.903914451599121, + "learning_rate": 9.522058698106475e-07, + "loss": 0.5739, + "step": 1296 + }, + { + "epoch": 1.2765748031496063, + "grad_norm": 8.298938751220703, + "learning_rate": 9.518777054621582e-07, + "loss": 0.3519, + "step": 1297 + }, + { + "epoch": 1.2775590551181102, + "grad_norm": 18.943220138549805, + "learning_rate": 9.515484754888582e-07, + "loss": 1.0849, + "step": 1298 + }, + { + "epoch": 1.2785433070866141, + "grad_norm": 11.27268123626709, + "learning_rate": 9.512181806700167e-07, + "loss": 0.5726, + "step": 1299 + }, + { + "epoch": 1.279527559055118, + "grad_norm": 14.286341667175293, + "learning_rate": 9.508868217874233e-07, + "loss": 0.8147, + "step": 1300 + }, + { + "epoch": 1.280511811023622, + "grad_norm": 11.98044490814209, + "learning_rate": 9.505543996253865e-07, + "loss": 0.6012, + "step": 1301 + }, + { + "epoch": 1.281496062992126, + "grad_norm": 11.58612060546875, + "learning_rate": 9.502209149707312e-07, + "loss": 0.5484, + "step": 1302 + }, + { + "epoch": 1.2824803149606299, + "grad_norm": 15.902400970458984, + "learning_rate": 9.498863686127971e-07, + "loss": 0.831, + "step": 1303 + }, + { + "epoch": 1.2834645669291338, + "grad_norm": 12.765377044677734, + "learning_rate": 9.495507613434374e-07, + "loss": 0.5111, + "step": 1304 + }, + { + "epoch": 1.2844488188976377, + "grad_norm": 12.77717399597168, + "learning_rate": 9.492140939570159e-07, + "loss": 0.7046, + "step": 1305 + }, + { + "epoch": 1.2854330708661417, + "grad_norm": 21.239530563354492, + "learning_rate": 9.488763672504058e-07, + "loss": 1.3873, + "step": 1306 + }, + { + "epoch": 1.2864173228346456, + "grad_norm": 13.046000480651855, + "learning_rate": 9.485375820229876e-07, + "loss": 0.504, + "step": 1307 + }, + { + "epoch": 1.2874015748031495, + "grad_norm": 14.194633483886719, + "learning_rate": 9.481977390766475e-07, + "loss": 0.5431, + "step": 1308 + }, + { + "epoch": 1.2883858267716535, + "grad_norm": 13.691484451293945, + "learning_rate": 9.47856839215775e-07, + "loss": 0.6305, + "step": 1309 + }, + { + "epoch": 1.2893700787401574, + "grad_norm": 14.59090518951416, + "learning_rate": 9.475148832472613e-07, + "loss": 1.0762, + "step": 1310 + }, + { + "epoch": 1.2903543307086613, + "grad_norm": 10.760035514831543, + "learning_rate": 9.471718719804975e-07, + "loss": 0.5877, + "step": 1311 + }, + { + "epoch": 1.2913385826771653, + "grad_norm": 16.073673248291016, + "learning_rate": 9.468278062273726e-07, + "loss": 0.9715, + "step": 1312 + }, + { + "epoch": 1.2923228346456692, + "grad_norm": 12.38337230682373, + "learning_rate": 9.464826868022709e-07, + "loss": 0.6389, + "step": 1313 + }, + { + "epoch": 1.2933070866141732, + "grad_norm": 14.141475677490234, + "learning_rate": 9.461365145220715e-07, + "loss": 0.9998, + "step": 1314 + }, + { + "epoch": 1.294291338582677, + "grad_norm": 7.767505168914795, + "learning_rate": 9.457892902061452e-07, + "loss": 0.3464, + "step": 1315 + }, + { + "epoch": 1.295275590551181, + "grad_norm": 12.913028717041016, + "learning_rate": 9.454410146763525e-07, + "loss": 0.8084, + "step": 1316 + }, + { + "epoch": 1.296259842519685, + "grad_norm": 13.283123016357422, + "learning_rate": 9.450916887570429e-07, + "loss": 0.8931, + "step": 1317 + }, + { + "epoch": 1.297244094488189, + "grad_norm": 11.093908309936523, + "learning_rate": 9.447413132750515e-07, + "loss": 0.7312, + "step": 1318 + }, + { + "epoch": 1.298228346456693, + "grad_norm": 9.144620895385742, + "learning_rate": 9.443898890596977e-07, + "loss": 0.4963, + "step": 1319 + }, + { + "epoch": 1.2992125984251968, + "grad_norm": 9.624670028686523, + "learning_rate": 9.440374169427834e-07, + "loss": 0.6836, + "step": 1320 + }, + { + "epoch": 1.300196850393701, + "grad_norm": 13.262110710144043, + "learning_rate": 9.43683897758591e-07, + "loss": 0.9576, + "step": 1321 + }, + { + "epoch": 1.3011811023622046, + "grad_norm": 9.179229736328125, + "learning_rate": 9.433293323438807e-07, + "loss": 0.3098, + "step": 1322 + }, + { + "epoch": 1.3021653543307088, + "grad_norm": 9.003008842468262, + "learning_rate": 9.429737215378894e-07, + "loss": 0.4355, + "step": 1323 + }, + { + "epoch": 1.3031496062992125, + "grad_norm": 9.560617446899414, + "learning_rate": 9.426170661823285e-07, + "loss": 0.479, + "step": 1324 + }, + { + "epoch": 1.3041338582677167, + "grad_norm": 13.863944053649902, + "learning_rate": 9.422593671213816e-07, + "loss": 0.9156, + "step": 1325 + }, + { + "epoch": 1.3051181102362204, + "grad_norm": 15.559622764587402, + "learning_rate": 9.419006252017027e-07, + "loss": 0.8502, + "step": 1326 + }, + { + "epoch": 1.3061023622047245, + "grad_norm": 14.897092819213867, + "learning_rate": 9.415408412724141e-07, + "loss": 1.1117, + "step": 1327 + }, + { + "epoch": 1.3070866141732282, + "grad_norm": 14.557184219360352, + "learning_rate": 9.411800161851048e-07, + "loss": 0.6265, + "step": 1328 + }, + { + "epoch": 1.3080708661417324, + "grad_norm": 14.032054901123047, + "learning_rate": 9.408181507938279e-07, + "loss": 0.8953, + "step": 1329 + }, + { + "epoch": 1.3090551181102361, + "grad_norm": 9.538619041442871, + "learning_rate": 9.404552459550988e-07, + "loss": 0.3273, + "step": 1330 + }, + { + "epoch": 1.3100393700787403, + "grad_norm": 12.876323699951172, + "learning_rate": 9.400913025278937e-07, + "loss": 0.5961, + "step": 1331 + }, + { + "epoch": 1.311023622047244, + "grad_norm": 12.061412811279297, + "learning_rate": 9.397263213736464e-07, + "loss": 0.496, + "step": 1332 + }, + { + "epoch": 1.3120078740157481, + "grad_norm": 15.20823860168457, + "learning_rate": 9.393603033562471e-07, + "loss": 0.7741, + "step": 1333 + }, + { + "epoch": 1.3129921259842519, + "grad_norm": 14.523262977600098, + "learning_rate": 9.389932493420407e-07, + "loss": 0.7729, + "step": 1334 + }, + { + "epoch": 1.313976377952756, + "grad_norm": 13.138249397277832, + "learning_rate": 9.386251601998239e-07, + "loss": 0.6344, + "step": 1335 + }, + { + "epoch": 1.3149606299212597, + "grad_norm": 12.543108940124512, + "learning_rate": 9.38256036800843e-07, + "loss": 0.7011, + "step": 1336 + }, + { + "epoch": 1.3159448818897639, + "grad_norm": 12.87456226348877, + "learning_rate": 9.378858800187935e-07, + "loss": 0.8326, + "step": 1337 + }, + { + "epoch": 1.3169291338582676, + "grad_norm": 16.89803123474121, + "learning_rate": 9.375146907298157e-07, + "loss": 1.0148, + "step": 1338 + }, + { + "epoch": 1.3179133858267718, + "grad_norm": 12.465119361877441, + "learning_rate": 9.371424698124945e-07, + "loss": 0.6185, + "step": 1339 + }, + { + "epoch": 1.3188976377952755, + "grad_norm": 12.81068229675293, + "learning_rate": 9.367692181478564e-07, + "loss": 0.6741, + "step": 1340 + }, + { + "epoch": 1.3198818897637796, + "grad_norm": 12.96876335144043, + "learning_rate": 9.363949366193673e-07, + "loss": 0.5321, + "step": 1341 + }, + { + "epoch": 1.3208661417322833, + "grad_norm": 11.544039726257324, + "learning_rate": 9.360196261129313e-07, + "loss": 0.4967, + "step": 1342 + }, + { + "epoch": 1.3218503937007875, + "grad_norm": 12.825027465820312, + "learning_rate": 9.356432875168876e-07, + "loss": 0.5539, + "step": 1343 + }, + { + "epoch": 1.3228346456692912, + "grad_norm": 8.391443252563477, + "learning_rate": 9.352659217220091e-07, + "loss": 0.3033, + "step": 1344 + }, + { + "epoch": 1.3238188976377954, + "grad_norm": 10.91363525390625, + "learning_rate": 9.348875296215001e-07, + "loss": 0.3676, + "step": 1345 + }, + { + "epoch": 1.324803149606299, + "grad_norm": 14.17178726196289, + "learning_rate": 9.345081121109935e-07, + "loss": 0.9113, + "step": 1346 + }, + { + "epoch": 1.3257874015748032, + "grad_norm": 10.456360816955566, + "learning_rate": 9.341276700885501e-07, + "loss": 0.4146, + "step": 1347 + }, + { + "epoch": 1.326771653543307, + "grad_norm": 8.266129493713379, + "learning_rate": 9.337462044546548e-07, + "loss": 0.2253, + "step": 1348 + }, + { + "epoch": 1.327755905511811, + "grad_norm": 12.939192771911621, + "learning_rate": 9.333637161122166e-07, + "loss": 0.6699, + "step": 1349 + }, + { + "epoch": 1.328740157480315, + "grad_norm": 10.262524604797363, + "learning_rate": 9.329802059665637e-07, + "loss": 0.3462, + "step": 1350 + }, + { + "epoch": 1.329724409448819, + "grad_norm": 18.28936004638672, + "learning_rate": 9.325956749254437e-07, + "loss": 0.9698, + "step": 1351 + }, + { + "epoch": 1.330708661417323, + "grad_norm": 17.049875259399414, + "learning_rate": 9.322101238990206e-07, + "loss": 0.8619, + "step": 1352 + }, + { + "epoch": 1.3316929133858268, + "grad_norm": 11.190884590148926, + "learning_rate": 9.318235537998726e-07, + "loss": 0.5379, + "step": 1353 + }, + { + "epoch": 1.3326771653543308, + "grad_norm": 19.294692993164062, + "learning_rate": 9.314359655429897e-07, + "loss": 0.948, + "step": 1354 + }, + { + "epoch": 1.3336614173228347, + "grad_norm": 10.547256469726562, + "learning_rate": 9.310473600457721e-07, + "loss": 0.6026, + "step": 1355 + }, + { + "epoch": 1.3346456692913387, + "grad_norm": 11.508523941040039, + "learning_rate": 9.306577382280277e-07, + "loss": 0.485, + "step": 1356 + }, + { + "epoch": 1.3356299212598426, + "grad_norm": 15.994858741760254, + "learning_rate": 9.3026710101197e-07, + "loss": 0.9165, + "step": 1357 + }, + { + "epoch": 1.3366141732283465, + "grad_norm": 13.161906242370605, + "learning_rate": 9.298754493222155e-07, + "loss": 0.4844, + "step": 1358 + }, + { + "epoch": 1.3375984251968505, + "grad_norm": 17.234201431274414, + "learning_rate": 9.294827840857827e-07, + "loss": 0.7272, + "step": 1359 + }, + { + "epoch": 1.3385826771653544, + "grad_norm": 15.099180221557617, + "learning_rate": 9.290891062320884e-07, + "loss": 0.9066, + "step": 1360 + }, + { + "epoch": 1.3395669291338583, + "grad_norm": 18.66205406188965, + "learning_rate": 9.286944166929466e-07, + "loss": 1.0294, + "step": 1361 + }, + { + "epoch": 1.3405511811023623, + "grad_norm": 11.537887573242188, + "learning_rate": 9.282987164025658e-07, + "loss": 0.3974, + "step": 1362 + }, + { + "epoch": 1.3415354330708662, + "grad_norm": 15.751293182373047, + "learning_rate": 9.279020062975465e-07, + "loss": 0.9056, + "step": 1363 + }, + { + "epoch": 1.3425196850393701, + "grad_norm": 13.901390075683594, + "learning_rate": 9.275042873168802e-07, + "loss": 0.7404, + "step": 1364 + }, + { + "epoch": 1.343503937007874, + "grad_norm": 12.27584457397461, + "learning_rate": 9.271055604019458e-07, + "loss": 0.58, + "step": 1365 + }, + { + "epoch": 1.344488188976378, + "grad_norm": 13.272692680358887, + "learning_rate": 9.267058264965076e-07, + "loss": 0.7247, + "step": 1366 + }, + { + "epoch": 1.345472440944882, + "grad_norm": 10.464261054992676, + "learning_rate": 9.263050865467143e-07, + "loss": 0.629, + "step": 1367 + }, + { + "epoch": 1.3464566929133859, + "grad_norm": 13.73171329498291, + "learning_rate": 9.259033415010952e-07, + "loss": 0.848, + "step": 1368 + }, + { + "epoch": 1.3474409448818898, + "grad_norm": 12.19621753692627, + "learning_rate": 9.255005923105587e-07, + "loss": 0.7288, + "step": 1369 + }, + { + "epoch": 1.3484251968503937, + "grad_norm": 15.977216720581055, + "learning_rate": 9.250968399283902e-07, + "loss": 1.0294, + "step": 1370 + }, + { + "epoch": 1.3494094488188977, + "grad_norm": 10.58302116394043, + "learning_rate": 9.24692085310249e-07, + "loss": 0.715, + "step": 1371 + }, + { + "epoch": 1.3503937007874016, + "grad_norm": 12.168503761291504, + "learning_rate": 9.242863294141675e-07, + "loss": 0.9057, + "step": 1372 + }, + { + "epoch": 1.3513779527559056, + "grad_norm": 10.93958568572998, + "learning_rate": 9.238795732005475e-07, + "loss": 0.6427, + "step": 1373 + }, + { + "epoch": 1.3523622047244095, + "grad_norm": 12.779228210449219, + "learning_rate": 9.234718176321586e-07, + "loss": 0.6787, + "step": 1374 + }, + { + "epoch": 1.3533464566929134, + "grad_norm": 12.067599296569824, + "learning_rate": 9.230630636741358e-07, + "loss": 0.5274, + "step": 1375 + }, + { + "epoch": 1.3543307086614174, + "grad_norm": 11.218463897705078, + "learning_rate": 9.226533122939772e-07, + "loss": 0.3804, + "step": 1376 + }, + { + "epoch": 1.3553149606299213, + "grad_norm": 14.325352668762207, + "learning_rate": 9.222425644615421e-07, + "loss": 0.7182, + "step": 1377 + }, + { + "epoch": 1.3553149606299213, + "eval_Qnli-dev_cosine_accuracy": 0.6796875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.853144109249115, + "eval_Qnli-dev_cosine_ap": 0.7179080331231261, + "eval_Qnli-dev_cosine_f1": 0.6722408026755853, + "eval_Qnli-dev_cosine_f1_threshold": 0.7798495292663574, + "eval_Qnli-dev_cosine_precision": 0.5552486187845304, + "eval_Qnli-dev_cosine_recall": 0.8516949152542372, + "eval_Qnli-dev_dot_accuracy": 0.6796875, + "eval_Qnli-dev_dot_accuracy_threshold": 654.5177001953125, + "eval_Qnli-dev_dot_ap": 0.717394020945829, + "eval_Qnli-dev_dot_f1": 0.6722408026755853, + "eval_Qnli-dev_dot_f1_threshold": 598.5498657226562, + "eval_Qnli-dev_dot_precision": 0.5552486187845304, + "eval_Qnli-dev_dot_recall": 0.8516949152542372, + "eval_Qnli-dev_euclidean_accuracy": 0.6796875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.011018753051758, + "eval_Qnli-dev_euclidean_ap": 0.7178790801062136, + "eval_Qnli-dev_euclidean_f1": 0.6722408026755853, + "eval_Qnli-dev_euclidean_f1_threshold": 18.383129119873047, + "eval_Qnli-dev_euclidean_precision": 0.5552486187845304, + "eval_Qnli-dev_euclidean_recall": 0.8516949152542372, + "eval_Qnli-dev_manhattan_accuracy": 0.681640625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 343.0353698730469, + "eval_Qnli-dev_manhattan_ap": 0.7157757013285855, + "eval_Qnli-dev_manhattan_f1": 0.6722689075630252, + "eval_Qnli-dev_manhattan_f1_threshold": 403.62115478515625, + "eval_Qnli-dev_manhattan_precision": 0.5571030640668524, + "eval_Qnli-dev_manhattan_recall": 0.847457627118644, + "eval_Qnli-dev_max_accuracy": 0.681640625, + "eval_Qnli-dev_max_accuracy_threshold": 654.5177001953125, + "eval_Qnli-dev_max_ap": 0.7179080331231261, + "eval_Qnli-dev_max_f1": 0.6722689075630252, + "eval_Qnli-dev_max_f1_threshold": 598.5498657226562, + "eval_Qnli-dev_max_precision": 0.5571030640668524, + "eval_Qnli-dev_max_recall": 0.8516949152542372, + "eval_allNLI-dev_cosine_accuracy": 0.708984375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9117740392684937, + "eval_allNLI-dev_cosine_ap": 0.5411955037630216, + "eval_allNLI-dev_cosine_f1": 0.5541125541125541, + "eval_allNLI-dev_cosine_f1_threshold": 0.8525219559669495, + "eval_allNLI-dev_cosine_precision": 0.4429065743944637, + "eval_allNLI-dev_cosine_recall": 0.7398843930635838, + "eval_allNLI-dev_dot_accuracy": 0.7109375, + "eval_allNLI-dev_dot_accuracy_threshold": 698.7433471679688, + "eval_allNLI-dev_dot_ap": 0.5420203919369292, + "eval_allNLI-dev_dot_f1": 0.5550660792951542, + "eval_allNLI-dev_dot_f1_threshold": 656.9251708984375, + "eval_allNLI-dev_dot_precision": 0.4483985765124555, + "eval_allNLI-dev_dot_recall": 0.7283236994219653, + "eval_allNLI-dev_euclidean_accuracy": 0.708984375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.635345458984375, + "eval_allNLI-dev_euclidean_ap": 0.5412314172441877, + "eval_allNLI-dev_euclidean_f1": 0.5541125541125541, + "eval_allNLI-dev_euclidean_f1_threshold": 15.043611526489258, + "eval_allNLI-dev_euclidean_precision": 0.4429065743944637, + "eval_allNLI-dev_euclidean_recall": 0.7398843930635838, + "eval_allNLI-dev_manhattan_accuracy": 0.7109375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 256.5762939453125, + "eval_allNLI-dev_manhattan_ap": 0.5402439631602882, + "eval_allNLI-dev_manhattan_f1": 0.5557986870897156, + "eval_allNLI-dev_manhattan_f1_threshold": 328.49932861328125, + "eval_allNLI-dev_manhattan_precision": 0.4471830985915493, + "eval_allNLI-dev_manhattan_recall": 0.7341040462427746, + "eval_allNLI-dev_max_accuracy": 0.7109375, + "eval_allNLI-dev_max_accuracy_threshold": 698.7433471679688, + "eval_allNLI-dev_max_ap": 0.5420203919369292, + "eval_allNLI-dev_max_f1": 0.5557986870897156, + "eval_allNLI-dev_max_f1_threshold": 656.9251708984375, + "eval_allNLI-dev_max_precision": 0.4483985765124555, + "eval_allNLI-dev_max_recall": 0.7398843930635838, + "eval_loss": 0.9106085896492004, + "eval_runtime": 50.1244, + "eval_samples_per_second": 33.197, + "eval_sequential_score": 0.7179080331231261, + "eval_steps_per_second": 0.14, + "eval_sts-test_pearson_cosine": 0.6743663504496755, + "eval_sts-test_pearson_dot": 0.6743355546121415, + "eval_sts-test_pearson_euclidean": 0.695890205831361, + "eval_sts-test_pearson_manhattan": 0.6970364427571867, + "eval_sts-test_pearson_max": 0.6970364427571867, + "eval_sts-test_spearman_cosine": 0.6833160946913416, + "eval_sts-test_spearman_dot": 0.6834700837817894, + "eval_sts-test_spearman_euclidean": 0.683329520069759, + "eval_sts-test_spearman_manhattan": 0.6844960064491873, + "eval_sts-test_spearman_max": 0.6844960064491873, + "step": 1377 + }, + { + "epoch": 1.3562992125984252, + "grad_norm": 9.487778663635254, + "learning_rate": 9.218308211490479e-07, + "loss": 0.2835, + "step": 1378 + }, + { + "epoch": 1.3572834645669292, + "grad_norm": 17.76441764831543, + "learning_rate": 9.214180833310684e-07, + "loss": 0.9879, + "step": 1379 + }, + { + "epoch": 1.358267716535433, + "grad_norm": 13.848097801208496, + "learning_rate": 9.210043519845313e-07, + "loss": 0.7156, + "step": 1380 + }, + { + "epoch": 1.359251968503937, + "grad_norm": 13.420242309570312, + "learning_rate": 9.205896280887163e-07, + "loss": 0.6815, + "step": 1381 + }, + { + "epoch": 1.360236220472441, + "grad_norm": 19.13620948791504, + "learning_rate": 9.201739126252518e-07, + "loss": 1.3848, + "step": 1382 + }, + { + "epoch": 1.361220472440945, + "grad_norm": 8.881099700927734, + "learning_rate": 9.197572065781133e-07, + "loss": 0.3454, + "step": 1383 + }, + { + "epoch": 1.3622047244094488, + "grad_norm": 18.03289222717285, + "learning_rate": 9.193395109336216e-07, + "loss": 0.767, + "step": 1384 + }, + { + "epoch": 1.3631889763779528, + "grad_norm": 20.932422637939453, + "learning_rate": 9.189208266804389e-07, + "loss": 1.4349, + "step": 1385 + }, + { + "epoch": 1.3641732283464567, + "grad_norm": 18.77211570739746, + "learning_rate": 9.185011548095682e-07, + "loss": 1.0088, + "step": 1386 + }, + { + "epoch": 1.3651574803149606, + "grad_norm": 16.08696937561035, + "learning_rate": 9.180804963143496e-07, + "loss": 0.8369, + "step": 1387 + }, + { + "epoch": 1.3661417322834646, + "grad_norm": 16.0799503326416, + "learning_rate": 9.176588521904588e-07, + "loss": 0.8239, + "step": 1388 + }, + { + "epoch": 1.3671259842519685, + "grad_norm": 11.53136157989502, + "learning_rate": 9.172362234359041e-07, + "loss": 0.712, + "step": 1389 + }, + { + "epoch": 1.3681102362204725, + "grad_norm": 12.719602584838867, + "learning_rate": 9.168126110510246e-07, + "loss": 0.6316, + "step": 1390 + }, + { + "epoch": 1.3690944881889764, + "grad_norm": 11.988082885742188, + "learning_rate": 9.163880160384877e-07, + "loss": 0.539, + "step": 1391 + }, + { + "epoch": 1.3700787401574803, + "grad_norm": 16.022050857543945, + "learning_rate": 9.159624394032864e-07, + "loss": 0.9446, + "step": 1392 + }, + { + "epoch": 1.3710629921259843, + "grad_norm": 7.294853687286377, + "learning_rate": 9.155358821527371e-07, + "loss": 0.3727, + "step": 1393 + }, + { + "epoch": 1.3720472440944882, + "grad_norm": 10.92911434173584, + "learning_rate": 9.151083452964775e-07, + "loss": 0.7056, + "step": 1394 + }, + { + "epoch": 1.3730314960629921, + "grad_norm": 13.118983268737793, + "learning_rate": 9.146798298464642e-07, + "loss": 0.7193, + "step": 1395 + }, + { + "epoch": 1.374015748031496, + "grad_norm": 16.991138458251953, + "learning_rate": 9.14250336816969e-07, + "loss": 1.1535, + "step": 1396 + }, + { + "epoch": 1.375, + "grad_norm": 8.522515296936035, + "learning_rate": 9.13819867224579e-07, + "loss": 0.469, + "step": 1397 + }, + { + "epoch": 1.375984251968504, + "grad_norm": 7.388733863830566, + "learning_rate": 9.133884220881917e-07, + "loss": 0.3409, + "step": 1398 + }, + { + "epoch": 1.3769685039370079, + "grad_norm": 8.352952003479004, + "learning_rate": 9.129560024290145e-07, + "loss": 0.3558, + "step": 1399 + }, + { + "epoch": 1.3779527559055118, + "grad_norm": 13.402681350708008, + "learning_rate": 9.125226092705603e-07, + "loss": 0.8896, + "step": 1400 + }, + { + "epoch": 1.3789370078740157, + "grad_norm": 10.454214096069336, + "learning_rate": 9.120882436386474e-07, + "loss": 0.5629, + "step": 1401 + }, + { + "epoch": 1.3799212598425197, + "grad_norm": 13.530871391296387, + "learning_rate": 9.116529065613954e-07, + "loss": 0.8229, + "step": 1402 + }, + { + "epoch": 1.3809055118110236, + "grad_norm": 13.828805923461914, + "learning_rate": 9.112165990692231e-07, + "loss": 0.5177, + "step": 1403 + }, + { + "epoch": 1.3818897637795275, + "grad_norm": 16.07158088684082, + "learning_rate": 9.107793221948464e-07, + "loss": 0.8918, + "step": 1404 + }, + { + "epoch": 1.3828740157480315, + "grad_norm": 10.79946231842041, + "learning_rate": 9.103410769732755e-07, + "loss": 0.4176, + "step": 1405 + }, + { + "epoch": 1.3838582677165354, + "grad_norm": 11.43078899383545, + "learning_rate": 9.099018644418133e-07, + "loss": 0.5783, + "step": 1406 + }, + { + "epoch": 1.3848425196850394, + "grad_norm": 14.772703170776367, + "learning_rate": 9.094616856400512e-07, + "loss": 0.6884, + "step": 1407 + }, + { + "epoch": 1.3858267716535433, + "grad_norm": 16.311328887939453, + "learning_rate": 9.090205416098685e-07, + "loss": 0.584, + "step": 1408 + }, + { + "epoch": 1.3868110236220472, + "grad_norm": 10.50403881072998, + "learning_rate": 9.08578433395429e-07, + "loss": 0.2709, + "step": 1409 + }, + { + "epoch": 1.3877952755905512, + "grad_norm": 15.796014785766602, + "learning_rate": 9.081353620431783e-07, + "loss": 1.0391, + "step": 1410 + }, + { + "epoch": 1.388779527559055, + "grad_norm": 15.348348617553711, + "learning_rate": 9.07691328601842e-07, + "loss": 0.7955, + "step": 1411 + }, + { + "epoch": 1.389763779527559, + "grad_norm": 10.050495147705078, + "learning_rate": 9.072463341224232e-07, + "loss": 0.3269, + "step": 1412 + }, + { + "epoch": 1.390748031496063, + "grad_norm": 19.240175247192383, + "learning_rate": 9.068003796581991e-07, + "loss": 1.1099, + "step": 1413 + }, + { + "epoch": 1.391732283464567, + "grad_norm": 14.656514167785645, + "learning_rate": 9.063534662647195e-07, + "loss": 0.666, + "step": 1414 + }, + { + "epoch": 1.3927165354330708, + "grad_norm": 10.956231117248535, + "learning_rate": 9.059055949998039e-07, + "loss": 0.3297, + "step": 1415 + }, + { + "epoch": 1.3937007874015748, + "grad_norm": 11.060700416564941, + "learning_rate": 9.05456766923539e-07, + "loss": 0.6201, + "step": 1416 + }, + { + "epoch": 1.3946850393700787, + "grad_norm": 11.633536338806152, + "learning_rate": 9.05006983098276e-07, + "loss": 0.563, + "step": 1417 + }, + { + "epoch": 1.3956692913385826, + "grad_norm": 12.422494888305664, + "learning_rate": 9.045562445886289e-07, + "loss": 0.4748, + "step": 1418 + }, + { + "epoch": 1.3966535433070866, + "grad_norm": 7.529232501983643, + "learning_rate": 9.041045524614707e-07, + "loss": 0.2349, + "step": 1419 + }, + { + "epoch": 1.3976377952755905, + "grad_norm": 17.310638427734375, + "learning_rate": 9.03651907785932e-07, + "loss": 0.9462, + "step": 1420 + }, + { + "epoch": 1.3986220472440944, + "grad_norm": 8.823171615600586, + "learning_rate": 9.03198311633398e-07, + "loss": 0.3161, + "step": 1421 + }, + { + "epoch": 1.3996062992125984, + "grad_norm": 11.668909072875977, + "learning_rate": 9.027437650775055e-07, + "loss": 0.572, + "step": 1422 + }, + { + "epoch": 1.4005905511811023, + "grad_norm": 9.653392791748047, + "learning_rate": 9.022882691941417e-07, + "loss": 0.2561, + "step": 1423 + }, + { + "epoch": 1.4015748031496063, + "grad_norm": 10.670364379882812, + "learning_rate": 9.018318250614402e-07, + "loss": 0.3842, + "step": 1424 + }, + { + "epoch": 1.4025590551181102, + "grad_norm": 15.111839294433594, + "learning_rate": 9.01374433759779e-07, + "loss": 0.5957, + "step": 1425 + }, + { + "epoch": 1.4035433070866141, + "grad_norm": 16.207548141479492, + "learning_rate": 9.009160963717784e-07, + "loss": 0.7951, + "step": 1426 + }, + { + "epoch": 1.404527559055118, + "grad_norm": 5.495473861694336, + "learning_rate": 9.004568139822978e-07, + "loss": 0.1247, + "step": 1427 + }, + { + "epoch": 1.405511811023622, + "grad_norm": 17.916967391967773, + "learning_rate": 8.999965876784333e-07, + "loss": 1.1933, + "step": 1428 + }, + { + "epoch": 1.406496062992126, + "grad_norm": 9.349445343017578, + "learning_rate": 8.995354185495155e-07, + "loss": 0.3577, + "step": 1429 + }, + { + "epoch": 1.4074803149606299, + "grad_norm": 8.227457046508789, + "learning_rate": 8.990733076871063e-07, + "loss": 0.2469, + "step": 1430 + }, + { + "epoch": 1.4084645669291338, + "grad_norm": 16.075143814086914, + "learning_rate": 8.986102561849967e-07, + "loss": 0.6525, + "step": 1431 + }, + { + "epoch": 1.4094488188976377, + "grad_norm": 20.092731475830078, + "learning_rate": 8.981462651392041e-07, + "loss": 1.0177, + "step": 1432 + }, + { + "epoch": 1.4104330708661417, + "grad_norm": 21.532272338867188, + "learning_rate": 8.9768133564797e-07, + "loss": 1.1509, + "step": 1433 + }, + { + "epoch": 1.4114173228346456, + "grad_norm": 16.083892822265625, + "learning_rate": 8.972154688117568e-07, + "loss": 0.4675, + "step": 1434 + }, + { + "epoch": 1.4124015748031495, + "grad_norm": 16.772615432739258, + "learning_rate": 8.967486657332462e-07, + "loss": 0.6547, + "step": 1435 + }, + { + "epoch": 1.4133858267716535, + "grad_norm": 14.79323959350586, + "learning_rate": 8.962809275173347e-07, + "loss": 0.5524, + "step": 1436 + }, + { + "epoch": 1.4143700787401574, + "grad_norm": 16.565717697143555, + "learning_rate": 8.958122552711336e-07, + "loss": 1.0243, + "step": 1437 + }, + { + "epoch": 1.4153543307086613, + "grad_norm": 15.325310707092285, + "learning_rate": 8.953426501039641e-07, + "loss": 0.806, + "step": 1438 + }, + { + "epoch": 1.4163385826771653, + "grad_norm": 9.061014175415039, + "learning_rate": 8.94872113127356e-07, + "loss": 0.2455, + "step": 1439 + }, + { + "epoch": 1.4173228346456692, + "grad_norm": 12.157710075378418, + "learning_rate": 8.944006454550443e-07, + "loss": 0.5622, + "step": 1440 + }, + { + "epoch": 1.4183070866141732, + "grad_norm": 8.2247314453125, + "learning_rate": 8.939282482029671e-07, + "loss": 0.2282, + "step": 1441 + }, + { + "epoch": 1.419291338582677, + "grad_norm": 9.974912643432617, + "learning_rate": 8.934549224892629e-07, + "loss": 0.3157, + "step": 1442 + }, + { + "epoch": 1.420275590551181, + "grad_norm": 12.080114364624023, + "learning_rate": 8.929806694342677e-07, + "loss": 0.6138, + "step": 1443 + }, + { + "epoch": 1.421259842519685, + "grad_norm": 9.459236145019531, + "learning_rate": 8.925054901605124e-07, + "loss": 0.3076, + "step": 1444 + }, + { + "epoch": 1.422244094488189, + "grad_norm": 13.363850593566895, + "learning_rate": 8.9202938579272e-07, + "loss": 0.6837, + "step": 1445 + }, + { + "epoch": 1.423228346456693, + "grad_norm": 9.763716697692871, + "learning_rate": 8.915523574578038e-07, + "loss": 0.3984, + "step": 1446 + }, + { + "epoch": 1.4242125984251968, + "grad_norm": 18.2073917388916, + "learning_rate": 8.910744062848633e-07, + "loss": 1.1718, + "step": 1447 + }, + { + "epoch": 1.425196850393701, + "grad_norm": 12.019681930541992, + "learning_rate": 8.90595533405183e-07, + "loss": 0.5737, + "step": 1448 + }, + { + "epoch": 1.4261811023622046, + "grad_norm": 17.33150863647461, + "learning_rate": 8.901157399522286e-07, + "loss": 1.1418, + "step": 1449 + }, + { + "epoch": 1.4271653543307088, + "grad_norm": 11.90988826751709, + "learning_rate": 8.896350270616449e-07, + "loss": 0.6953, + "step": 1450 + }, + { + "epoch": 1.4281496062992125, + "grad_norm": 10.953588485717773, + "learning_rate": 8.891533958712529e-07, + "loss": 0.5401, + "step": 1451 + }, + { + "epoch": 1.4291338582677167, + "grad_norm": 10.7542724609375, + "learning_rate": 8.88670847521047e-07, + "loss": 0.6473, + "step": 1452 + }, + { + "epoch": 1.4301181102362204, + "grad_norm": 9.825413703918457, + "learning_rate": 8.881873831531928e-07, + "loss": 0.7759, + "step": 1453 + }, + { + "epoch": 1.4311023622047245, + "grad_norm": 11.431859970092773, + "learning_rate": 8.877030039120242e-07, + "loss": 0.7438, + "step": 1454 + }, + { + "epoch": 1.4320866141732282, + "grad_norm": 11.18675422668457, + "learning_rate": 8.872177109440398e-07, + "loss": 0.9636, + "step": 1455 + }, + { + "epoch": 1.4330708661417324, + "grad_norm": 11.937495231628418, + "learning_rate": 8.867315053979016e-07, + "loss": 0.8507, + "step": 1456 + }, + { + "epoch": 1.4340551181102361, + "grad_norm": 12.715885162353516, + "learning_rate": 8.862443884244314e-07, + "loss": 0.8462, + "step": 1457 + }, + { + "epoch": 1.4350393700787403, + "grad_norm": 8.754853248596191, + "learning_rate": 8.857563611766085e-07, + "loss": 0.3837, + "step": 1458 + }, + { + "epoch": 1.436023622047244, + "grad_norm": 9.570255279541016, + "learning_rate": 8.852674248095663e-07, + "loss": 0.2155, + "step": 1459 + }, + { + "epoch": 1.4370078740157481, + "grad_norm": 11.375158309936523, + "learning_rate": 8.847775804805907e-07, + "loss": 0.7255, + "step": 1460 + }, + { + "epoch": 1.4379921259842519, + "grad_norm": 12.96324634552002, + "learning_rate": 8.84286829349116e-07, + "loss": 0.6524, + "step": 1461 + }, + { + "epoch": 1.438976377952756, + "grad_norm": 9.111471176147461, + "learning_rate": 8.837951725767234e-07, + "loss": 0.3466, + "step": 1462 + }, + { + "epoch": 1.4399606299212597, + "grad_norm": 10.544808387756348, + "learning_rate": 8.833026113271374e-07, + "loss": 0.2716, + "step": 1463 + }, + { + "epoch": 1.4409448818897639, + "grad_norm": 8.537788391113281, + "learning_rate": 8.828091467662234e-07, + "loss": 0.3398, + "step": 1464 + }, + { + "epoch": 1.4419291338582676, + "grad_norm": 14.746088027954102, + "learning_rate": 8.82314780061985e-07, + "loss": 0.5691, + "step": 1465 + }, + { + "epoch": 1.4429133858267718, + "grad_norm": 17.383953094482422, + "learning_rate": 8.81819512384561e-07, + "loss": 0.865, + "step": 1466 + }, + { + "epoch": 1.4438976377952755, + "grad_norm": 11.366089820861816, + "learning_rate": 8.813233449062229e-07, + "loss": 0.3672, + "step": 1467 + }, + { + "epoch": 1.4448818897637796, + "grad_norm": 10.128392219543457, + "learning_rate": 8.808262788013718e-07, + "loss": 0.3376, + "step": 1468 + }, + { + "epoch": 1.4458661417322833, + "grad_norm": 14.306221008300781, + "learning_rate": 8.803283152465357e-07, + "loss": 0.8082, + "step": 1469 + }, + { + "epoch": 1.4468503937007875, + "grad_norm": 15.288216590881348, + "learning_rate": 8.798294554203674e-07, + "loss": 0.8659, + "step": 1470 + }, + { + "epoch": 1.4478346456692912, + "grad_norm": 17.740705490112305, + "learning_rate": 8.793297005036402e-07, + "loss": 0.6866, + "step": 1471 + }, + { + "epoch": 1.4488188976377954, + "grad_norm": 18.309263229370117, + "learning_rate": 8.788290516792471e-07, + "loss": 0.8329, + "step": 1472 + }, + { + "epoch": 1.449803149606299, + "grad_norm": 14.249978065490723, + "learning_rate": 8.783275101321962e-07, + "loss": 0.3695, + "step": 1473 + }, + { + "epoch": 1.4507874015748032, + "grad_norm": 17.316293716430664, + "learning_rate": 8.778250770496084e-07, + "loss": 0.8269, + "step": 1474 + }, + { + "epoch": 1.451771653543307, + "grad_norm": 16.028146743774414, + "learning_rate": 8.773217536207158e-07, + "loss": 0.6318, + "step": 1475 + }, + { + "epoch": 1.452755905511811, + "grad_norm": 10.932854652404785, + "learning_rate": 8.76817541036857e-07, + "loss": 0.4768, + "step": 1476 + }, + { + "epoch": 1.453740157480315, + "grad_norm": 21.78103256225586, + "learning_rate": 8.763124404914754e-07, + "loss": 1.113, + "step": 1477 + }, + { + "epoch": 1.454724409448819, + "grad_norm": 7.4316725730896, + "learning_rate": 8.758064531801166e-07, + "loss": 0.2887, + "step": 1478 + }, + { + "epoch": 1.455708661417323, + "grad_norm": 15.07807445526123, + "learning_rate": 8.752995803004247e-07, + "loss": 0.5199, + "step": 1479 + }, + { + "epoch": 1.4566929133858268, + "grad_norm": 15.122557640075684, + "learning_rate": 8.747918230521398e-07, + "loss": 0.6446, + "step": 1480 + }, + { + "epoch": 1.4576771653543308, + "grad_norm": 10.244226455688477, + "learning_rate": 8.742831826370956e-07, + "loss": 0.3986, + "step": 1481 + }, + { + "epoch": 1.4586614173228347, + "grad_norm": 11.979135513305664, + "learning_rate": 8.737736602592158e-07, + "loss": 0.479, + "step": 1482 + }, + { + "epoch": 1.4596456692913387, + "grad_norm": 9.878643035888672, + "learning_rate": 8.732632571245121e-07, + "loss": 0.404, + "step": 1483 + }, + { + "epoch": 1.4606299212598426, + "grad_norm": 12.459001541137695, + "learning_rate": 8.727519744410806e-07, + "loss": 0.7923, + "step": 1484 + }, + { + "epoch": 1.4616141732283465, + "grad_norm": 11.021946907043457, + "learning_rate": 8.722398134190995e-07, + "loss": 0.4652, + "step": 1485 + }, + { + "epoch": 1.4625984251968505, + "grad_norm": 9.663249969482422, + "learning_rate": 8.717267752708257e-07, + "loss": 0.2954, + "step": 1486 + }, + { + "epoch": 1.4635826771653544, + "grad_norm": 11.624634742736816, + "learning_rate": 8.712128612105924e-07, + "loss": 0.553, + "step": 1487 + }, + { + "epoch": 1.4645669291338583, + "grad_norm": 10.029973983764648, + "learning_rate": 8.706980724548058e-07, + "loss": 0.4778, + "step": 1488 + }, + { + "epoch": 1.4655511811023623, + "grad_norm": 14.350408554077148, + "learning_rate": 8.701824102219428e-07, + "loss": 0.7567, + "step": 1489 + }, + { + "epoch": 1.4665354330708662, + "grad_norm": 8.486898422241211, + "learning_rate": 8.696658757325474e-07, + "loss": 0.3517, + "step": 1490 + }, + { + "epoch": 1.4675196850393701, + "grad_norm": 11.681425094604492, + "learning_rate": 8.691484702092286e-07, + "loss": 0.4175, + "step": 1491 + }, + { + "epoch": 1.468503937007874, + "grad_norm": 10.027353286743164, + "learning_rate": 8.686301948766568e-07, + "loss": 0.4908, + "step": 1492 + }, + { + "epoch": 1.469488188976378, + "grad_norm": 9.131977081298828, + "learning_rate": 8.681110509615612e-07, + "loss": 0.2792, + "step": 1493 + }, + { + "epoch": 1.470472440944882, + "grad_norm": 9.341715812683105, + "learning_rate": 8.675910396927269e-07, + "loss": 0.4523, + "step": 1494 + }, + { + "epoch": 1.4714566929133859, + "grad_norm": 14.984341621398926, + "learning_rate": 8.670701623009921e-07, + "loss": 0.5355, + "step": 1495 + }, + { + "epoch": 1.4724409448818898, + "grad_norm": 13.480386734008789, + "learning_rate": 8.66548420019245e-07, + "loss": 0.5229, + "step": 1496 + }, + { + "epoch": 1.4734251968503937, + "grad_norm": 17.546493530273438, + "learning_rate": 8.66025814082421e-07, + "loss": 0.6926, + "step": 1497 + }, + { + "epoch": 1.4744094488188977, + "grad_norm": 16.34644317626953, + "learning_rate": 8.655023457274994e-07, + "loss": 0.8461, + "step": 1498 + }, + { + "epoch": 1.4753937007874016, + "grad_norm": 9.8609619140625, + "learning_rate": 8.649780161935012e-07, + "loss": 0.2661, + "step": 1499 + }, + { + "epoch": 1.4763779527559056, + "grad_norm": 15.149656295776367, + "learning_rate": 8.644528267214859e-07, + "loss": 0.6715, + "step": 1500 + }, + { + "epoch": 1.4773622047244095, + "grad_norm": 8.556966781616211, + "learning_rate": 8.639267785545478e-07, + "loss": 0.2671, + "step": 1501 + }, + { + "epoch": 1.4783464566929134, + "grad_norm": 17.204381942749023, + "learning_rate": 8.63399872937814e-07, + "loss": 1.0064, + "step": 1502 + }, + { + "epoch": 1.4793307086614174, + "grad_norm": 14.100544929504395, + "learning_rate": 8.628721111184414e-07, + "loss": 0.7656, + "step": 1503 + }, + { + "epoch": 1.4803149606299213, + "grad_norm": 8.810201644897461, + "learning_rate": 8.623434943456132e-07, + "loss": 0.2789, + "step": 1504 + }, + { + "epoch": 1.4812992125984252, + "grad_norm": 11.421368598937988, + "learning_rate": 8.61814023870536e-07, + "loss": 0.6413, + "step": 1505 + }, + { + "epoch": 1.4822834645669292, + "grad_norm": 10.439610481262207, + "learning_rate": 8.612837009464377e-07, + "loss": 0.3223, + "step": 1506 + }, + { + "epoch": 1.483267716535433, + "grad_norm": 11.429130554199219, + "learning_rate": 8.607525268285634e-07, + "loss": 0.464, + "step": 1507 + }, + { + "epoch": 1.484251968503937, + "grad_norm": 12.121988296508789, + "learning_rate": 8.602205027741731e-07, + "loss": 0.6149, + "step": 1508 + }, + { + "epoch": 1.485236220472441, + "grad_norm": 13.45350456237793, + "learning_rate": 8.596876300425384e-07, + "loss": 0.8092, + "step": 1509 + }, + { + "epoch": 1.486220472440945, + "grad_norm": 10.397159576416016, + "learning_rate": 8.591539098949401e-07, + "loss": 0.4549, + "step": 1510 + }, + { + "epoch": 1.4872047244094488, + "grad_norm": 11.52188777923584, + "learning_rate": 8.586193435946641e-07, + "loss": 0.6682, + "step": 1511 + }, + { + "epoch": 1.4881889763779528, + "grad_norm": 10.674049377441406, + "learning_rate": 8.580839324069998e-07, + "loss": 0.5025, + "step": 1512 + }, + { + "epoch": 1.4891732283464567, + "grad_norm": 8.651293754577637, + "learning_rate": 8.57547677599236e-07, + "loss": 0.1873, + "step": 1513 + }, + { + "epoch": 1.4901574803149606, + "grad_norm": 16.00859260559082, + "learning_rate": 8.570105804406583e-07, + "loss": 0.9181, + "step": 1514 + }, + { + "epoch": 1.4911417322834646, + "grad_norm": 9.345109939575195, + "learning_rate": 8.564726422025462e-07, + "loss": 0.3487, + "step": 1515 + }, + { + "epoch": 1.4921259842519685, + "grad_norm": 14.682822227478027, + "learning_rate": 8.5593386415817e-07, + "loss": 0.8338, + "step": 1516 + }, + { + "epoch": 1.4931102362204725, + "grad_norm": 13.973051071166992, + "learning_rate": 8.553942475827879e-07, + "loss": 0.8586, + "step": 1517 + }, + { + "epoch": 1.4940944881889764, + "grad_norm": 13.441484451293945, + "learning_rate": 8.548537937536425e-07, + "loss": 0.6894, + "step": 1518 + }, + { + "epoch": 1.4950787401574803, + "grad_norm": 15.134810447692871, + "learning_rate": 8.543125039499582e-07, + "loss": 0.7321, + "step": 1519 + }, + { + "epoch": 1.4960629921259843, + "grad_norm": 11.011528968811035, + "learning_rate": 8.537703794529387e-07, + "loss": 0.33, + "step": 1520 + }, + { + "epoch": 1.4970472440944882, + "grad_norm": 17.388662338256836, + "learning_rate": 8.532274215457625e-07, + "loss": 0.7501, + "step": 1521 + }, + { + "epoch": 1.4980314960629921, + "grad_norm": 14.477554321289062, + "learning_rate": 8.526836315135815e-07, + "loss": 0.6217, + "step": 1522 + }, + { + "epoch": 1.499015748031496, + "grad_norm": 12.057535171508789, + "learning_rate": 8.521390106435168e-07, + "loss": 0.6856, + "step": 1523 + }, + { + "epoch": 1.5, + "grad_norm": 11.690994262695312, + "learning_rate": 8.515935602246558e-07, + "loss": 0.4751, + "step": 1524 + }, + { + "epoch": 1.500984251968504, + "grad_norm": 12.48520278930664, + "learning_rate": 8.510472815480498e-07, + "loss": 0.4743, + "step": 1525 + }, + { + "epoch": 1.5019685039370079, + "grad_norm": 9.85148811340332, + "learning_rate": 8.505001759067106e-07, + "loss": 0.3669, + "step": 1526 + }, + { + "epoch": 1.5029527559055118, + "grad_norm": 10.154548645019531, + "learning_rate": 8.499522445956071e-07, + "loss": 0.5173, + "step": 1527 + }, + { + "epoch": 1.5039370078740157, + "grad_norm": 12.616178512573242, + "learning_rate": 8.494034889116626e-07, + "loss": 0.5997, + "step": 1528 + }, + { + "epoch": 1.5049212598425197, + "grad_norm": 9.951362609863281, + "learning_rate": 8.488539101537518e-07, + "loss": 0.5956, + "step": 1529 + }, + { + "epoch": 1.5059055118110236, + "grad_norm": 9.793926239013672, + "learning_rate": 8.483035096226975e-07, + "loss": 0.3, + "step": 1530 + }, + { + "epoch": 1.5059055118110236, + "eval_Qnli-dev_cosine_accuracy": 0.673828125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8367668986320496, + "eval_Qnli-dev_cosine_ap": 0.692443142157041, + "eval_Qnli-dev_cosine_f1": 0.6955074875207986, + "eval_Qnli-dev_cosine_f1_threshold": 0.7842345237731934, + "eval_Qnli-dev_cosine_precision": 0.5726027397260274, + "eval_Qnli-dev_cosine_recall": 0.885593220338983, + "eval_Qnli-dev_dot_accuracy": 0.671875, + "eval_Qnli-dev_dot_accuracy_threshold": 664.6419677734375, + "eval_Qnli-dev_dot_ap": 0.6914995082305544, + "eval_Qnli-dev_dot_f1": 0.6955074875207986, + "eval_Qnli-dev_dot_f1_threshold": 600.91162109375, + "eval_Qnli-dev_dot_precision": 0.5726027397260274, + "eval_Qnli-dev_dot_recall": 0.885593220338983, + "eval_Qnli-dev_euclidean_accuracy": 0.673828125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.815683364868164, + "eval_Qnli-dev_euclidean_ap": 0.6924753355581168, + "eval_Qnli-dev_euclidean_f1": 0.6955074875207986, + "eval_Qnli-dev_euclidean_f1_threshold": 18.18395233154297, + "eval_Qnli-dev_euclidean_precision": 0.5726027397260274, + "eval_Qnli-dev_euclidean_recall": 0.885593220338983, + "eval_Qnli-dev_manhattan_accuracy": 0.6796875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 314.6761169433594, + "eval_Qnli-dev_manhattan_ap": 0.6938588857164959, + "eval_Qnli-dev_manhattan_f1": 0.6910299003322258, + "eval_Qnli-dev_manhattan_f1_threshold": 402.5721435546875, + "eval_Qnli-dev_manhattan_precision": 0.5683060109289617, + "eval_Qnli-dev_manhattan_recall": 0.8813559322033898, + "eval_Qnli-dev_max_accuracy": 0.6796875, + "eval_Qnli-dev_max_accuracy_threshold": 664.6419677734375, + "eval_Qnli-dev_max_ap": 0.6938588857164959, + "eval_Qnli-dev_max_f1": 0.6955074875207986, + "eval_Qnli-dev_max_f1_threshold": 600.91162109375, + "eval_Qnli-dev_max_precision": 0.5726027397260274, + "eval_Qnli-dev_max_recall": 0.885593220338983, + "eval_allNLI-dev_cosine_accuracy": 0.69140625, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9463135600090027, + "eval_allNLI-dev_cosine_ap": 0.5314451457789088, + "eval_allNLI-dev_cosine_f1": 0.5584905660377358, + "eval_allNLI-dev_cosine_f1_threshold": 0.8485152721405029, + "eval_allNLI-dev_cosine_precision": 0.41456582633053224, + "eval_allNLI-dev_cosine_recall": 0.8554913294797688, + "eval_allNLI-dev_dot_accuracy": 0.69140625, + "eval_allNLI-dev_dot_accuracy_threshold": 725.194091796875, + "eval_allNLI-dev_dot_ap": 0.5315330035625176, + "eval_allNLI-dev_dot_f1": 0.5582417582417583, + "eval_allNLI-dev_dot_f1_threshold": 668.3628540039062, + "eval_allNLI-dev_dot_precision": 0.450354609929078, + "eval_allNLI-dev_dot_recall": 0.7341040462427746, + "eval_allNLI-dev_euclidean_accuracy": 0.69140625, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.070939064025879, + "eval_allNLI-dev_euclidean_ap": 0.5313673193329587, + "eval_allNLI-dev_euclidean_f1": 0.5584905660377358, + "eval_allNLI-dev_euclidean_f1_threshold": 15.238346099853516, + "eval_allNLI-dev_euclidean_precision": 0.41456582633053224, + "eval_allNLI-dev_euclidean_recall": 0.8554913294797688, + "eval_allNLI-dev_manhattan_accuracy": 0.69140625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 200.66372680664062, + "eval_allNLI-dev_manhattan_ap": 0.5306649040508786, + "eval_allNLI-dev_manhattan_f1": 0.5594713656387665, + "eval_allNLI-dev_manhattan_f1_threshold": 307.2613525390625, + "eval_allNLI-dev_manhattan_precision": 0.45195729537366547, + "eval_allNLI-dev_manhattan_recall": 0.7341040462427746, + "eval_allNLI-dev_max_accuracy": 0.69140625, + "eval_allNLI-dev_max_accuracy_threshold": 725.194091796875, + "eval_allNLI-dev_max_ap": 0.5315330035625176, + "eval_allNLI-dev_max_f1": 0.5594713656387665, + "eval_allNLI-dev_max_f1_threshold": 668.3628540039062, + "eval_allNLI-dev_max_precision": 0.45195729537366547, + "eval_allNLI-dev_max_recall": 0.8554913294797688, + "eval_loss": 0.9290673136711121, + "eval_runtime": 50.5796, + "eval_samples_per_second": 32.899, + "eval_sequential_score": 0.6938588857164959, + "eval_steps_per_second": 0.138, + "eval_sts-test_pearson_cosine": 0.6456792798211939, + "eval_sts-test_pearson_dot": 0.6455590013984889, + "eval_sts-test_pearson_euclidean": 0.6713297771774944, + "eval_sts-test_pearson_manhattan": 0.6717117227082865, + "eval_sts-test_pearson_max": 0.6717117227082865, + "eval_sts-test_spearman_cosine": 0.6659394036493681, + "eval_sts-test_spearman_dot": 0.6656832026779024, + "eval_sts-test_spearman_euclidean": 0.6658860601457895, + "eval_sts-test_spearman_manhattan": 0.6661336241238068, + "eval_sts-test_spearman_max": 0.6661336241238068, + "step": 1530 + }, + { + "epoch": 1.5068897637795275, + "grad_norm": 11.409811973571777, + "learning_rate": 8.477522886212673e-07, + "loss": 0.4104, + "step": 1531 + }, + { + "epoch": 1.5078740157480315, + "grad_norm": 7.25697135925293, + "learning_rate": 8.472002484541711e-07, + "loss": 0.1822, + "step": 1532 + }, + { + "epoch": 1.5088582677165354, + "grad_norm": 10.846244812011719, + "learning_rate": 8.466473904280579e-07, + "loss": 0.5589, + "step": 1533 + }, + { + "epoch": 1.5098425196850394, + "grad_norm": 12.356818199157715, + "learning_rate": 8.46093715851512e-07, + "loss": 0.7207, + "step": 1534 + }, + { + "epoch": 1.5108267716535433, + "grad_norm": 8.84495735168457, + "learning_rate": 8.455392260350507e-07, + "loss": 0.5159, + "step": 1535 + }, + { + "epoch": 1.5118110236220472, + "grad_norm": 13.065821647644043, + "learning_rate": 8.449839222911214e-07, + "loss": 0.5168, + "step": 1536 + }, + { + "epoch": 1.5127952755905512, + "grad_norm": 15.312318801879883, + "learning_rate": 8.444278059340971e-07, + "loss": 0.8943, + "step": 1537 + }, + { + "epoch": 1.513779527559055, + "grad_norm": 8.150442123413086, + "learning_rate": 8.438708782802748e-07, + "loss": 0.2219, + "step": 1538 + }, + { + "epoch": 1.514763779527559, + "grad_norm": 9.8677396774292, + "learning_rate": 8.43313140647872e-07, + "loss": 0.5076, + "step": 1539 + }, + { + "epoch": 1.515748031496063, + "grad_norm": 10.45460033416748, + "learning_rate": 8.427545943570226e-07, + "loss": 0.3928, + "step": 1540 + }, + { + "epoch": 1.516732283464567, + "grad_norm": 13.580182075500488, + "learning_rate": 8.421952407297752e-07, + "loss": 0.5391, + "step": 1541 + }, + { + "epoch": 1.5177165354330708, + "grad_norm": 11.698064804077148, + "learning_rate": 8.416350810900892e-07, + "loss": 0.4186, + "step": 1542 + }, + { + "epoch": 1.5187007874015748, + "grad_norm": 10.180648803710938, + "learning_rate": 8.410741167638317e-07, + "loss": 0.4881, + "step": 1543 + }, + { + "epoch": 1.5196850393700787, + "grad_norm": 17.744869232177734, + "learning_rate": 8.405123490787741e-07, + "loss": 0.9253, + "step": 1544 + }, + { + "epoch": 1.5206692913385826, + "grad_norm": 11.025421142578125, + "learning_rate": 8.3994977936459e-07, + "loss": 0.4182, + "step": 1545 + }, + { + "epoch": 1.5216535433070866, + "grad_norm": 10.687175750732422, + "learning_rate": 8.393864089528514e-07, + "loss": 0.5401, + "step": 1546 + }, + { + "epoch": 1.5226377952755905, + "grad_norm": 14.252847671508789, + "learning_rate": 8.388222391770245e-07, + "loss": 0.5194, + "step": 1547 + }, + { + "epoch": 1.5236220472440944, + "grad_norm": 13.089218139648438, + "learning_rate": 8.382572713724683e-07, + "loss": 0.5134, + "step": 1548 + }, + { + "epoch": 1.5246062992125984, + "grad_norm": 14.467389106750488, + "learning_rate": 8.376915068764308e-07, + "loss": 0.5772, + "step": 1549 + }, + { + "epoch": 1.5255905511811023, + "grad_norm": 7.608703136444092, + "learning_rate": 8.371249470280453e-07, + "loss": 0.231, + "step": 1550 + }, + { + "epoch": 1.5265748031496063, + "grad_norm": 14.168994903564453, + "learning_rate": 8.36557593168328e-07, + "loss": 0.859, + "step": 1551 + }, + { + "epoch": 1.5275590551181102, + "grad_norm": 12.910221099853516, + "learning_rate": 8.359894466401743e-07, + "loss": 0.4484, + "step": 1552 + }, + { + "epoch": 1.5285433070866141, + "grad_norm": 15.980766296386719, + "learning_rate": 8.354205087883556e-07, + "loss": 0.6175, + "step": 1553 + }, + { + "epoch": 1.529527559055118, + "grad_norm": 8.659918785095215, + "learning_rate": 8.348507809595169e-07, + "loss": 0.234, + "step": 1554 + }, + { + "epoch": 1.530511811023622, + "grad_norm": 13.041356086730957, + "learning_rate": 8.342802645021727e-07, + "loss": 0.9227, + "step": 1555 + }, + { + "epoch": 1.531496062992126, + "grad_norm": 7.043632984161377, + "learning_rate": 8.337089607667039e-07, + "loss": 0.1934, + "step": 1556 + }, + { + "epoch": 1.53248031496063, + "grad_norm": 10.571005821228027, + "learning_rate": 8.331368711053552e-07, + "loss": 0.405, + "step": 1557 + }, + { + "epoch": 1.5334645669291338, + "grad_norm": 19.392261505126953, + "learning_rate": 8.325639968722313e-07, + "loss": 1.234, + "step": 1558 + }, + { + "epoch": 1.534448818897638, + "grad_norm": 15.383576393127441, + "learning_rate": 8.319903394232942e-07, + "loss": 0.8299, + "step": 1559 + }, + { + "epoch": 1.5354330708661417, + "grad_norm": 11.755831718444824, + "learning_rate": 8.314159001163596e-07, + "loss": 0.3714, + "step": 1560 + }, + { + "epoch": 1.5364173228346458, + "grad_norm": 12.31116008758545, + "learning_rate": 8.308406803110936e-07, + "loss": 0.5257, + "step": 1561 + }, + { + "epoch": 1.5374015748031495, + "grad_norm": 13.971604347229004, + "learning_rate": 8.302646813690101e-07, + "loss": 0.6182, + "step": 1562 + }, + { + "epoch": 1.5383858267716537, + "grad_norm": 13.13081169128418, + "learning_rate": 8.296879046534668e-07, + "loss": 0.7028, + "step": 1563 + }, + { + "epoch": 1.5393700787401574, + "grad_norm": 13.45482063293457, + "learning_rate": 8.291103515296628e-07, + "loss": 0.4544, + "step": 1564 + }, + { + "epoch": 1.5403543307086616, + "grad_norm": 11.91739559173584, + "learning_rate": 8.285320233646345e-07, + "loss": 0.3698, + "step": 1565 + }, + { + "epoch": 1.5413385826771653, + "grad_norm": 10.708499908447266, + "learning_rate": 8.279529215272526e-07, + "loss": 0.5062, + "step": 1566 + }, + { + "epoch": 1.5423228346456694, + "grad_norm": 12.124205589294434, + "learning_rate": 8.273730473882198e-07, + "loss": 0.669, + "step": 1567 + }, + { + "epoch": 1.5433070866141732, + "grad_norm": 11.772865295410156, + "learning_rate": 8.267924023200662e-07, + "loss": 0.4676, + "step": 1568 + }, + { + "epoch": 1.5442913385826773, + "grad_norm": 10.194809913635254, + "learning_rate": 8.262109876971466e-07, + "loss": 0.351, + "step": 1569 + }, + { + "epoch": 1.545275590551181, + "grad_norm": 9.81915283203125, + "learning_rate": 8.256288048956379e-07, + "loss": 0.4107, + "step": 1570 + }, + { + "epoch": 1.5462598425196852, + "grad_norm": 14.78946304321289, + "learning_rate": 8.250458552935343e-07, + "loss": 0.8612, + "step": 1571 + }, + { + "epoch": 1.547244094488189, + "grad_norm": 14.549205780029297, + "learning_rate": 8.244621402706462e-07, + "loss": 0.669, + "step": 1572 + }, + { + "epoch": 1.548228346456693, + "grad_norm": 13.066632270812988, + "learning_rate": 8.238776612085943e-07, + "loss": 0.8359, + "step": 1573 + }, + { + "epoch": 1.5492125984251968, + "grad_norm": 11.537317276000977, + "learning_rate": 8.232924194908088e-07, + "loss": 0.4907, + "step": 1574 + }, + { + "epoch": 1.550196850393701, + "grad_norm": 13.774821281433105, + "learning_rate": 8.227064165025247e-07, + "loss": 0.7319, + "step": 1575 + }, + { + "epoch": 1.5511811023622046, + "grad_norm": 12.936616897583008, + "learning_rate": 8.221196536307788e-07, + "loss": 0.6074, + "step": 1576 + }, + { + "epoch": 1.5521653543307088, + "grad_norm": 12.369338989257812, + "learning_rate": 8.215321322644067e-07, + "loss": 0.3403, + "step": 1577 + }, + { + "epoch": 1.5531496062992125, + "grad_norm": 9.890181541442871, + "learning_rate": 8.209438537940393e-07, + "loss": 0.2655, + "step": 1578 + }, + { + "epoch": 1.5541338582677167, + "grad_norm": 12.744134902954102, + "learning_rate": 8.203548196120994e-07, + "loss": 0.5543, + "step": 1579 + }, + { + "epoch": 1.5551181102362204, + "grad_norm": 12.227079391479492, + "learning_rate": 8.197650311127981e-07, + "loss": 0.5024, + "step": 1580 + }, + { + "epoch": 1.5561023622047245, + "grad_norm": 12.182401657104492, + "learning_rate": 8.191744896921331e-07, + "loss": 0.4707, + "step": 1581 + }, + { + "epoch": 1.5570866141732282, + "grad_norm": 11.062856674194336, + "learning_rate": 8.185831967478831e-07, + "loss": 0.4331, + "step": 1582 + }, + { + "epoch": 1.5580708661417324, + "grad_norm": 10.086639404296875, + "learning_rate": 8.179911536796062e-07, + "loss": 0.2716, + "step": 1583 + }, + { + "epoch": 1.5590551181102361, + "grad_norm": 15.219388008117676, + "learning_rate": 8.173983618886358e-07, + "loss": 0.5427, + "step": 1584 + }, + { + "epoch": 1.5600393700787403, + "grad_norm": 10.917632102966309, + "learning_rate": 8.168048227780775e-07, + "loss": 0.4009, + "step": 1585 + }, + { + "epoch": 1.561023622047244, + "grad_norm": 13.485771179199219, + "learning_rate": 8.162105377528057e-07, + "loss": 0.4137, + "step": 1586 + }, + { + "epoch": 1.5620078740157481, + "grad_norm": 17.750478744506836, + "learning_rate": 8.156155082194607e-07, + "loss": 0.9003, + "step": 1587 + }, + { + "epoch": 1.5629921259842519, + "grad_norm": 15.75775146484375, + "learning_rate": 8.150197355864446e-07, + "loss": 0.7676, + "step": 1588 + }, + { + "epoch": 1.563976377952756, + "grad_norm": 10.124483108520508, + "learning_rate": 8.144232212639184e-07, + "loss": 0.3106, + "step": 1589 + }, + { + "epoch": 1.5649606299212597, + "grad_norm": 11.900956153869629, + "learning_rate": 8.138259666637987e-07, + "loss": 0.4484, + "step": 1590 + }, + { + "epoch": 1.5659448818897639, + "grad_norm": 13.709794998168945, + "learning_rate": 8.132279731997544e-07, + "loss": 0.5018, + "step": 1591 + }, + { + "epoch": 1.5669291338582676, + "grad_norm": 13.284584999084473, + "learning_rate": 8.126292422872034e-07, + "loss": 0.5949, + "step": 1592 + }, + { + "epoch": 1.5679133858267718, + "grad_norm": 8.018437385559082, + "learning_rate": 8.120297753433087e-07, + "loss": 0.2563, + "step": 1593 + }, + { + "epoch": 1.5688976377952755, + "grad_norm": 10.370757102966309, + "learning_rate": 8.114295737869759e-07, + "loss": 0.3456, + "step": 1594 + }, + { + "epoch": 1.5698818897637796, + "grad_norm": 12.048629760742188, + "learning_rate": 8.108286390388487e-07, + "loss": 0.7007, + "step": 1595 + }, + { + "epoch": 1.5708661417322833, + "grad_norm": 16.407394409179688, + "learning_rate": 8.102269725213071e-07, + "loss": 0.5122, + "step": 1596 + }, + { + "epoch": 1.5718503937007875, + "grad_norm": 10.813406944274902, + "learning_rate": 8.096245756584625e-07, + "loss": 0.5203, + "step": 1597 + }, + { + "epoch": 1.5728346456692912, + "grad_norm": 16.364131927490234, + "learning_rate": 8.090214498761551e-07, + "loss": 0.7651, + "step": 1598 + }, + { + "epoch": 1.5738188976377954, + "grad_norm": 11.831172943115234, + "learning_rate": 8.084175966019507e-07, + "loss": 0.5278, + "step": 1599 + }, + { + "epoch": 1.574803149606299, + "grad_norm": 13.435785293579102, + "learning_rate": 8.078130172651368e-07, + "loss": 0.5975, + "step": 1600 + }, + { + "epoch": 1.5757874015748032, + "grad_norm": 17.21648597717285, + "learning_rate": 8.072077132967193e-07, + "loss": 1.0192, + "step": 1601 + }, + { + "epoch": 1.576771653543307, + "grad_norm": 15.198051452636719, + "learning_rate": 8.066016861294198e-07, + "loss": 0.5717, + "step": 1602 + }, + { + "epoch": 1.577755905511811, + "grad_norm": 10.746184349060059, + "learning_rate": 8.059949371976711e-07, + "loss": 0.3574, + "step": 1603 + }, + { + "epoch": 1.5787401574803148, + "grad_norm": 12.760784149169922, + "learning_rate": 8.053874679376147e-07, + "loss": 0.6425, + "step": 1604 + }, + { + "epoch": 1.579724409448819, + "grad_norm": 8.901346206665039, + "learning_rate": 8.047792797870968e-07, + "loss": 0.2945, + "step": 1605 + }, + { + "epoch": 1.5807086614173227, + "grad_norm": 8.448673248291016, + "learning_rate": 8.041703741856657e-07, + "loss": 0.3246, + "step": 1606 + }, + { + "epoch": 1.5816929133858268, + "grad_norm": 11.05006217956543, + "learning_rate": 8.035607525745672e-07, + "loss": 0.6352, + "step": 1607 + }, + { + "epoch": 1.5826771653543306, + "grad_norm": 11.4871244430542, + "learning_rate": 8.029504163967423e-07, + "loss": 0.6531, + "step": 1608 + }, + { + "epoch": 1.5836614173228347, + "grad_norm": 7.738185882568359, + "learning_rate": 8.023393670968235e-07, + "loss": 0.251, + "step": 1609 + }, + { + "epoch": 1.5846456692913384, + "grad_norm": 14.763057708740234, + "learning_rate": 8.017276061211303e-07, + "loss": 0.6983, + "step": 1610 + }, + { + "epoch": 1.5856299212598426, + "grad_norm": 13.369607925415039, + "learning_rate": 8.011151349176681e-07, + "loss": 0.5046, + "step": 1611 + }, + { + "epoch": 1.5866141732283463, + "grad_norm": 10.108794212341309, + "learning_rate": 8.005019549361222e-07, + "loss": 0.3996, + "step": 1612 + }, + { + "epoch": 1.5875984251968505, + "grad_norm": 11.757378578186035, + "learning_rate": 7.998880676278558e-07, + "loss": 0.3577, + "step": 1613 + }, + { + "epoch": 1.5885826771653542, + "grad_norm": 12.969524383544922, + "learning_rate": 7.992734744459068e-07, + "loss": 0.6171, + "step": 1614 + }, + { + "epoch": 1.5895669291338583, + "grad_norm": 13.642068862915039, + "learning_rate": 7.986581768449833e-07, + "loss": 0.6428, + "step": 1615 + }, + { + "epoch": 1.590551181102362, + "grad_norm": 13.31014347076416, + "learning_rate": 7.98042176281461e-07, + "loss": 0.6469, + "step": 1616 + }, + { + "epoch": 1.5915354330708662, + "grad_norm": 10.514156341552734, + "learning_rate": 7.974254742133795e-07, + "loss": 0.2742, + "step": 1617 + }, + { + "epoch": 1.59251968503937, + "grad_norm": 12.404247283935547, + "learning_rate": 7.968080721004386e-07, + "loss": 0.4386, + "step": 1618 + }, + { + "epoch": 1.593503937007874, + "grad_norm": 14.54157543182373, + "learning_rate": 7.961899714039954e-07, + "loss": 0.6856, + "step": 1619 + }, + { + "epoch": 1.594488188976378, + "grad_norm": 10.717940330505371, + "learning_rate": 7.955711735870601e-07, + "loss": 0.4403, + "step": 1620 + }, + { + "epoch": 1.595472440944882, + "grad_norm": 10.8330717086792, + "learning_rate": 7.949516801142932e-07, + "loss": 0.2275, + "step": 1621 + }, + { + "epoch": 1.5964566929133859, + "grad_norm": 11.137269973754883, + "learning_rate": 7.943314924520017e-07, + "loss": 0.2993, + "step": 1622 + }, + { + "epoch": 1.5974409448818898, + "grad_norm": 11.189386367797852, + "learning_rate": 7.93710612068136e-07, + "loss": 0.4506, + "step": 1623 + }, + { + "epoch": 1.5984251968503937, + "grad_norm": 15.803678512573242, + "learning_rate": 7.930890404322857e-07, + "loss": 0.9605, + "step": 1624 + }, + { + "epoch": 1.5994094488188977, + "grad_norm": 13.05101203918457, + "learning_rate": 7.924667790156766e-07, + "loss": 0.3964, + "step": 1625 + }, + { + "epoch": 1.6003937007874016, + "grad_norm": 11.266312599182129, + "learning_rate": 7.918438292911674e-07, + "loss": 0.2884, + "step": 1626 + }, + { + "epoch": 1.6013779527559056, + "grad_norm": 20.96393394470215, + "learning_rate": 7.912201927332461e-07, + "loss": 0.9598, + "step": 1627 + }, + { + "epoch": 1.6023622047244095, + "grad_norm": 15.09812068939209, + "learning_rate": 7.905958708180258e-07, + "loss": 0.4528, + "step": 1628 + }, + { + "epoch": 1.6033464566929134, + "grad_norm": 14.083560943603516, + "learning_rate": 7.899708650232426e-07, + "loss": 0.5177, + "step": 1629 + }, + { + "epoch": 1.6043307086614174, + "grad_norm": 11.798062324523926, + "learning_rate": 7.893451768282508e-07, + "loss": 0.4524, + "step": 1630 + }, + { + "epoch": 1.6053149606299213, + "grad_norm": 12.8540678024292, + "learning_rate": 7.8871880771402e-07, + "loss": 0.4653, + "step": 1631 + }, + { + "epoch": 1.6062992125984252, + "grad_norm": 7.517286777496338, + "learning_rate": 7.880917591631312e-07, + "loss": 0.2467, + "step": 1632 + }, + { + "epoch": 1.6072834645669292, + "grad_norm": 14.2374267578125, + "learning_rate": 7.874640326597741e-07, + "loss": 0.4882, + "step": 1633 + }, + { + "epoch": 1.608267716535433, + "grad_norm": 14.022074699401855, + "learning_rate": 7.868356296897431e-07, + "loss": 0.5083, + "step": 1634 + }, + { + "epoch": 1.609251968503937, + "grad_norm": 18.19550895690918, + "learning_rate": 7.862065517404334e-07, + "loss": 1.1841, + "step": 1635 + }, + { + "epoch": 1.610236220472441, + "grad_norm": 18.766563415527344, + "learning_rate": 7.855768003008376e-07, + "loss": 0.7497, + "step": 1636 + }, + { + "epoch": 1.611220472440945, + "grad_norm": 18.93699836730957, + "learning_rate": 7.849463768615432e-07, + "loss": 1.2472, + "step": 1637 + }, + { + "epoch": 1.6122047244094488, + "grad_norm": 13.13559341430664, + "learning_rate": 7.843152829147278e-07, + "loss": 0.4748, + "step": 1638 + }, + { + "epoch": 1.6131889763779528, + "grad_norm": 6.983983993530273, + "learning_rate": 7.836835199541562e-07, + "loss": 0.2488, + "step": 1639 + }, + { + "epoch": 1.6141732283464567, + "grad_norm": 11.50744915008545, + "learning_rate": 7.830510894751764e-07, + "loss": 0.2891, + "step": 1640 + }, + { + "epoch": 1.6151574803149606, + "grad_norm": 8.553048133850098, + "learning_rate": 7.824179929747168e-07, + "loss": 0.1761, + "step": 1641 + }, + { + "epoch": 1.6161417322834646, + "grad_norm": 7.506103992462158, + "learning_rate": 7.817842319512818e-07, + "loss": 0.1631, + "step": 1642 + }, + { + "epoch": 1.6171259842519685, + "grad_norm": 12.174046516418457, + "learning_rate": 7.811498079049492e-07, + "loss": 0.7441, + "step": 1643 + }, + { + "epoch": 1.6181102362204725, + "grad_norm": 13.682055473327637, + "learning_rate": 7.805147223373658e-07, + "loss": 0.4533, + "step": 1644 + }, + { + "epoch": 1.6190944881889764, + "grad_norm": 8.8323392868042, + "learning_rate": 7.798789767517444e-07, + "loss": 0.2456, + "step": 1645 + }, + { + "epoch": 1.6200787401574803, + "grad_norm": 18.149415969848633, + "learning_rate": 7.792425726528596e-07, + "loss": 0.8355, + "step": 1646 + }, + { + "epoch": 1.6210629921259843, + "grad_norm": 14.547651290893555, + "learning_rate": 7.786055115470454e-07, + "loss": 0.5344, + "step": 1647 + }, + { + "epoch": 1.6220472440944882, + "grad_norm": 7.435754299163818, + "learning_rate": 7.779677949421902e-07, + "loss": 0.2838, + "step": 1648 + }, + { + "epoch": 1.6230314960629921, + "grad_norm": 12.586782455444336, + "learning_rate": 7.77329424347734e-07, + "loss": 0.5601, + "step": 1649 + }, + { + "epoch": 1.624015748031496, + "grad_norm": 11.83624267578125, + "learning_rate": 7.766904012746653e-07, + "loss": 0.4486, + "step": 1650 + }, + { + "epoch": 1.625, + "grad_norm": 5.291337490081787, + "learning_rate": 7.760507272355165e-07, + "loss": 0.1496, + "step": 1651 + }, + { + "epoch": 1.625984251968504, + "grad_norm": 13.773070335388184, + "learning_rate": 7.754104037443609e-07, + "loss": 0.7022, + "step": 1652 + }, + { + "epoch": 1.6269685039370079, + "grad_norm": 8.214975357055664, + "learning_rate": 7.747694323168091e-07, + "loss": 0.3057, + "step": 1653 + }, + { + "epoch": 1.6279527559055118, + "grad_norm": 9.962435722351074, + "learning_rate": 7.741278144700052e-07, + "loss": 0.1891, + "step": 1654 + }, + { + "epoch": 1.6289370078740157, + "grad_norm": 13.932888984680176, + "learning_rate": 7.734855517226234e-07, + "loss": 0.5046, + "step": 1655 + }, + { + "epoch": 1.6299212598425197, + "grad_norm": 10.47618579864502, + "learning_rate": 7.728426455948644e-07, + "loss": 0.3195, + "step": 1656 + }, + { + "epoch": 1.6309055118110236, + "grad_norm": 9.896806716918945, + "learning_rate": 7.721990976084517e-07, + "loss": 0.3934, + "step": 1657 + }, + { + "epoch": 1.6318897637795275, + "grad_norm": 12.589936256408691, + "learning_rate": 7.715549092866282e-07, + "loss": 0.5143, + "step": 1658 + }, + { + "epoch": 1.6328740157480315, + "grad_norm": 7.899942398071289, + "learning_rate": 7.70910082154152e-07, + "loss": 0.229, + "step": 1659 + }, + { + "epoch": 1.6338582677165354, + "grad_norm": 16.23533058166504, + "learning_rate": 7.702646177372938e-07, + "loss": 0.856, + "step": 1660 + }, + { + "epoch": 1.6348425196850394, + "grad_norm": 15.442038536071777, + "learning_rate": 7.696185175638325e-07, + "loss": 0.7887, + "step": 1661 + }, + { + "epoch": 1.6358267716535433, + "grad_norm": 14.28133773803711, + "learning_rate": 7.689717831630513e-07, + "loss": 0.6338, + "step": 1662 + }, + { + "epoch": 1.6368110236220472, + "grad_norm": 11.173591613769531, + "learning_rate": 7.683244160657353e-07, + "loss": 0.2418, + "step": 1663 + }, + { + "epoch": 1.6377952755905512, + "grad_norm": 8.022150993347168, + "learning_rate": 7.676764178041668e-07, + "loss": 0.2384, + "step": 1664 + }, + { + "epoch": 1.638779527559055, + "grad_norm": 16.501663208007812, + "learning_rate": 7.670277899121223e-07, + "loss": 0.8926, + "step": 1665 + }, + { + "epoch": 1.639763779527559, + "grad_norm": 14.105144500732422, + "learning_rate": 7.663785339248683e-07, + "loss": 0.607, + "step": 1666 + }, + { + "epoch": 1.640748031496063, + "grad_norm": 14.766664505004883, + "learning_rate": 7.657286513791578e-07, + "loss": 0.6438, + "step": 1667 + }, + { + "epoch": 1.641732283464567, + "grad_norm": 11.59138011932373, + "learning_rate": 7.650781438132276e-07, + "loss": 0.5639, + "step": 1668 + }, + { + "epoch": 1.6427165354330708, + "grad_norm": 9.606240272521973, + "learning_rate": 7.644270127667932e-07, + "loss": 0.3296, + "step": 1669 + }, + { + "epoch": 1.6437007874015748, + "grad_norm": 13.641767501831055, + "learning_rate": 7.637752597810456e-07, + "loss": 0.5224, + "step": 1670 + }, + { + "epoch": 1.6446850393700787, + "grad_norm": 16.4365291595459, + "learning_rate": 7.631228863986491e-07, + "loss": 0.8004, + "step": 1671 + }, + { + "epoch": 1.6456692913385826, + "grad_norm": 11.884933471679688, + "learning_rate": 7.624698941637352e-07, + "loss": 0.3586, + "step": 1672 + }, + { + "epoch": 1.6466535433070866, + "grad_norm": 12.327584266662598, + "learning_rate": 7.618162846219006e-07, + "loss": 0.3812, + "step": 1673 + }, + { + "epoch": 1.6476377952755905, + "grad_norm": 12.210389137268066, + "learning_rate": 7.611620593202037e-07, + "loss": 0.505, + "step": 1674 + }, + { + "epoch": 1.6486220472440944, + "grad_norm": 11.03765869140625, + "learning_rate": 7.605072198071592e-07, + "loss": 0.4731, + "step": 1675 + }, + { + "epoch": 1.6496062992125984, + "grad_norm": 15.110365867614746, + "learning_rate": 7.598517676327368e-07, + "loss": 0.7902, + "step": 1676 + }, + { + "epoch": 1.6505905511811023, + "grad_norm": 11.443836212158203, + "learning_rate": 7.591957043483555e-07, + "loss": 0.3093, + "step": 1677 + }, + { + "epoch": 1.6515748031496063, + "grad_norm": 15.809693336486816, + "learning_rate": 7.585390315068812e-07, + "loss": 0.7674, + "step": 1678 + }, + { + "epoch": 1.6525590551181102, + "grad_norm": 10.619746208190918, + "learning_rate": 7.578817506626224e-07, + "loss": 0.6227, + "step": 1679 + }, + { + "epoch": 1.6535433070866141, + "grad_norm": 12.997358322143555, + "learning_rate": 7.572238633713268e-07, + "loss": 0.7726, + "step": 1680 + }, + { + "epoch": 1.654527559055118, + "grad_norm": 10.946554183959961, + "learning_rate": 7.565653711901778e-07, + "loss": 0.5234, + "step": 1681 + }, + { + "epoch": 1.655511811023622, + "grad_norm": 11.154240608215332, + "learning_rate": 7.559062756777898e-07, + "loss": 0.575, + "step": 1682 + }, + { + "epoch": 1.656496062992126, + "grad_norm": 13.539304733276367, + "learning_rate": 7.552465783942058e-07, + "loss": 0.5614, + "step": 1683 + }, + { + "epoch": 1.656496062992126, + "eval_Qnli-dev_cosine_accuracy": 0.66796875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8061826229095459, + "eval_Qnli-dev_cosine_ap": 0.6684281236316783, + "eval_Qnli-dev_cosine_f1": 0.6877192982456141, + "eval_Qnli-dev_cosine_f1_threshold": 0.7795990705490112, + "eval_Qnli-dev_cosine_precision": 0.5868263473053892, + "eval_Qnli-dev_cosine_recall": 0.8305084745762712, + "eval_Qnli-dev_dot_accuracy": 0.666015625, + "eval_Qnli-dev_dot_accuracy_threshold": 631.760986328125, + "eval_Qnli-dev_dot_ap": 0.6690229905089764, + "eval_Qnli-dev_dot_f1": 0.6877192982456141, + "eval_Qnli-dev_dot_f1_threshold": 596.4973754882812, + "eval_Qnli-dev_dot_precision": 0.5868263473053892, + "eval_Qnli-dev_dot_recall": 0.8305084745762712, + "eval_Qnli-dev_euclidean_accuracy": 0.666015625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.140230178833008, + "eval_Qnli-dev_euclidean_ap": 0.6681454561406752, + "eval_Qnli-dev_euclidean_f1": 0.6877192982456141, + "eval_Qnli-dev_euclidean_f1_threshold": 18.370464324951172, + "eval_Qnli-dev_euclidean_precision": 0.5868263473053892, + "eval_Qnli-dev_euclidean_recall": 0.8305084745762712, + "eval_Qnli-dev_manhattan_accuracy": 0.671875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 375.9737548828125, + "eval_Qnli-dev_manhattan_ap": 0.6704993708527434, + "eval_Qnli-dev_manhattan_f1": 0.6889279437609841, + "eval_Qnli-dev_manhattan_f1_threshold": 407.16436767578125, + "eval_Qnli-dev_manhattan_precision": 0.5885885885885885, + "eval_Qnli-dev_manhattan_recall": 0.8305084745762712, + "eval_Qnli-dev_max_accuracy": 0.671875, + "eval_Qnli-dev_max_accuracy_threshold": 631.760986328125, + "eval_Qnli-dev_max_ap": 0.6704993708527434, + "eval_Qnli-dev_max_f1": 0.6889279437609841, + "eval_Qnli-dev_max_f1_threshold": 596.4973754882812, + "eval_Qnli-dev_max_precision": 0.5885885885885885, + "eval_Qnli-dev_max_recall": 0.8305084745762712, + "eval_allNLI-dev_cosine_accuracy": 0.703125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9278714656829834, + "eval_allNLI-dev_cosine_ap": 0.5360006571581784, + "eval_allNLI-dev_cosine_f1": 0.5560975609756097, + "eval_allNLI-dev_cosine_f1_threshold": 0.8716906905174255, + "eval_allNLI-dev_cosine_precision": 0.4810126582278481, + "eval_allNLI-dev_cosine_recall": 0.6589595375722543, + "eval_allNLI-dev_dot_accuracy": 0.703125, + "eval_allNLI-dev_dot_accuracy_threshold": 710.1931762695312, + "eval_allNLI-dev_dot_ap": 0.5364012669017512, + "eval_allNLI-dev_dot_f1": 0.5563549160671463, + "eval_allNLI-dev_dot_f1_threshold": 664.9071044921875, + "eval_allNLI-dev_dot_precision": 0.47540983606557374, + "eval_allNLI-dev_dot_recall": 0.6705202312138728, + "eval_allNLI-dev_euclidean_accuracy": 0.703125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.507229804992676, + "eval_allNLI-dev_euclidean_ap": 0.5360253625403568, + "eval_allNLI-dev_euclidean_f1": 0.5560975609756097, + "eval_allNLI-dev_euclidean_f1_threshold": 14.014714241027832, + "eval_allNLI-dev_euclidean_precision": 0.4810126582278481, + "eval_allNLI-dev_euclidean_recall": 0.6589595375722543, + "eval_allNLI-dev_manhattan_accuracy": 0.701171875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 230.21136474609375, + "eval_allNLI-dev_manhattan_ap": 0.5368395859693288, + "eval_allNLI-dev_manhattan_f1": 0.5547445255474452, + "eval_allNLI-dev_manhattan_f1_threshold": 309.4046630859375, + "eval_allNLI-dev_manhattan_precision": 0.4789915966386555, + "eval_allNLI-dev_manhattan_recall": 0.6589595375722543, + "eval_allNLI-dev_max_accuracy": 0.703125, + "eval_allNLI-dev_max_accuracy_threshold": 710.1931762695312, + "eval_allNLI-dev_max_ap": 0.5368395859693288, + "eval_allNLI-dev_max_f1": 0.5563549160671463, + "eval_allNLI-dev_max_f1_threshold": 664.9071044921875, + "eval_allNLI-dev_max_precision": 0.4810126582278481, + "eval_allNLI-dev_max_recall": 0.6705202312138728, + "eval_loss": 0.8721151947975159, + "eval_runtime": 50.1849, + "eval_samples_per_second": 33.157, + "eval_sequential_score": 0.6704993708527434, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.6826785247340771, + "eval_sts-test_pearson_dot": 0.6828021136504023, + "eval_sts-test_pearson_euclidean": 0.7015768382234913, + "eval_sts-test_pearson_manhattan": 0.701278706250807, + "eval_sts-test_pearson_max": 0.7015768382234913, + "eval_sts-test_spearman_cosine": 0.6912253878783962, + "eval_sts-test_spearman_dot": 0.6908982114063636, + "eval_sts-test_spearman_euclidean": 0.6912322348213891, + "eval_sts-test_spearman_manhattan": 0.6907121356614981, + "eval_sts-test_spearman_max": 0.6912322348213891, + "step": 1683 + }, + { + "epoch": 1.65748031496063, + "grad_norm": 13.697303771972656, + "learning_rate": 7.545862809008931e-07, + "loss": 0.496, + "step": 1684 + }, + { + "epoch": 1.6584645669291338, + "grad_norm": 10.742180824279785, + "learning_rate": 7.539253847607397e-07, + "loss": 0.358, + "step": 1685 + }, + { + "epoch": 1.659448818897638, + "grad_norm": 10.653914451599121, + "learning_rate": 7.532638915380503e-07, + "loss": 0.3925, + "step": 1686 + }, + { + "epoch": 1.6604330708661417, + "grad_norm": 15.769359588623047, + "learning_rate": 7.526018027985434e-07, + "loss": 0.9602, + "step": 1687 + }, + { + "epoch": 1.6614173228346458, + "grad_norm": 10.051492691040039, + "learning_rate": 7.519391201093462e-07, + "loss": 0.3946, + "step": 1688 + }, + { + "epoch": 1.6624015748031495, + "grad_norm": 14.521549224853516, + "learning_rate": 7.512758450389925e-07, + "loss": 0.8932, + "step": 1689 + }, + { + "epoch": 1.6633858267716537, + "grad_norm": 9.777917861938477, + "learning_rate": 7.50611979157418e-07, + "loss": 0.3817, + "step": 1690 + }, + { + "epoch": 1.6643700787401574, + "grad_norm": 8.56822395324707, + "learning_rate": 7.499475240359567e-07, + "loss": 0.2362, + "step": 1691 + }, + { + "epoch": 1.6653543307086616, + "grad_norm": 8.25054931640625, + "learning_rate": 7.492824812473376e-07, + "loss": 0.2459, + "step": 1692 + }, + { + "epoch": 1.6663385826771653, + "grad_norm": 13.631850242614746, + "learning_rate": 7.4861685236568e-07, + "loss": 0.5824, + "step": 1693 + }, + { + "epoch": 1.6673228346456694, + "grad_norm": 14.922378540039062, + "learning_rate": 7.479506389664914e-07, + "loss": 0.7861, + "step": 1694 + }, + { + "epoch": 1.6683070866141732, + "grad_norm": 14.515337944030762, + "learning_rate": 7.47283842626662e-07, + "loss": 0.625, + "step": 1695 + }, + { + "epoch": 1.6692913385826773, + "grad_norm": 7.668905735015869, + "learning_rate": 7.466164649244625e-07, + "loss": 0.2845, + "step": 1696 + }, + { + "epoch": 1.670275590551181, + "grad_norm": 11.19218921661377, + "learning_rate": 7.45948507439539e-07, + "loss": 0.3401, + "step": 1697 + }, + { + "epoch": 1.6712598425196852, + "grad_norm": 10.618613243103027, + "learning_rate": 7.452799717529102e-07, + "loss": 0.3849, + "step": 1698 + }, + { + "epoch": 1.672244094488189, + "grad_norm": 14.857203483581543, + "learning_rate": 7.446108594469634e-07, + "loss": 0.7924, + "step": 1699 + }, + { + "epoch": 1.673228346456693, + "grad_norm": 9.791996955871582, + "learning_rate": 7.439411721054508e-07, + "loss": 0.2839, + "step": 1700 + }, + { + "epoch": 1.6742125984251968, + "grad_norm": 8.478402137756348, + "learning_rate": 7.432709113134856e-07, + "loss": 0.2974, + "step": 1701 + }, + { + "epoch": 1.675196850393701, + "grad_norm": 9.687871932983398, + "learning_rate": 7.42600078657538e-07, + "loss": 0.3049, + "step": 1702 + }, + { + "epoch": 1.6761811023622046, + "grad_norm": 9.937713623046875, + "learning_rate": 7.419286757254324e-07, + "loss": 0.4319, + "step": 1703 + }, + { + "epoch": 1.6771653543307088, + "grad_norm": 13.37419319152832, + "learning_rate": 7.412567041063424e-07, + "loss": 0.6378, + "step": 1704 + }, + { + "epoch": 1.6781496062992125, + "grad_norm": 13.778456687927246, + "learning_rate": 7.40584165390788e-07, + "loss": 0.6112, + "step": 1705 + }, + { + "epoch": 1.6791338582677167, + "grad_norm": 13.216461181640625, + "learning_rate": 7.399110611706316e-07, + "loss": 0.4503, + "step": 1706 + }, + { + "epoch": 1.6801181102362204, + "grad_norm": 16.38566780090332, + "learning_rate": 7.392373930390738e-07, + "loss": 0.7275, + "step": 1707 + }, + { + "epoch": 1.6811023622047245, + "grad_norm": 13.268678665161133, + "learning_rate": 7.3856316259065e-07, + "loss": 0.498, + "step": 1708 + }, + { + "epoch": 1.6820866141732282, + "grad_norm": 16.93975067138672, + "learning_rate": 7.378883714212268e-07, + "loss": 1.0963, + "step": 1709 + }, + { + "epoch": 1.6830708661417324, + "grad_norm": 13.136117935180664, + "learning_rate": 7.372130211279977e-07, + "loss": 0.5383, + "step": 1710 + }, + { + "epoch": 1.6840551181102361, + "grad_norm": 18.57037353515625, + "learning_rate": 7.365371133094799e-07, + "loss": 0.5373, + "step": 1711 + }, + { + "epoch": 1.6850393700787403, + "grad_norm": 14.358283996582031, + "learning_rate": 7.3586064956551e-07, + "loss": 0.6593, + "step": 1712 + }, + { + "epoch": 1.686023622047244, + "grad_norm": 7.687897682189941, + "learning_rate": 7.351836314972408e-07, + "loss": 0.1884, + "step": 1713 + }, + { + "epoch": 1.6870078740157481, + "grad_norm": 12.682760238647461, + "learning_rate": 7.345060607071365e-07, + "loss": 0.5496, + "step": 1714 + }, + { + "epoch": 1.6879921259842519, + "grad_norm": 11.868288040161133, + "learning_rate": 7.338279387989703e-07, + "loss": 0.5201, + "step": 1715 + }, + { + "epoch": 1.688976377952756, + "grad_norm": 9.233494758605957, + "learning_rate": 7.331492673778193e-07, + "loss": 0.3061, + "step": 1716 + }, + { + "epoch": 1.6899606299212597, + "grad_norm": 12.900493621826172, + "learning_rate": 7.324700480500615e-07, + "loss": 0.534, + "step": 1717 + }, + { + "epoch": 1.6909448818897639, + "grad_norm": 13.235301971435547, + "learning_rate": 7.317902824233719e-07, + "loss": 0.6459, + "step": 1718 + }, + { + "epoch": 1.6919291338582676, + "grad_norm": 15.071056365966797, + "learning_rate": 7.311099721067183e-07, + "loss": 1.1266, + "step": 1719 + }, + { + "epoch": 1.6929133858267718, + "grad_norm": 11.357483863830566, + "learning_rate": 7.304291187103578e-07, + "loss": 0.5472, + "step": 1720 + }, + { + "epoch": 1.6938976377952755, + "grad_norm": 12.90341567993164, + "learning_rate": 7.297477238458333e-07, + "loss": 0.4939, + "step": 1721 + }, + { + "epoch": 1.6948818897637796, + "grad_norm": 13.244677543640137, + "learning_rate": 7.290657891259688e-07, + "loss": 0.6837, + "step": 1722 + }, + { + "epoch": 1.6958661417322833, + "grad_norm": 10.282441139221191, + "learning_rate": 7.283833161648664e-07, + "loss": 0.463, + "step": 1723 + }, + { + "epoch": 1.6968503937007875, + "grad_norm": 7.344255447387695, + "learning_rate": 7.277003065779022e-07, + "loss": 0.2378, + "step": 1724 + }, + { + "epoch": 1.6978346456692912, + "grad_norm": 13.763047218322754, + "learning_rate": 7.270167619817222e-07, + "loss": 0.7928, + "step": 1725 + }, + { + "epoch": 1.6988188976377954, + "grad_norm": 11.851907730102539, + "learning_rate": 7.26332683994239e-07, + "loss": 0.389, + "step": 1726 + }, + { + "epoch": 1.699803149606299, + "grad_norm": 14.237005233764648, + "learning_rate": 7.256480742346278e-07, + "loss": 0.7841, + "step": 1727 + }, + { + "epoch": 1.7007874015748032, + "grad_norm": 9.115606307983398, + "learning_rate": 7.249629343233221e-07, + "loss": 0.3297, + "step": 1728 + }, + { + "epoch": 1.701771653543307, + "grad_norm": 13.66873836517334, + "learning_rate": 7.242772658820105e-07, + "loss": 0.5783, + "step": 1729 + }, + { + "epoch": 1.702755905511811, + "grad_norm": 10.461472511291504, + "learning_rate": 7.235910705336327e-07, + "loss": 0.6329, + "step": 1730 + }, + { + "epoch": 1.7037401574803148, + "grad_norm": 12.690411567687988, + "learning_rate": 7.229043499023753e-07, + "loss": 0.6091, + "step": 1731 + }, + { + "epoch": 1.704724409448819, + "grad_norm": 6.315324306488037, + "learning_rate": 7.222171056136683e-07, + "loss": 0.272, + "step": 1732 + }, + { + "epoch": 1.7057086614173227, + "grad_norm": 13.229377746582031, + "learning_rate": 7.215293392941814e-07, + "loss": 0.8095, + "step": 1733 + }, + { + "epoch": 1.7066929133858268, + "grad_norm": 8.663484573364258, + "learning_rate": 7.208410525718195e-07, + "loss": 0.4847, + "step": 1734 + }, + { + "epoch": 1.7076771653543306, + "grad_norm": 7.8202433586120605, + "learning_rate": 7.201522470757196e-07, + "loss": 0.2548, + "step": 1735 + }, + { + "epoch": 1.7086614173228347, + "grad_norm": 12.125920295715332, + "learning_rate": 7.194629244362462e-07, + "loss": 0.6772, + "step": 1736 + }, + { + "epoch": 1.7096456692913384, + "grad_norm": 14.174428939819336, + "learning_rate": 7.187730862849886e-07, + "loss": 0.6348, + "step": 1737 + }, + { + "epoch": 1.7106299212598426, + "grad_norm": 11.810324668884277, + "learning_rate": 7.180827342547556e-07, + "loss": 0.5227, + "step": 1738 + }, + { + "epoch": 1.7116141732283463, + "grad_norm": 6.058929920196533, + "learning_rate": 7.173918699795726e-07, + "loss": 0.2096, + "step": 1739 + }, + { + "epoch": 1.7125984251968505, + "grad_norm": 10.328184127807617, + "learning_rate": 7.167004950946774e-07, + "loss": 0.4727, + "step": 1740 + }, + { + "epoch": 1.7135826771653542, + "grad_norm": 6.936435699462891, + "learning_rate": 7.160086112365163e-07, + "loss": 0.2488, + "step": 1741 + }, + { + "epoch": 1.7145669291338583, + "grad_norm": 10.65377426147461, + "learning_rate": 7.153162200427404e-07, + "loss": 0.4559, + "step": 1742 + }, + { + "epoch": 1.715551181102362, + "grad_norm": 9.537163734436035, + "learning_rate": 7.146233231522018e-07, + "loss": 0.4249, + "step": 1743 + }, + { + "epoch": 1.7165354330708662, + "grad_norm": 7.669369220733643, + "learning_rate": 7.139299222049493e-07, + "loss": 0.2523, + "step": 1744 + }, + { + "epoch": 1.71751968503937, + "grad_norm": 13.894774436950684, + "learning_rate": 7.132360188422249e-07, + "loss": 0.5707, + "step": 1745 + }, + { + "epoch": 1.718503937007874, + "grad_norm": 14.338887214660645, + "learning_rate": 7.125416147064597e-07, + "loss": 0.8353, + "step": 1746 + }, + { + "epoch": 1.719488188976378, + "grad_norm": 11.37844181060791, + "learning_rate": 7.118467114412702e-07, + "loss": 0.5417, + "step": 1747 + }, + { + "epoch": 1.720472440944882, + "grad_norm": 7.538876056671143, + "learning_rate": 7.111513106914544e-07, + "loss": 0.253, + "step": 1748 + }, + { + "epoch": 1.7214566929133859, + "grad_norm": 15.313557624816895, + "learning_rate": 7.104554141029874e-07, + "loss": 0.609, + "step": 1749 + }, + { + "epoch": 1.7224409448818898, + "grad_norm": 11.29182243347168, + "learning_rate": 7.097590233230186e-07, + "loss": 0.2546, + "step": 1750 + }, + { + "epoch": 1.7234251968503937, + "grad_norm": 15.192030906677246, + "learning_rate": 7.090621399998665e-07, + "loss": 0.8198, + "step": 1751 + }, + { + "epoch": 1.7244094488188977, + "grad_norm": 5.506044864654541, + "learning_rate": 7.083647657830155e-07, + "loss": 0.1099, + "step": 1752 + }, + { + "epoch": 1.7253937007874016, + "grad_norm": 8.552281379699707, + "learning_rate": 7.076669023231126e-07, + "loss": 0.1621, + "step": 1753 + }, + { + "epoch": 1.7263779527559056, + "grad_norm": 7.1006927490234375, + "learning_rate": 7.069685512719619e-07, + "loss": 0.1905, + "step": 1754 + }, + { + "epoch": 1.7273622047244095, + "grad_norm": 12.157682418823242, + "learning_rate": 7.06269714282522e-07, + "loss": 0.5979, + "step": 1755 + }, + { + "epoch": 1.7283464566929134, + "grad_norm": 12.502230644226074, + "learning_rate": 7.055703930089018e-07, + "loss": 0.5411, + "step": 1756 + }, + { + "epoch": 1.7293307086614174, + "grad_norm": 7.2131757736206055, + "learning_rate": 7.048705891063562e-07, + "loss": 0.1829, + "step": 1757 + }, + { + "epoch": 1.7303149606299213, + "grad_norm": 18.08379364013672, + "learning_rate": 7.04170304231283e-07, + "loss": 0.9215, + "step": 1758 + }, + { + "epoch": 1.7312992125984252, + "grad_norm": 12.267788887023926, + "learning_rate": 7.034695400412177e-07, + "loss": 0.3832, + "step": 1759 + }, + { + "epoch": 1.7322834645669292, + "grad_norm": 8.47931957244873, + "learning_rate": 7.027682981948306e-07, + "loss": 0.2209, + "step": 1760 + }, + { + "epoch": 1.733267716535433, + "grad_norm": 11.93918228149414, + "learning_rate": 7.020665803519229e-07, + "loss": 0.483, + "step": 1761 + }, + { + "epoch": 1.734251968503937, + "grad_norm": 16.401823043823242, + "learning_rate": 7.013643881734222e-07, + "loss": 0.6515, + "step": 1762 + }, + { + "epoch": 1.735236220472441, + "grad_norm": 14.30345630645752, + "learning_rate": 7.006617233213787e-07, + "loss": 0.4831, + "step": 1763 + }, + { + "epoch": 1.736220472440945, + "grad_norm": 14.988460540771484, + "learning_rate": 6.999585874589616e-07, + "loss": 0.9117, + "step": 1764 + }, + { + "epoch": 1.7372047244094488, + "grad_norm": 9.57097053527832, + "learning_rate": 6.992549822504548e-07, + "loss": 0.3845, + "step": 1765 + }, + { + "epoch": 1.7381889763779528, + "grad_norm": 7.5431342124938965, + "learning_rate": 6.985509093612534e-07, + "loss": 0.1861, + "step": 1766 + }, + { + "epoch": 1.7391732283464567, + "grad_norm": 16.113262176513672, + "learning_rate": 6.978463704578589e-07, + "loss": 1.1416, + "step": 1767 + }, + { + "epoch": 1.7401574803149606, + "grad_norm": 9.348328590393066, + "learning_rate": 6.971413672078766e-07, + "loss": 0.5021, + "step": 1768 + }, + { + "epoch": 1.7411417322834646, + "grad_norm": 12.102070808410645, + "learning_rate": 6.9643590128001e-07, + "loss": 0.594, + "step": 1769 + }, + { + "epoch": 1.7421259842519685, + "grad_norm": 13.739825248718262, + "learning_rate": 6.957299743440589e-07, + "loss": 0.5797, + "step": 1770 + }, + { + "epoch": 1.7431102362204725, + "grad_norm": 7.500216960906982, + "learning_rate": 6.950235880709131e-07, + "loss": 0.2171, + "step": 1771 + }, + { + "epoch": 1.7440944881889764, + "grad_norm": 12.756651878356934, + "learning_rate": 6.943167441325504e-07, + "loss": 0.4669, + "step": 1772 + }, + { + "epoch": 1.7450787401574803, + "grad_norm": 16.936201095581055, + "learning_rate": 6.936094442020314e-07, + "loss": 0.9011, + "step": 1773 + }, + { + "epoch": 1.7460629921259843, + "grad_norm": 12.409344673156738, + "learning_rate": 6.929016899534963e-07, + "loss": 0.4342, + "step": 1774 + }, + { + "epoch": 1.7470472440944882, + "grad_norm": 17.04608154296875, + "learning_rate": 6.921934830621603e-07, + "loss": 0.746, + "step": 1775 + }, + { + "epoch": 1.7480314960629921, + "grad_norm": 16.811267852783203, + "learning_rate": 6.914848252043106e-07, + "loss": 0.6163, + "step": 1776 + }, + { + "epoch": 1.749015748031496, + "grad_norm": 12.833903312683105, + "learning_rate": 6.907757180573013e-07, + "loss": 0.3791, + "step": 1777 + }, + { + "epoch": 1.75, + "grad_norm": 14.699774742126465, + "learning_rate": 6.9006616329955e-07, + "loss": 0.5592, + "step": 1778 + }, + { + "epoch": 1.750984251968504, + "grad_norm": 11.65503215789795, + "learning_rate": 6.893561626105338e-07, + "loss": 0.49, + "step": 1779 + }, + { + "epoch": 1.7519685039370079, + "grad_norm": 11.93079662322998, + "learning_rate": 6.886457176707855e-07, + "loss": 0.5305, + "step": 1780 + }, + { + "epoch": 1.7529527559055118, + "grad_norm": 22.919675827026367, + "learning_rate": 6.879348301618894e-07, + "loss": 1.2776, + "step": 1781 + }, + { + "epoch": 1.7539370078740157, + "grad_norm": 9.736113548278809, + "learning_rate": 6.872235017664769e-07, + "loss": 0.308, + "step": 1782 + }, + { + "epoch": 1.7549212598425197, + "grad_norm": 11.094430923461914, + "learning_rate": 6.865117341682234e-07, + "loss": 0.332, + "step": 1783 + }, + { + "epoch": 1.7559055118110236, + "grad_norm": 14.460293769836426, + "learning_rate": 6.857995290518439e-07, + "loss": 0.531, + "step": 1784 + }, + { + "epoch": 1.7568897637795275, + "grad_norm": 11.97847843170166, + "learning_rate": 6.850868881030883e-07, + "loss": 0.3992, + "step": 1785 + }, + { + "epoch": 1.7578740157480315, + "grad_norm": 9.501277923583984, + "learning_rate": 6.843738130087393e-07, + "loss": 0.3785, + "step": 1786 + }, + { + "epoch": 1.7588582677165354, + "grad_norm": 10.422222137451172, + "learning_rate": 6.836603054566059e-07, + "loss": 0.6424, + "step": 1787 + }, + { + "epoch": 1.7598425196850394, + "grad_norm": 9.03464412689209, + "learning_rate": 6.829463671355216e-07, + "loss": 0.2961, + "step": 1788 + }, + { + "epoch": 1.7608267716535433, + "grad_norm": 9.770003318786621, + "learning_rate": 6.822319997353391e-07, + "loss": 0.3291, + "step": 1789 + }, + { + "epoch": 1.7618110236220472, + "grad_norm": 9.269059181213379, + "learning_rate": 6.815172049469268e-07, + "loss": 0.3368, + "step": 1790 + }, + { + "epoch": 1.7627952755905512, + "grad_norm": 8.973724365234375, + "learning_rate": 6.808019844621649e-07, + "loss": 0.3546, + "step": 1791 + }, + { + "epoch": 1.763779527559055, + "grad_norm": 12.605461120605469, + "learning_rate": 6.800863399739409e-07, + "loss": 0.4336, + "step": 1792 + }, + { + "epoch": 1.764763779527559, + "grad_norm": 14.863327026367188, + "learning_rate": 6.793702731761459e-07, + "loss": 0.899, + "step": 1793 + }, + { + "epoch": 1.765748031496063, + "grad_norm": 6.504664421081543, + "learning_rate": 6.786537857636707e-07, + "loss": 0.1529, + "step": 1794 + }, + { + "epoch": 1.766732283464567, + "grad_norm": 13.2537841796875, + "learning_rate": 6.77936879432402e-07, + "loss": 0.5044, + "step": 1795 + }, + { + "epoch": 1.7677165354330708, + "grad_norm": 11.545211791992188, + "learning_rate": 6.772195558792175e-07, + "loss": 0.4481, + "step": 1796 + }, + { + "epoch": 1.7687007874015748, + "grad_norm": 10.334686279296875, + "learning_rate": 6.765018168019825e-07, + "loss": 0.5129, + "step": 1797 + }, + { + "epoch": 1.7696850393700787, + "grad_norm": 11.80605411529541, + "learning_rate": 6.757836638995461e-07, + "loss": 0.67, + "step": 1798 + }, + { + "epoch": 1.7706692913385826, + "grad_norm": 11.868819236755371, + "learning_rate": 6.75065098871737e-07, + "loss": 0.5876, + "step": 1799 + }, + { + "epoch": 1.7716535433070866, + "grad_norm": 13.558659553527832, + "learning_rate": 6.74346123419359e-07, + "loss": 0.5667, + "step": 1800 + }, + { + "epoch": 1.7726377952755905, + "grad_norm": 10.476797103881836, + "learning_rate": 6.736267392441874e-07, + "loss": 0.3875, + "step": 1801 + }, + { + "epoch": 1.7736220472440944, + "grad_norm": 10.431046485900879, + "learning_rate": 6.729069480489652e-07, + "loss": 0.3771, + "step": 1802 + }, + { + "epoch": 1.7746062992125984, + "grad_norm": 7.647192001342773, + "learning_rate": 6.721867515373987e-07, + "loss": 0.4124, + "step": 1803 + }, + { + "epoch": 1.7755905511811023, + "grad_norm": 12.491352081298828, + "learning_rate": 6.714661514141535e-07, + "loss": 0.5689, + "step": 1804 + }, + { + "epoch": 1.7765748031496063, + "grad_norm": 11.710945129394531, + "learning_rate": 6.707451493848502e-07, + "loss": 0.634, + "step": 1805 + }, + { + "epoch": 1.7775590551181102, + "grad_norm": 14.7250394821167, + "learning_rate": 6.700237471560614e-07, + "loss": 0.6412, + "step": 1806 + }, + { + "epoch": 1.7785433070866141, + "grad_norm": 15.322542190551758, + "learning_rate": 6.693019464353065e-07, + "loss": 0.852, + "step": 1807 + }, + { + "epoch": 1.779527559055118, + "grad_norm": 10.991961479187012, + "learning_rate": 6.685797489310481e-07, + "loss": 0.4046, + "step": 1808 + }, + { + "epoch": 1.780511811023622, + "grad_norm": 14.14397144317627, + "learning_rate": 6.67857156352688e-07, + "loss": 0.8435, + "step": 1809 + }, + { + "epoch": 1.781496062992126, + "grad_norm": 12.458358764648438, + "learning_rate": 6.671341704105632e-07, + "loss": 0.4458, + "step": 1810 + }, + { + "epoch": 1.78248031496063, + "grad_norm": 11.952537536621094, + "learning_rate": 6.664107928159419e-07, + "loss": 0.4458, + "step": 1811 + }, + { + "epoch": 1.7834645669291338, + "grad_norm": 9.018787384033203, + "learning_rate": 6.656870252810187e-07, + "loss": 0.234, + "step": 1812 + }, + { + "epoch": 1.784448818897638, + "grad_norm": 8.081368446350098, + "learning_rate": 6.649628695189123e-07, + "loss": 0.2844, + "step": 1813 + }, + { + "epoch": 1.7854330708661417, + "grad_norm": 7.770649433135986, + "learning_rate": 6.642383272436589e-07, + "loss": 0.4315, + "step": 1814 + }, + { + "epoch": 1.7864173228346458, + "grad_norm": 12.0542573928833, + "learning_rate": 6.635134001702107e-07, + "loss": 0.6663, + "step": 1815 + }, + { + "epoch": 1.7874015748031495, + "grad_norm": 13.215538024902344, + "learning_rate": 6.627880900144303e-07, + "loss": 0.7126, + "step": 1816 + }, + { + "epoch": 1.7883858267716537, + "grad_norm": 7.426309108734131, + "learning_rate": 6.620623984930869e-07, + "loss": 0.2318, + "step": 1817 + }, + { + "epoch": 1.7893700787401574, + "grad_norm": 9.918778419494629, + "learning_rate": 6.613363273238527e-07, + "loss": 0.3844, + "step": 1818 + }, + { + "epoch": 1.7903543307086616, + "grad_norm": 7.417040824890137, + "learning_rate": 6.606098782252978e-07, + "loss": 0.2853, + "step": 1819 + }, + { + "epoch": 1.7913385826771653, + "grad_norm": 9.51761531829834, + "learning_rate": 6.598830529168877e-07, + "loss": 0.3053, + "step": 1820 + }, + { + "epoch": 1.7923228346456694, + "grad_norm": 12.628561019897461, + "learning_rate": 6.591558531189781e-07, + "loss": 0.6893, + "step": 1821 + }, + { + "epoch": 1.7933070866141732, + "grad_norm": 10.970913887023926, + "learning_rate": 6.584282805528106e-07, + "loss": 0.3448, + "step": 1822 + }, + { + "epoch": 1.7942913385826773, + "grad_norm": 11.655643463134766, + "learning_rate": 6.577003369405099e-07, + "loss": 0.4703, + "step": 1823 + }, + { + "epoch": 1.795275590551181, + "grad_norm": 7.324519157409668, + "learning_rate": 6.569720240050784e-07, + "loss": 0.2449, + "step": 1824 + }, + { + "epoch": 1.7962598425196852, + "grad_norm": 11.098455429077148, + "learning_rate": 6.562433434703928e-07, + "loss": 0.4558, + "step": 1825 + }, + { + "epoch": 1.797244094488189, + "grad_norm": 3.6354446411132812, + "learning_rate": 6.555142970612001e-07, + "loss": 0.0693, + "step": 1826 + }, + { + "epoch": 1.798228346456693, + "grad_norm": 11.416316032409668, + "learning_rate": 6.54784886503113e-07, + "loss": 0.4178, + "step": 1827 + }, + { + "epoch": 1.7992125984251968, + "grad_norm": 11.04200267791748, + "learning_rate": 6.540551135226061e-07, + "loss": 0.3354, + "step": 1828 + }, + { + "epoch": 1.800196850393701, + "grad_norm": 14.586440086364746, + "learning_rate": 6.533249798470123e-07, + "loss": 0.6727, + "step": 1829 + }, + { + "epoch": 1.8011811023622046, + "grad_norm": 14.171919822692871, + "learning_rate": 6.525944872045175e-07, + "loss": 0.4957, + "step": 1830 + }, + { + "epoch": 1.8021653543307088, + "grad_norm": 14.926817893981934, + "learning_rate": 6.51863637324158e-07, + "loss": 0.5577, + "step": 1831 + }, + { + "epoch": 1.8031496062992125, + "grad_norm": 12.324542045593262, + "learning_rate": 6.511324319358154e-07, + "loss": 0.3628, + "step": 1832 + }, + { + "epoch": 1.8041338582677167, + "grad_norm": 8.232148170471191, + "learning_rate": 6.504008727702123e-07, + "loss": 0.1764, + "step": 1833 + }, + { + "epoch": 1.8051181102362204, + "grad_norm": 8.707999229431152, + "learning_rate": 6.496689615589096e-07, + "loss": 0.2427, + "step": 1834 + }, + { + "epoch": 1.8061023622047245, + "grad_norm": 13.219265937805176, + "learning_rate": 6.489367000343005e-07, + "loss": 0.3703, + "step": 1835 + }, + { + "epoch": 1.8070866141732282, + "grad_norm": 9.929430961608887, + "learning_rate": 6.482040899296081e-07, + "loss": 0.2839, + "step": 1836 + }, + { + "epoch": 1.8070866141732282, + "eval_Qnli-dev_cosine_accuracy": 0.697265625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.775582492351532, + "eval_Qnli-dev_cosine_ap": 0.688738787235431, + "eval_Qnli-dev_cosine_f1": 0.6951456310679611, + "eval_Qnli-dev_cosine_f1_threshold": 0.7566580772399902, + "eval_Qnli-dev_cosine_precision": 0.6415770609318996, + "eval_Qnli-dev_cosine_recall": 0.7584745762711864, + "eval_Qnli-dev_dot_accuracy": 0.697265625, + "eval_Qnli-dev_dot_accuracy_threshold": 593.4371337890625, + "eval_Qnli-dev_dot_ap": 0.6890728187486499, + "eval_Qnli-dev_dot_f1": 0.6951456310679611, + "eval_Qnli-dev_dot_f1_threshold": 578.9700317382812, + "eval_Qnli-dev_dot_precision": 0.6415770609318996, + "eval_Qnli-dev_dot_recall": 0.7584745762711864, + "eval_Qnli-dev_euclidean_accuracy": 0.697265625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 18.531177520751953, + "eval_Qnli-dev_euclidean_ap": 0.6886470121465607, + "eval_Qnli-dev_euclidean_f1": 0.6951456310679611, + "eval_Qnli-dev_euclidean_f1_threshold": 19.3006591796875, + "eval_Qnli-dev_euclidean_precision": 0.6415770609318996, + "eval_Qnli-dev_euclidean_recall": 0.7584745762711864, + "eval_Qnli-dev_manhattan_accuracy": 0.705078125, + "eval_Qnli-dev_manhattan_accuracy_threshold": 392.95538330078125, + "eval_Qnli-dev_manhattan_ap": 0.690347526905352, + "eval_Qnli-dev_manhattan_f1": 0.6912621359223301, + "eval_Qnli-dev_manhattan_f1_threshold": 425.84490966796875, + "eval_Qnli-dev_manhattan_precision": 0.6379928315412187, + "eval_Qnli-dev_manhattan_recall": 0.7542372881355932, + "eval_Qnli-dev_max_accuracy": 0.705078125, + "eval_Qnli-dev_max_accuracy_threshold": 593.4371337890625, + "eval_Qnli-dev_max_ap": 0.690347526905352, + "eval_Qnli-dev_max_f1": 0.6951456310679611, + "eval_Qnli-dev_max_f1_threshold": 578.9700317382812, + "eval_Qnli-dev_max_precision": 0.6415770609318996, + "eval_Qnli-dev_max_recall": 0.7584745762711864, + "eval_allNLI-dev_cosine_accuracy": 0.6953125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9281036257743835, + "eval_allNLI-dev_cosine_ap": 0.5317153845037348, + "eval_allNLI-dev_cosine_f1": 0.577319587628866, + "eval_allNLI-dev_cosine_f1_threshold": 0.8576892614364624, + "eval_allNLI-dev_cosine_precision": 0.5209302325581395, + "eval_allNLI-dev_cosine_recall": 0.6473988439306358, + "eval_allNLI-dev_dot_accuracy": 0.6953125, + "eval_allNLI-dev_dot_accuracy_threshold": 710.734375, + "eval_allNLI-dev_dot_ap": 0.5325389254628002, + "eval_allNLI-dev_dot_f1": 0.5788113695090439, + "eval_allNLI-dev_dot_f1_threshold": 656.3677368164062, + "eval_allNLI-dev_dot_precision": 0.5233644859813084, + "eval_allNLI-dev_dot_recall": 0.6473988439306358, + "eval_allNLI-dev_euclidean_accuracy": 0.6953125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.493587493896484, + "eval_allNLI-dev_euclidean_ap": 0.5319987489812823, + "eval_allNLI-dev_euclidean_f1": 0.577319587628866, + "eval_allNLI-dev_euclidean_f1_threshold": 14.756336212158203, + "eval_allNLI-dev_euclidean_precision": 0.5209302325581395, + "eval_allNLI-dev_euclidean_recall": 0.6473988439306358, + "eval_allNLI-dev_manhattan_accuracy": 0.693359375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 231.1400146484375, + "eval_allNLI-dev_manhattan_ap": 0.5333786665126131, + "eval_allNLI-dev_manhattan_f1": 0.5749385749385749, + "eval_allNLI-dev_manhattan_f1_threshold": 334.93621826171875, + "eval_allNLI-dev_manhattan_precision": 0.5, + "eval_allNLI-dev_manhattan_recall": 0.6763005780346821, + "eval_allNLI-dev_max_accuracy": 0.6953125, + "eval_allNLI-dev_max_accuracy_threshold": 710.734375, + "eval_allNLI-dev_max_ap": 0.5333786665126131, + "eval_allNLI-dev_max_f1": 0.5788113695090439, + "eval_allNLI-dev_max_f1_threshold": 656.3677368164062, + "eval_allNLI-dev_max_precision": 0.5233644859813084, + "eval_allNLI-dev_max_recall": 0.6763005780346821, + "eval_loss": 0.8638383150100708, + "eval_runtime": 50.5746, + "eval_samples_per_second": 32.902, + "eval_sequential_score": 0.690347526905352, + "eval_steps_per_second": 0.138, + "eval_sts-test_pearson_cosine": 0.6744927184612423, + "eval_sts-test_pearson_dot": 0.6743365606265737, + "eval_sts-test_pearson_euclidean": 0.6937012781730647, + "eval_sts-test_pearson_manhattan": 0.6939985907312147, + "eval_sts-test_pearson_max": 0.6939985907312147, + "eval_sts-test_spearman_cosine": 0.6827461426262607, + "eval_sts-test_spearman_dot": 0.682627149022221, + "eval_sts-test_spearman_euclidean": 0.6828339446011107, + "eval_sts-test_spearman_manhattan": 0.6836835472986263, + "eval_sts-test_spearman_max": 0.6836835472986263, + "step": 1836 + }, + { + "epoch": 1.8080708661417324, + "grad_norm": 10.339896202087402, + "learning_rate": 6.474711329788802e-07, + "loss": 0.3651, + "step": 1837 + }, + { + "epoch": 1.8090551181102361, + "grad_norm": 12.040072441101074, + "learning_rate": 6.467378309169858e-07, + "loss": 0.3415, + "step": 1838 + }, + { + "epoch": 1.8100393700787403, + "grad_norm": 10.304866790771484, + "learning_rate": 6.460041854796105e-07, + "loss": 0.312, + "step": 1839 + }, + { + "epoch": 1.811023622047244, + "grad_norm": 7.809697151184082, + "learning_rate": 6.452701984032526e-07, + "loss": 0.2208, + "step": 1840 + }, + { + "epoch": 1.8120078740157481, + "grad_norm": 9.799861907958984, + "learning_rate": 6.445358714252195e-07, + "loss": 0.4022, + "step": 1841 + }, + { + "epoch": 1.8129921259842519, + "grad_norm": 15.075982093811035, + "learning_rate": 6.438012062836226e-07, + "loss": 0.8372, + "step": 1842 + }, + { + "epoch": 1.813976377952756, + "grad_norm": 10.279742240905762, + "learning_rate": 6.430662047173742e-07, + "loss": 0.2395, + "step": 1843 + }, + { + "epoch": 1.8149606299212597, + "grad_norm": 8.526269912719727, + "learning_rate": 6.423308684661825e-07, + "loss": 0.2016, + "step": 1844 + }, + { + "epoch": 1.8159448818897639, + "grad_norm": 6.898560047149658, + "learning_rate": 6.415951992705482e-07, + "loss": 0.1323, + "step": 1845 + }, + { + "epoch": 1.8169291338582676, + "grad_norm": 13.116087913513184, + "learning_rate": 6.408591988717594e-07, + "loss": 0.4582, + "step": 1846 + }, + { + "epoch": 1.8179133858267718, + "grad_norm": 11.618614196777344, + "learning_rate": 6.401228690118893e-07, + "loss": 0.3896, + "step": 1847 + }, + { + "epoch": 1.8188976377952755, + "grad_norm": 12.383076667785645, + "learning_rate": 6.3938621143379e-07, + "loss": 0.4897, + "step": 1848 + }, + { + "epoch": 1.8198818897637796, + "grad_norm": 8.645105361938477, + "learning_rate": 6.386492278810894e-07, + "loss": 0.1942, + "step": 1849 + }, + { + "epoch": 1.8208661417322833, + "grad_norm": 7.458933353424072, + "learning_rate": 6.37911920098187e-07, + "loss": 0.1479, + "step": 1850 + }, + { + "epoch": 1.8218503937007875, + "grad_norm": 12.923534393310547, + "learning_rate": 6.371742898302503e-07, + "loss": 0.4026, + "step": 1851 + }, + { + "epoch": 1.8228346456692912, + "grad_norm": 11.39023208618164, + "learning_rate": 6.364363388232091e-07, + "loss": 0.4361, + "step": 1852 + }, + { + "epoch": 1.8238188976377954, + "grad_norm": 12.149271011352539, + "learning_rate": 6.356980688237535e-07, + "loss": 0.3709, + "step": 1853 + }, + { + "epoch": 1.824803149606299, + "grad_norm": 7.508931636810303, + "learning_rate": 6.349594815793275e-07, + "loss": 0.1513, + "step": 1854 + }, + { + "epoch": 1.8257874015748032, + "grad_norm": 16.406944274902344, + "learning_rate": 6.34220578838127e-07, + "loss": 1.1054, + "step": 1855 + }, + { + "epoch": 1.826771653543307, + "grad_norm": 10.299270629882812, + "learning_rate": 6.334813623490939e-07, + "loss": 0.3779, + "step": 1856 + }, + { + "epoch": 1.827755905511811, + "grad_norm": 10.28286361694336, + "learning_rate": 6.327418338619132e-07, + "loss": 0.3918, + "step": 1857 + }, + { + "epoch": 1.8287401574803148, + "grad_norm": 17.7500057220459, + "learning_rate": 6.320019951270082e-07, + "loss": 0.8661, + "step": 1858 + }, + { + "epoch": 1.829724409448819, + "grad_norm": 15.081620216369629, + "learning_rate": 6.312618478955368e-07, + "loss": 0.5442, + "step": 1859 + }, + { + "epoch": 1.8307086614173227, + "grad_norm": 9.177066802978516, + "learning_rate": 6.305213939193864e-07, + "loss": 0.2266, + "step": 1860 + }, + { + "epoch": 1.8316929133858268, + "grad_norm": 11.408110618591309, + "learning_rate": 6.297806349511711e-07, + "loss": 0.2871, + "step": 1861 + }, + { + "epoch": 1.8326771653543306, + "grad_norm": 13.180274963378906, + "learning_rate": 6.290395727442269e-07, + "loss": 0.6227, + "step": 1862 + }, + { + "epoch": 1.8336614173228347, + "grad_norm": 12.962007522583008, + "learning_rate": 6.282982090526071e-07, + "loss": 0.5195, + "step": 1863 + }, + { + "epoch": 1.8346456692913384, + "grad_norm": 8.186065673828125, + "learning_rate": 6.27556545631079e-07, + "loss": 0.1537, + "step": 1864 + }, + { + "epoch": 1.8356299212598426, + "grad_norm": 13.374753952026367, + "learning_rate": 6.268145842351191e-07, + "loss": 0.4054, + "step": 1865 + }, + { + "epoch": 1.8366141732283463, + "grad_norm": 18.404796600341797, + "learning_rate": 6.260723266209094e-07, + "loss": 1.0866, + "step": 1866 + }, + { + "epoch": 1.8375984251968505, + "grad_norm": 9.623383522033691, + "learning_rate": 6.253297745453329e-07, + "loss": 0.3088, + "step": 1867 + }, + { + "epoch": 1.8385826771653542, + "grad_norm": 15.913948059082031, + "learning_rate": 6.245869297659694e-07, + "loss": 0.8792, + "step": 1868 + }, + { + "epoch": 1.8395669291338583, + "grad_norm": 8.207590103149414, + "learning_rate": 6.238437940410921e-07, + "loss": 0.2253, + "step": 1869 + }, + { + "epoch": 1.840551181102362, + "grad_norm": 9.02070426940918, + "learning_rate": 6.231003691296622e-07, + "loss": 0.2422, + "step": 1870 + }, + { + "epoch": 1.8415354330708662, + "grad_norm": 9.788644790649414, + "learning_rate": 6.223566567913256e-07, + "loss": 0.3587, + "step": 1871 + }, + { + "epoch": 1.84251968503937, + "grad_norm": 15.468562126159668, + "learning_rate": 6.216126587864084e-07, + "loss": 1.0157, + "step": 1872 + }, + { + "epoch": 1.843503937007874, + "grad_norm": 13.664466857910156, + "learning_rate": 6.208683768759136e-07, + "loss": 0.5268, + "step": 1873 + }, + { + "epoch": 1.844488188976378, + "grad_norm": 12.072665214538574, + "learning_rate": 6.201238128215152e-07, + "loss": 0.3945, + "step": 1874 + }, + { + "epoch": 1.845472440944882, + "grad_norm": 6.9167070388793945, + "learning_rate": 6.193789683855555e-07, + "loss": 0.1593, + "step": 1875 + }, + { + "epoch": 1.8464566929133859, + "grad_norm": 13.289469718933105, + "learning_rate": 6.186338453310404e-07, + "loss": 0.4943, + "step": 1876 + }, + { + "epoch": 1.8474409448818898, + "grad_norm": 14.335954666137695, + "learning_rate": 6.178884454216355e-07, + "loss": 0.5563, + "step": 1877 + }, + { + "epoch": 1.8484251968503937, + "grad_norm": 12.79931640625, + "learning_rate": 6.171427704216609e-07, + "loss": 0.674, + "step": 1878 + }, + { + "epoch": 1.8494094488188977, + "grad_norm": 13.337133407592773, + "learning_rate": 6.163968220960891e-07, + "loss": 0.5021, + "step": 1879 + }, + { + "epoch": 1.8503937007874016, + "grad_norm": 10.655539512634277, + "learning_rate": 6.156506022105382e-07, + "loss": 0.256, + "step": 1880 + }, + { + "epoch": 1.8513779527559056, + "grad_norm": 13.640843391418457, + "learning_rate": 6.1490411253127e-07, + "loss": 0.4215, + "step": 1881 + }, + { + "epoch": 1.8523622047244095, + "grad_norm": 14.875307083129883, + "learning_rate": 6.141573548251847e-07, + "loss": 0.4112, + "step": 1882 + }, + { + "epoch": 1.8533464566929134, + "grad_norm": 13.641471862792969, + "learning_rate": 6.134103308598166e-07, + "loss": 0.7106, + "step": 1883 + }, + { + "epoch": 1.8543307086614174, + "grad_norm": 12.023168563842773, + "learning_rate": 6.126630424033303e-07, + "loss": 0.2984, + "step": 1884 + }, + { + "epoch": 1.8553149606299213, + "grad_norm": 17.169565200805664, + "learning_rate": 6.119154912245167e-07, + "loss": 0.7731, + "step": 1885 + }, + { + "epoch": 1.8562992125984252, + "grad_norm": 11.882272720336914, + "learning_rate": 6.111676790927884e-07, + "loss": 0.3873, + "step": 1886 + }, + { + "epoch": 1.8572834645669292, + "grad_norm": 13.995811462402344, + "learning_rate": 6.104196077781757e-07, + "loss": 0.5798, + "step": 1887 + }, + { + "epoch": 1.858267716535433, + "grad_norm": 6.9644646644592285, + "learning_rate": 6.096712790513222e-07, + "loss": 0.1436, + "step": 1888 + }, + { + "epoch": 1.859251968503937, + "grad_norm": 10.488465309143066, + "learning_rate": 6.08922694683481e-07, + "loss": 0.2872, + "step": 1889 + }, + { + "epoch": 1.860236220472441, + "grad_norm": 15.655674934387207, + "learning_rate": 6.081738564465103e-07, + "loss": 1.0389, + "step": 1890 + }, + { + "epoch": 1.861220472440945, + "grad_norm": 12.279995918273926, + "learning_rate": 6.07424766112869e-07, + "loss": 0.4683, + "step": 1891 + }, + { + "epoch": 1.8622047244094488, + "grad_norm": 12.133166313171387, + "learning_rate": 6.066754254556127e-07, + "loss": 0.3775, + "step": 1892 + }, + { + "epoch": 1.8631889763779528, + "grad_norm": 10.459905624389648, + "learning_rate": 6.059258362483897e-07, + "loss": 0.374, + "step": 1893 + }, + { + "epoch": 1.8641732283464567, + "grad_norm": 7.63214635848999, + "learning_rate": 6.051760002654364e-07, + "loss": 0.183, + "step": 1894 + }, + { + "epoch": 1.8651574803149606, + "grad_norm": 12.797852516174316, + "learning_rate": 6.044259192815735e-07, + "loss": 0.4589, + "step": 1895 + }, + { + "epoch": 1.8661417322834646, + "grad_norm": 15.862738609313965, + "learning_rate": 6.036755950722014e-07, + "loss": 0.5661, + "step": 1896 + }, + { + "epoch": 1.8671259842519685, + "grad_norm": 10.23412799835205, + "learning_rate": 6.029250294132965e-07, + "loss": 0.3746, + "step": 1897 + }, + { + "epoch": 1.8681102362204725, + "grad_norm": 11.018589973449707, + "learning_rate": 6.021742240814067e-07, + "loss": 0.3988, + "step": 1898 + }, + { + "epoch": 1.8690944881889764, + "grad_norm": 9.83062744140625, + "learning_rate": 6.014231808536463e-07, + "loss": 0.3589, + "step": 1899 + }, + { + "epoch": 1.8700787401574803, + "grad_norm": 8.018756866455078, + "learning_rate": 6.006719015076938e-07, + "loss": 0.3373, + "step": 1900 + }, + { + "epoch": 1.8710629921259843, + "grad_norm": 10.258496284484863, + "learning_rate": 5.999203878217861e-07, + "loss": 0.3848, + "step": 1901 + }, + { + "epoch": 1.8720472440944882, + "grad_norm": 9.544556617736816, + "learning_rate": 5.99168641574715e-07, + "loss": 0.2908, + "step": 1902 + }, + { + "epoch": 1.8730314960629921, + "grad_norm": 11.74750804901123, + "learning_rate": 5.984166645458224e-07, + "loss": 0.511, + "step": 1903 + }, + { + "epoch": 1.874015748031496, + "grad_norm": 7.742142200469971, + "learning_rate": 5.976644585149967e-07, + "loss": 0.2879, + "step": 1904 + }, + { + "epoch": 1.875, + "grad_norm": 7.75278377532959, + "learning_rate": 5.969120252626684e-07, + "loss": 0.213, + "step": 1905 + }, + { + "epoch": 1.875984251968504, + "grad_norm": 12.225879669189453, + "learning_rate": 5.961593665698055e-07, + "loss": 0.4193, + "step": 1906 + }, + { + "epoch": 1.8769685039370079, + "grad_norm": 10.169327735900879, + "learning_rate": 5.954064842179099e-07, + "loss": 0.357, + "step": 1907 + }, + { + "epoch": 1.8779527559055118, + "grad_norm": 13.49083423614502, + "learning_rate": 5.946533799890127e-07, + "loss": 0.5527, + "step": 1908 + }, + { + "epoch": 1.8789370078740157, + "grad_norm": 14.641011238098145, + "learning_rate": 5.939000556656704e-07, + "loss": 0.7, + "step": 1909 + }, + { + "epoch": 1.8799212598425197, + "grad_norm": 7.070704936981201, + "learning_rate": 5.931465130309604e-07, + "loss": 0.1586, + "step": 1910 + }, + { + "epoch": 1.8809055118110236, + "grad_norm": 13.22642707824707, + "learning_rate": 5.923927538684764e-07, + "loss": 0.4659, + "step": 1911 + }, + { + "epoch": 1.8818897637795275, + "grad_norm": 6.765905857086182, + "learning_rate": 5.916387799623253e-07, + "loss": 0.1581, + "step": 1912 + }, + { + "epoch": 1.8828740157480315, + "grad_norm": 11.993915557861328, + "learning_rate": 5.908845930971217e-07, + "loss": 0.4552, + "step": 1913 + }, + { + "epoch": 1.8838582677165354, + "grad_norm": 11.672133445739746, + "learning_rate": 5.901301950579845e-07, + "loss": 0.4731, + "step": 1914 + }, + { + "epoch": 1.8848425196850394, + "grad_norm": 13.491580963134766, + "learning_rate": 5.893755876305328e-07, + "loss": 0.7688, + "step": 1915 + }, + { + "epoch": 1.8858267716535433, + "grad_norm": 13.972790718078613, + "learning_rate": 5.886207726008805e-07, + "loss": 0.4149, + "step": 1916 + }, + { + "epoch": 1.8868110236220472, + "grad_norm": 16.476404190063477, + "learning_rate": 5.878657517556336e-07, + "loss": 0.6552, + "step": 1917 + }, + { + "epoch": 1.8877952755905512, + "grad_norm": 13.266674041748047, + "learning_rate": 5.871105268818848e-07, + "loss": 0.3994, + "step": 1918 + }, + { + "epoch": 1.888779527559055, + "grad_norm": 9.897150039672852, + "learning_rate": 5.863550997672101e-07, + "loss": 0.2387, + "step": 1919 + }, + { + "epoch": 1.889763779527559, + "grad_norm": 10.708202362060547, + "learning_rate": 5.855994721996638e-07, + "loss": 0.4503, + "step": 1920 + }, + { + "epoch": 1.890748031496063, + "grad_norm": 15.079380989074707, + "learning_rate": 5.848436459677752e-07, + "loss": 0.8661, + "step": 1921 + }, + { + "epoch": 1.891732283464567, + "grad_norm": 8.452141761779785, + "learning_rate": 5.840876228605432e-07, + "loss": 0.1919, + "step": 1922 + }, + { + "epoch": 1.8927165354330708, + "grad_norm": 8.290151596069336, + "learning_rate": 5.833314046674329e-07, + "loss": 0.2893, + "step": 1923 + }, + { + "epoch": 1.8937007874015748, + "grad_norm": 10.83450698852539, + "learning_rate": 5.825749931783715e-07, + "loss": 0.4982, + "step": 1924 + }, + { + "epoch": 1.8946850393700787, + "grad_norm": 12.19471549987793, + "learning_rate": 5.818183901837435e-07, + "loss": 0.5492, + "step": 1925 + }, + { + "epoch": 1.8956692913385826, + "grad_norm": 10.35559368133545, + "learning_rate": 5.810615974743864e-07, + "loss": 0.4138, + "step": 1926 + }, + { + "epoch": 1.8966535433070866, + "grad_norm": 11.441749572753906, + "learning_rate": 5.803046168415872e-07, + "loss": 0.5311, + "step": 1927 + }, + { + "epoch": 1.8976377952755905, + "grad_norm": 12.828340530395508, + "learning_rate": 5.795474500770772e-07, + "loss": 0.7912, + "step": 1928 + }, + { + "epoch": 1.8986220472440944, + "grad_norm": 14.586247444152832, + "learning_rate": 5.787900989730291e-07, + "loss": 0.8013, + "step": 1929 + }, + { + "epoch": 1.8996062992125984, + "grad_norm": 9.356870651245117, + "learning_rate": 5.780325653220509e-07, + "loss": 0.3011, + "step": 1930 + }, + { + "epoch": 1.9005905511811023, + "grad_norm": 3.8600447177886963, + "learning_rate": 5.772748509171833e-07, + "loss": 0.0675, + "step": 1931 + }, + { + "epoch": 1.9015748031496063, + "grad_norm": 10.75484848022461, + "learning_rate": 5.765169575518948e-07, + "loss": 0.4471, + "step": 1932 + }, + { + "epoch": 1.9025590551181102, + "grad_norm": 6.319433689117432, + "learning_rate": 5.757588870200772e-07, + "loss": 0.1335, + "step": 1933 + }, + { + "epoch": 1.9035433070866141, + "grad_norm": 9.709519386291504, + "learning_rate": 5.750006411160421e-07, + "loss": 0.4238, + "step": 1934 + }, + { + "epoch": 1.904527559055118, + "grad_norm": 14.068886756896973, + "learning_rate": 5.742422216345157e-07, + "loss": 0.4881, + "step": 1935 + }, + { + "epoch": 1.905511811023622, + "grad_norm": 13.283921241760254, + "learning_rate": 5.734836303706352e-07, + "loss": 0.677, + "step": 1936 + }, + { + "epoch": 1.906496062992126, + "grad_norm": 11.908031463623047, + "learning_rate": 5.727248691199446e-07, + "loss": 0.4711, + "step": 1937 + }, + { + "epoch": 1.90748031496063, + "grad_norm": 6.094554424285889, + "learning_rate": 5.719659396783902e-07, + "loss": 0.1324, + "step": 1938 + }, + { + "epoch": 1.9084645669291338, + "grad_norm": 7.862908363342285, + "learning_rate": 5.712068438423161e-07, + "loss": 0.3842, + "step": 1939 + }, + { + "epoch": 1.909448818897638, + "grad_norm": 11.08634090423584, + "learning_rate": 5.704475834084608e-07, + "loss": 0.4307, + "step": 1940 + }, + { + "epoch": 1.9104330708661417, + "grad_norm": 7.747585296630859, + "learning_rate": 5.696881601739517e-07, + "loss": 0.2825, + "step": 1941 + }, + { + "epoch": 1.9114173228346458, + "grad_norm": 10.382781982421875, + "learning_rate": 5.689285759363022e-07, + "loss": 0.269, + "step": 1942 + }, + { + "epoch": 1.9124015748031495, + "grad_norm": 12.933969497680664, + "learning_rate": 5.681688324934064e-07, + "loss": 0.5368, + "step": 1943 + }, + { + "epoch": 1.9133858267716537, + "grad_norm": 13.612123489379883, + "learning_rate": 5.674089316435353e-07, + "loss": 0.5181, + "step": 1944 + }, + { + "epoch": 1.9143700787401574, + "grad_norm": 11.407240867614746, + "learning_rate": 5.666488751853321e-07, + "loss": 0.2569, + "step": 1945 + }, + { + "epoch": 1.9153543307086616, + "grad_norm": 13.681182861328125, + "learning_rate": 5.658886649178096e-07, + "loss": 0.5495, + "step": 1946 + }, + { + "epoch": 1.9163385826771653, + "grad_norm": 10.186968803405762, + "learning_rate": 5.651283026403427e-07, + "loss": 0.2403, + "step": 1947 + }, + { + "epoch": 1.9173228346456694, + "grad_norm": 16.009565353393555, + "learning_rate": 5.64367790152668e-07, + "loss": 0.5904, + "step": 1948 + }, + { + "epoch": 1.9183070866141732, + "grad_norm": 14.810846328735352, + "learning_rate": 5.636071292548766e-07, + "loss": 0.841, + "step": 1949 + }, + { + "epoch": 1.9192913385826773, + "grad_norm": 12.214828491210938, + "learning_rate": 5.628463217474108e-07, + "loss": 0.5092, + "step": 1950 + }, + { + "epoch": 1.920275590551181, + "grad_norm": 13.162872314453125, + "learning_rate": 5.620853694310604e-07, + "loss": 0.5407, + "step": 1951 + }, + { + "epoch": 1.9212598425196852, + "grad_norm": 9.507124900817871, + "learning_rate": 5.613242741069578e-07, + "loss": 0.447, + "step": 1952 + }, + { + "epoch": 1.922244094488189, + "grad_norm": 10.588579177856445, + "learning_rate": 5.605630375765739e-07, + "loss": 0.3349, + "step": 1953 + }, + { + "epoch": 1.923228346456693, + "grad_norm": 10.573832511901855, + "learning_rate": 5.598016616417136e-07, + "loss": 0.3061, + "step": 1954 + }, + { + "epoch": 1.9242125984251968, + "grad_norm": 11.5529203414917, + "learning_rate": 5.590401481045118e-07, + "loss": 0.4172, + "step": 1955 + }, + { + "epoch": 1.925196850393701, + "grad_norm": 17.73041343688965, + "learning_rate": 5.582784987674296e-07, + "loss": 0.9333, + "step": 1956 + }, + { + "epoch": 1.9261811023622046, + "grad_norm": 6.100178241729736, + "learning_rate": 5.575167154332489e-07, + "loss": 0.1431, + "step": 1957 + }, + { + "epoch": 1.9271653543307088, + "grad_norm": 18.10417366027832, + "learning_rate": 5.567547999050692e-07, + "loss": 1.287, + "step": 1958 + }, + { + "epoch": 1.9281496062992125, + "grad_norm": 9.070444107055664, + "learning_rate": 5.559927539863024e-07, + "loss": 0.4819, + "step": 1959 + }, + { + "epoch": 1.9291338582677167, + "grad_norm": 9.977766990661621, + "learning_rate": 5.552305794806699e-07, + "loss": 0.4332, + "step": 1960 + }, + { + "epoch": 1.9301181102362204, + "grad_norm": 14.155847549438477, + "learning_rate": 5.544682781921964e-07, + "loss": 0.5793, + "step": 1961 + }, + { + "epoch": 1.9311023622047245, + "grad_norm": 7.486117362976074, + "learning_rate": 5.537058519252076e-07, + "loss": 0.1754, + "step": 1962 + }, + { + "epoch": 1.9320866141732282, + "grad_norm": 9.620716094970703, + "learning_rate": 5.529433024843244e-07, + "loss": 0.3553, + "step": 1963 + }, + { + "epoch": 1.9330708661417324, + "grad_norm": 8.661392211914062, + "learning_rate": 5.521806316744594e-07, + "loss": 0.3444, + "step": 1964 + }, + { + "epoch": 1.9340551181102361, + "grad_norm": 11.182474136352539, + "learning_rate": 5.514178413008125e-07, + "loss": 0.5331, + "step": 1965 + }, + { + "epoch": 1.9350393700787403, + "grad_norm": 5.605703353881836, + "learning_rate": 5.506549331688669e-07, + "loss": 0.1572, + "step": 1966 + }, + { + "epoch": 1.936023622047244, + "grad_norm": 11.598262786865234, + "learning_rate": 5.49891909084384e-07, + "loss": 0.3608, + "step": 1967 + }, + { + "epoch": 1.9370078740157481, + "grad_norm": 8.981062889099121, + "learning_rate": 5.491287708534e-07, + "loss": 0.2908, + "step": 1968 + }, + { + "epoch": 1.9379921259842519, + "grad_norm": 13.642875671386719, + "learning_rate": 5.483655202822214e-07, + "loss": 0.6125, + "step": 1969 + }, + { + "epoch": 1.938976377952756, + "grad_norm": 12.759238243103027, + "learning_rate": 5.4760215917742e-07, + "loss": 0.258, + "step": 1970 + }, + { + "epoch": 1.9399606299212597, + "grad_norm": 11.04817008972168, + "learning_rate": 5.4683868934583e-07, + "loss": 0.5867, + "step": 1971 + }, + { + "epoch": 1.9409448818897639, + "grad_norm": 9.991081237792969, + "learning_rate": 5.460751125945424e-07, + "loss": 0.4119, + "step": 1972 + }, + { + "epoch": 1.9419291338582676, + "grad_norm": 14.327850341796875, + "learning_rate": 5.453114307309015e-07, + "loss": 0.4027, + "step": 1973 + }, + { + "epoch": 1.9429133858267718, + "grad_norm": 7.812295436859131, + "learning_rate": 5.445476455625002e-07, + "loss": 0.2298, + "step": 1974 + }, + { + "epoch": 1.9438976377952755, + "grad_norm": 8.687687873840332, + "learning_rate": 5.437837588971761e-07, + "loss": 0.3679, + "step": 1975 + }, + { + "epoch": 1.9448818897637796, + "grad_norm": 12.897151947021484, + "learning_rate": 5.430197725430071e-07, + "loss": 0.3998, + "step": 1976 + }, + { + "epoch": 1.9458661417322833, + "grad_norm": 9.91887092590332, + "learning_rate": 5.422556883083069e-07, + "loss": 0.2195, + "step": 1977 + }, + { + "epoch": 1.9468503937007875, + "grad_norm": 8.012516021728516, + "learning_rate": 5.414915080016209e-07, + "loss": 0.1621, + "step": 1978 + }, + { + "epoch": 1.9478346456692912, + "grad_norm": 13.478154182434082, + "learning_rate": 5.407272334317221e-07, + "loss": 0.5747, + "step": 1979 + }, + { + "epoch": 1.9488188976377954, + "grad_norm": 18.646350860595703, + "learning_rate": 5.399628664076062e-07, + "loss": 1.0614, + "step": 1980 + }, + { + "epoch": 1.949803149606299, + "grad_norm": 9.263113021850586, + "learning_rate": 5.391984087384882e-07, + "loss": 0.3013, + "step": 1981 + }, + { + "epoch": 1.9507874015748032, + "grad_norm": 12.43802261352539, + "learning_rate": 5.384338622337973e-07, + "loss": 0.3753, + "step": 1982 + }, + { + "epoch": 1.951771653543307, + "grad_norm": 8.433222770690918, + "learning_rate": 5.37669228703173e-07, + "loss": 0.1918, + "step": 1983 + }, + { + "epoch": 1.952755905511811, + "grad_norm": 11.056380271911621, + "learning_rate": 5.369045099564613e-07, + "loss": 0.2864, + "step": 1984 + }, + { + "epoch": 1.9537401574803148, + "grad_norm": 7.701892375946045, + "learning_rate": 5.361397078037089e-07, + "loss": 0.1933, + "step": 1985 + }, + { + "epoch": 1.954724409448819, + "grad_norm": 9.080361366271973, + "learning_rate": 5.353748240551612e-07, + "loss": 0.1951, + "step": 1986 + }, + { + "epoch": 1.9557086614173227, + "grad_norm": 10.842450141906738, + "learning_rate": 5.346098605212553e-07, + "loss": 0.6032, + "step": 1987 + }, + { + "epoch": 1.9566929133858268, + "grad_norm": 17.33802604675293, + "learning_rate": 5.338448190126184e-07, + "loss": 0.9978, + "step": 1988 + }, + { + "epoch": 1.9576771653543306, + "grad_norm": 12.883622169494629, + "learning_rate": 5.330797013400614e-07, + "loss": 0.5123, + "step": 1989 + }, + { + "epoch": 1.9576771653543306, + "eval_Qnli-dev_cosine_accuracy": 0.701171875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8180251121520996, + "eval_Qnli-dev_cosine_ap": 0.7143839210403344, + "eval_Qnli-dev_cosine_f1": 0.6892655367231638, + "eval_Qnli-dev_cosine_f1_threshold": 0.7824835181236267, + "eval_Qnli-dev_cosine_precision": 0.6203389830508474, + "eval_Qnli-dev_cosine_recall": 0.7754237288135594, + "eval_Qnli-dev_dot_accuracy": 0.69921875, + "eval_Qnli-dev_dot_accuracy_threshold": 625.5216064453125, + "eval_Qnli-dev_dot_ap": 0.7147411844337869, + "eval_Qnli-dev_dot_f1": 0.690566037735849, + "eval_Qnli-dev_dot_f1_threshold": 598.4419555664062, + "eval_Qnli-dev_dot_precision": 0.6224489795918368, + "eval_Qnli-dev_dot_recall": 0.7754237288135594, + "eval_Qnli-dev_euclidean_accuracy": 0.701171875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 16.680295944213867, + "eval_Qnli-dev_euclidean_ap": 0.7145208759421537, + "eval_Qnli-dev_euclidean_f1": 0.6892655367231638, + "eval_Qnli-dev_euclidean_f1_threshold": 18.235549926757812, + "eval_Qnli-dev_euclidean_precision": 0.6203389830508474, + "eval_Qnli-dev_euclidean_recall": 0.7754237288135594, + "eval_Qnli-dev_manhattan_accuracy": 0.69921875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 369.07568359375, + "eval_Qnli-dev_manhattan_ap": 0.7158568897840752, + "eval_Qnli-dev_manhattan_f1": 0.6934865900383143, + "eval_Qnli-dev_manhattan_f1_threshold": 400.7629699707031, + "eval_Qnli-dev_manhattan_precision": 0.6328671328671329, + "eval_Qnli-dev_manhattan_recall": 0.7669491525423728, + "eval_Qnli-dev_max_accuracy": 0.701171875, + "eval_Qnli-dev_max_accuracy_threshold": 625.5216064453125, + "eval_Qnli-dev_max_ap": 0.7158568897840752, + "eval_Qnli-dev_max_f1": 0.6934865900383143, + "eval_Qnli-dev_max_f1_threshold": 598.4419555664062, + "eval_Qnli-dev_max_precision": 0.6328671328671329, + "eval_Qnli-dev_max_recall": 0.7754237288135594, + "eval_allNLI-dev_cosine_accuracy": 0.68359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9415509104728699, + "eval_allNLI-dev_cosine_ap": 0.5172239749490007, + "eval_allNLI-dev_cosine_f1": 0.5598377281947262, + "eval_allNLI-dev_cosine_f1_threshold": 0.8371573686599731, + "eval_allNLI-dev_cosine_precision": 0.43125, + "eval_allNLI-dev_cosine_recall": 0.7976878612716763, + "eval_allNLI-dev_dot_accuracy": 0.685546875, + "eval_allNLI-dev_dot_accuracy_threshold": 720.062744140625, + "eval_allNLI-dev_dot_ap": 0.5173952870938315, + "eval_allNLI-dev_dot_f1": 0.5591836734693878, + "eval_allNLI-dev_dot_f1_threshold": 640.5137939453125, + "eval_allNLI-dev_dot_precision": 0.43217665615141954, + "eval_allNLI-dev_dot_recall": 0.791907514450867, + "eval_allNLI-dev_euclidean_accuracy": 0.68359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.455118179321289, + "eval_allNLI-dev_euclidean_ap": 0.5173366676727166, + "eval_allNLI-dev_euclidean_f1": 0.5598377281947262, + "eval_allNLI-dev_euclidean_f1_threshold": 15.775871276855469, + "eval_allNLI-dev_euclidean_precision": 0.43125, + "eval_allNLI-dev_euclidean_recall": 0.7976878612716763, + "eval_allNLI-dev_manhattan_accuracy": 0.68359375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 209.49310302734375, + "eval_allNLI-dev_manhattan_ap": 0.5162658773602664, + "eval_allNLI-dev_manhattan_f1": 0.55863539445629, + "eval_allNLI-dev_manhattan_f1_threshold": 338.36199951171875, + "eval_allNLI-dev_manhattan_precision": 0.44256756756756754, + "eval_allNLI-dev_manhattan_recall": 0.7572254335260116, + "eval_allNLI-dev_max_accuracy": 0.685546875, + "eval_allNLI-dev_max_accuracy_threshold": 720.062744140625, + "eval_allNLI-dev_max_ap": 0.5173952870938315, + "eval_allNLI-dev_max_f1": 0.5598377281947262, + "eval_allNLI-dev_max_f1_threshold": 640.5137939453125, + "eval_allNLI-dev_max_precision": 0.44256756756756754, + "eval_allNLI-dev_max_recall": 0.7976878612716763, + "eval_loss": 0.7938111424446106, + "eval_runtime": 50.2082, + "eval_samples_per_second": 33.142, + "eval_sequential_score": 0.7158568897840752, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.6960379723167603, + "eval_sts-test_pearson_dot": 0.6957512228793926, + "eval_sts-test_pearson_euclidean": 0.7188787382645466, + "eval_sts-test_pearson_manhattan": 0.7182504864871246, + "eval_sts-test_pearson_max": 0.7188787382645466, + "eval_sts-test_spearman_cosine": 0.7094393750210027, + "eval_sts-test_spearman_dot": 0.7088395291133124, + "eval_sts-test_spearman_euclidean": 0.709505383131555, + "eval_sts-test_spearman_manhattan": 0.708816392711173, + "eval_sts-test_spearman_max": 0.709505383131555, + "step": 1989 + }, + { + "epoch": 1.9586614173228347, + "grad_norm": 12.853811264038086, + "learning_rate": 5.323145093145763e-07, + "loss": 0.5045, + "step": 1990 + }, + { + "epoch": 1.9596456692913384, + "grad_norm": 7.313170433044434, + "learning_rate": 5.3154924474733e-07, + "loss": 0.2195, + "step": 1991 + }, + { + "epoch": 1.9606299212598426, + "grad_norm": 11.035358428955078, + "learning_rate": 5.307839094496621e-07, + "loss": 0.4691, + "step": 1992 + }, + { + "epoch": 1.9616141732283463, + "grad_norm": 14.967765808105469, + "learning_rate": 5.300185052330793e-07, + "loss": 0.5583, + "step": 1993 + }, + { + "epoch": 1.9625984251968505, + "grad_norm": 10.805540084838867, + "learning_rate": 5.292530339092513e-07, + "loss": 0.396, + "step": 1994 + }, + { + "epoch": 1.9635826771653542, + "grad_norm": 12.279142379760742, + "learning_rate": 5.284874972900064e-07, + "loss": 0.5822, + "step": 1995 + }, + { + "epoch": 1.9645669291338583, + "grad_norm": 8.870162963867188, + "learning_rate": 5.277218971873281e-07, + "loss": 0.304, + "step": 1996 + }, + { + "epoch": 1.965551181102362, + "grad_norm": 12.810161590576172, + "learning_rate": 5.269562354133496e-07, + "loss": 0.4864, + "step": 1997 + }, + { + "epoch": 1.9665354330708662, + "grad_norm": 16.406719207763672, + "learning_rate": 5.261905137803504e-07, + "loss": 0.7415, + "step": 1998 + }, + { + "epoch": 1.96751968503937, + "grad_norm": 8.030189514160156, + "learning_rate": 5.254247341007514e-07, + "loss": 0.3624, + "step": 1999 + }, + { + "epoch": 1.968503937007874, + "grad_norm": 12.63553237915039, + "learning_rate": 5.246588981871113e-07, + "loss": 0.5553, + "step": 2000 + }, + { + "epoch": 1.969488188976378, + "grad_norm": 10.115273475646973, + "learning_rate": 5.238930078521215e-07, + "loss": 0.403, + "step": 2001 + }, + { + "epoch": 1.970472440944882, + "grad_norm": 10.957775115966797, + "learning_rate": 5.231270649086023e-07, + "loss": 0.3852, + "step": 2002 + }, + { + "epoch": 1.9714566929133859, + "grad_norm": 10.635632514953613, + "learning_rate": 5.22361071169499e-07, + "loss": 0.426, + "step": 2003 + }, + { + "epoch": 1.9724409448818898, + "grad_norm": 12.00613784790039, + "learning_rate": 5.215950284478763e-07, + "loss": 0.5756, + "step": 2004 + }, + { + "epoch": 1.9734251968503937, + "grad_norm": 10.560711860656738, + "learning_rate": 5.208289385569154e-07, + "loss": 0.5362, + "step": 2005 + }, + { + "epoch": 1.9744094488188977, + "grad_norm": 9.652769088745117, + "learning_rate": 5.20062803309909e-07, + "loss": 0.3039, + "step": 2006 + }, + { + "epoch": 1.9753937007874016, + "grad_norm": 10.32618236541748, + "learning_rate": 5.192966245202571e-07, + "loss": 0.5428, + "step": 2007 + }, + { + "epoch": 1.9763779527559056, + "grad_norm": 8.049816131591797, + "learning_rate": 5.18530404001463e-07, + "loss": 0.1809, + "step": 2008 + }, + { + "epoch": 1.9773622047244095, + "grad_norm": 7.6905717849731445, + "learning_rate": 5.177641435671284e-07, + "loss": 0.4187, + "step": 2009 + }, + { + "epoch": 1.9783464566929134, + "grad_norm": 9.481074333190918, + "learning_rate": 5.169978450309498e-07, + "loss": 0.3272, + "step": 2010 + }, + { + "epoch": 1.9793307086614174, + "grad_norm": 11.813081741333008, + "learning_rate": 5.162315102067139e-07, + "loss": 0.7026, + "step": 2011 + }, + { + "epoch": 1.9803149606299213, + "grad_norm": 12.26184368133545, + "learning_rate": 5.154651409082927e-07, + "loss": 0.4655, + "step": 2012 + }, + { + "epoch": 1.9812992125984252, + "grad_norm": 9.2816801071167, + "learning_rate": 5.146987389496408e-07, + "loss": 0.3722, + "step": 2013 + }, + { + "epoch": 1.9822834645669292, + "grad_norm": 13.438984870910645, + "learning_rate": 5.139323061447892e-07, + "loss": 0.6931, + "step": 2014 + }, + { + "epoch": 1.983267716535433, + "grad_norm": 8.03709602355957, + "learning_rate": 5.131658443078426e-07, + "loss": 0.2296, + "step": 2015 + }, + { + "epoch": 1.984251968503937, + "grad_norm": 13.03205680847168, + "learning_rate": 5.123993552529736e-07, + "loss": 0.4416, + "step": 2016 + }, + { + "epoch": 1.985236220472441, + "grad_norm": 5.718056678771973, + "learning_rate": 5.116328407944201e-07, + "loss": 0.212, + "step": 2017 + }, + { + "epoch": 1.986220472440945, + "grad_norm": 16.513364791870117, + "learning_rate": 5.108663027464795e-07, + "loss": 0.8206, + "step": 2018 + }, + { + "epoch": 1.9872047244094488, + "grad_norm": 9.781304359436035, + "learning_rate": 5.100997429235055e-07, + "loss": 0.4023, + "step": 2019 + }, + { + "epoch": 1.9881889763779528, + "grad_norm": 12.696980476379395, + "learning_rate": 5.093331631399029e-07, + "loss": 0.6757, + "step": 2020 + }, + { + "epoch": 1.9891732283464567, + "grad_norm": 11.146982192993164, + "learning_rate": 5.085665652101244e-07, + "loss": 0.4231, + "step": 2021 + }, + { + "epoch": 1.9901574803149606, + "grad_norm": 13.091096878051758, + "learning_rate": 5.077999509486645e-07, + "loss": 0.5269, + "step": 2022 + }, + { + "epoch": 1.9911417322834646, + "grad_norm": 9.668133735656738, + "learning_rate": 5.070333221700575e-07, + "loss": 0.4245, + "step": 2023 + }, + { + "epoch": 1.9921259842519685, + "grad_norm": 8.33875560760498, + "learning_rate": 5.062666806888715e-07, + "loss": 0.1659, + "step": 2024 + }, + { + "epoch": 1.9931102362204725, + "grad_norm": 8.097585678100586, + "learning_rate": 5.055000283197051e-07, + "loss": 0.2194, + "step": 2025 + }, + { + "epoch": 1.9940944881889764, + "grad_norm": 17.49336051940918, + "learning_rate": 5.047333668771821e-07, + "loss": 0.7817, + "step": 2026 + }, + { + "epoch": 1.9950787401574803, + "grad_norm": 7.3635783195495605, + "learning_rate": 5.03966698175948e-07, + "loss": 0.1798, + "step": 2027 + }, + { + "epoch": 1.9960629921259843, + "grad_norm": 8.882747650146484, + "learning_rate": 5.032000240306655e-07, + "loss": 0.2914, + "step": 2028 + }, + { + "epoch": 1.9970472440944882, + "grad_norm": 6.346508502960205, + "learning_rate": 5.024333462560104e-07, + "loss": 0.2195, + "step": 2029 + }, + { + "epoch": 1.9980314960629921, + "grad_norm": 14.176541328430176, + "learning_rate": 5.016666666666668e-07, + "loss": 0.7062, + "step": 2030 + }, + { + "epoch": 1.999015748031496, + "grad_norm": 6.6057586669921875, + "learning_rate": 5.008999870773232e-07, + "loss": 0.3015, + "step": 2031 + }, + { + "epoch": 2.0, + "grad_norm": 4.027645587921143, + "learning_rate": 5.001333093026681e-07, + "loss": 0.0439, + "step": 2032 + }, + { + "epoch": 2.000984251968504, + "grad_norm": 13.882719039916992, + "learning_rate": 4.993666351573856e-07, + "loss": 0.4779, + "step": 2033 + }, + { + "epoch": 2.001968503937008, + "grad_norm": 17.731901168823242, + "learning_rate": 4.985999664561514e-07, + "loss": 0.9751, + "step": 2034 + }, + { + "epoch": 2.002952755905512, + "grad_norm": 6.450457572937012, + "learning_rate": 4.978333050136285e-07, + "loss": 0.1586, + "step": 2035 + }, + { + "epoch": 2.0039370078740157, + "grad_norm": 9.63836669921875, + "learning_rate": 4.970666526444619e-07, + "loss": 0.3471, + "step": 2036 + }, + { + "epoch": 2.00492125984252, + "grad_norm": 10.052136421203613, + "learning_rate": 4.96300011163276e-07, + "loss": 0.2625, + "step": 2037 + }, + { + "epoch": 2.0059055118110236, + "grad_norm": 7.936495304107666, + "learning_rate": 4.955333823846691e-07, + "loss": 0.2341, + "step": 2038 + }, + { + "epoch": 2.0068897637795278, + "grad_norm": 12.090235710144043, + "learning_rate": 4.947667681232094e-07, + "loss": 0.4475, + "step": 2039 + }, + { + "epoch": 2.0078740157480315, + "grad_norm": 13.9876070022583, + "learning_rate": 4.940001701934305e-07, + "loss": 0.4256, + "step": 2040 + }, + { + "epoch": 2.0088582677165356, + "grad_norm": 8.177584648132324, + "learning_rate": 4.932335904098282e-07, + "loss": 0.2214, + "step": 2041 + }, + { + "epoch": 2.0098425196850394, + "grad_norm": 13.374650001525879, + "learning_rate": 4.924670305868541e-07, + "loss": 0.4418, + "step": 2042 + }, + { + "epoch": 2.0108267716535435, + "grad_norm": 10.596186637878418, + "learning_rate": 4.917004925389137e-07, + "loss": 0.5155, + "step": 2043 + }, + { + "epoch": 2.0118110236220472, + "grad_norm": 9.714990615844727, + "learning_rate": 4.909339780803601e-07, + "loss": 0.2737, + "step": 2044 + }, + { + "epoch": 2.0127952755905514, + "grad_norm": 13.049437522888184, + "learning_rate": 4.901674890254911e-07, + "loss": 0.5087, + "step": 2045 + }, + { + "epoch": 2.013779527559055, + "grad_norm": 13.00695514678955, + "learning_rate": 4.894010271885444e-07, + "loss": 0.6449, + "step": 2046 + }, + { + "epoch": 2.0147637795275593, + "grad_norm": 11.481451988220215, + "learning_rate": 4.886345943836928e-07, + "loss": 0.2826, + "step": 2047 + }, + { + "epoch": 2.015748031496063, + "grad_norm": 8.891685485839844, + "learning_rate": 4.878681924250408e-07, + "loss": 0.2836, + "step": 2048 + }, + { + "epoch": 2.016732283464567, + "grad_norm": 9.778858184814453, + "learning_rate": 4.871018231266197e-07, + "loss": 0.3512, + "step": 2049 + }, + { + "epoch": 2.017716535433071, + "grad_norm": 11.535679817199707, + "learning_rate": 4.863354883023838e-07, + "loss": 0.5771, + "step": 2050 + }, + { + "epoch": 2.018700787401575, + "grad_norm": 12.165321350097656, + "learning_rate": 4.85569189766205e-07, + "loss": 0.4603, + "step": 2051 + }, + { + "epoch": 2.0196850393700787, + "grad_norm": 11.94356632232666, + "learning_rate": 4.848029293318705e-07, + "loss": 0.4311, + "step": 2052 + }, + { + "epoch": 2.020669291338583, + "grad_norm": 7.932766437530518, + "learning_rate": 4.840367088130764e-07, + "loss": 0.1954, + "step": 2053 + }, + { + "epoch": 2.0216535433070866, + "grad_norm": 11.730668067932129, + "learning_rate": 4.832705300234245e-07, + "loss": 0.4358, + "step": 2054 + }, + { + "epoch": 2.0226377952755907, + "grad_norm": 13.260199546813965, + "learning_rate": 4.825043947764181e-07, + "loss": 0.3992, + "step": 2055 + }, + { + "epoch": 2.0236220472440944, + "grad_norm": 10.06352710723877, + "learning_rate": 4.817383048854572e-07, + "loss": 0.3307, + "step": 2056 + }, + { + "epoch": 2.0246062992125986, + "grad_norm": 11.39328670501709, + "learning_rate": 4.809722621638345e-07, + "loss": 0.5604, + "step": 2057 + }, + { + "epoch": 2.0255905511811023, + "grad_norm": 11.53843879699707, + "learning_rate": 4.802062684247311e-07, + "loss": 0.3903, + "step": 2058 + }, + { + "epoch": 2.0265748031496065, + "grad_norm": 12.264870643615723, + "learning_rate": 4.79440325481212e-07, + "loss": 0.3189, + "step": 2059 + }, + { + "epoch": 2.02755905511811, + "grad_norm": 9.427846908569336, + "learning_rate": 4.786744351462222e-07, + "loss": 0.3665, + "step": 2060 + }, + { + "epoch": 2.0285433070866143, + "grad_norm": 12.669951438903809, + "learning_rate": 4.779085992325821e-07, + "loss": 0.4483, + "step": 2061 + }, + { + "epoch": 2.029527559055118, + "grad_norm": 15.742377281188965, + "learning_rate": 4.771428195529833e-07, + "loss": 0.5755, + "step": 2062 + }, + { + "epoch": 2.030511811023622, + "grad_norm": 13.098060607910156, + "learning_rate": 4.763770979199841e-07, + "loss": 0.4937, + "step": 2063 + }, + { + "epoch": 2.031496062992126, + "grad_norm": 12.316033363342285, + "learning_rate": 4.756114361460057e-07, + "loss": 0.5062, + "step": 2064 + }, + { + "epoch": 2.03248031496063, + "grad_norm": 10.440582275390625, + "learning_rate": 4.748458360433273e-07, + "loss": 0.377, + "step": 2065 + }, + { + "epoch": 2.033464566929134, + "grad_norm": 7.60792350769043, + "learning_rate": 4.7408029942408244e-07, + "loss": 0.2601, + "step": 2066 + }, + { + "epoch": 2.034448818897638, + "grad_norm": 12.326177597045898, + "learning_rate": 4.733148281002543e-07, + "loss": 0.5072, + "step": 2067 + }, + { + "epoch": 2.0354330708661417, + "grad_norm": 13.335978507995605, + "learning_rate": 4.7254942388367147e-07, + "loss": 0.4048, + "step": 2068 + }, + { + "epoch": 2.036417322834646, + "grad_norm": 17.156761169433594, + "learning_rate": 4.717840885860036e-07, + "loss": 0.8814, + "step": 2069 + }, + { + "epoch": 2.0374015748031495, + "grad_norm": 12.331279754638672, + "learning_rate": 4.7101882401875745e-07, + "loss": 0.5141, + "step": 2070 + }, + { + "epoch": 2.0383858267716537, + "grad_norm": 15.29952621459961, + "learning_rate": 4.7025363199327213e-07, + "loss": 0.6425, + "step": 2071 + }, + { + "epoch": 2.0393700787401574, + "grad_norm": 15.998370170593262, + "learning_rate": 4.694885143207153e-07, + "loss": 0.9713, + "step": 2072 + }, + { + "epoch": 2.0403543307086616, + "grad_norm": 8.79700756072998, + "learning_rate": 4.687234728120783e-07, + "loss": 0.3059, + "step": 2073 + }, + { + "epoch": 2.0413385826771653, + "grad_norm": 7.250189781188965, + "learning_rate": 4.6795850927817256e-07, + "loss": 0.1983, + "step": 2074 + }, + { + "epoch": 2.0423228346456694, + "grad_norm": 12.074629783630371, + "learning_rate": 4.6719362552962457e-07, + "loss": 0.3746, + "step": 2075 + }, + { + "epoch": 2.043307086614173, + "grad_norm": 9.357890129089355, + "learning_rate": 4.664288233768724e-07, + "loss": 0.3395, + "step": 2076 + }, + { + "epoch": 2.0442913385826773, + "grad_norm": 7.769244194030762, + "learning_rate": 4.6566410463016044e-07, + "loss": 0.2468, + "step": 2077 + }, + { + "epoch": 2.045275590551181, + "grad_norm": 13.112343788146973, + "learning_rate": 4.648994710995363e-07, + "loss": 0.6995, + "step": 2078 + }, + { + "epoch": 2.046259842519685, + "grad_norm": 10.218881607055664, + "learning_rate": 4.641349245948454e-07, + "loss": 0.3861, + "step": 2079 + }, + { + "epoch": 2.047244094488189, + "grad_norm": 12.14520263671875, + "learning_rate": 4.633704669257273e-07, + "loss": 0.4106, + "step": 2080 + }, + { + "epoch": 2.048228346456693, + "grad_norm": 8.871252059936523, + "learning_rate": 4.626060999016115e-07, + "loss": 0.2958, + "step": 2081 + }, + { + "epoch": 2.0492125984251968, + "grad_norm": 4.858777046203613, + "learning_rate": 4.6184182533171257e-07, + "loss": 0.1139, + "step": 2082 + }, + { + "epoch": 2.050196850393701, + "grad_norm": 10.624565124511719, + "learning_rate": 4.6107764502502684e-07, + "loss": 0.2553, + "step": 2083 + }, + { + "epoch": 2.0511811023622046, + "grad_norm": 9.890420913696289, + "learning_rate": 4.603135607903266e-07, + "loss": 0.2428, + "step": 2084 + }, + { + "epoch": 2.052165354330709, + "grad_norm": 8.641292572021484, + "learning_rate": 4.5954957443615765e-07, + "loss": 0.4406, + "step": 2085 + }, + { + "epoch": 2.0531496062992125, + "grad_norm": 7.16730260848999, + "learning_rate": 4.5878568777083356e-07, + "loss": 0.1801, + "step": 2086 + }, + { + "epoch": 2.0541338582677167, + "grad_norm": 11.3385009765625, + "learning_rate": 4.5802190260243226e-07, + "loss": 0.4199, + "step": 2087 + }, + { + "epoch": 2.0551181102362204, + "grad_norm": 8.774247169494629, + "learning_rate": 4.572582207387912e-07, + "loss": 0.2363, + "step": 2088 + }, + { + "epoch": 2.0561023622047245, + "grad_norm": 10.837448120117188, + "learning_rate": 4.5649464398750356e-07, + "loss": 0.348, + "step": 2089 + }, + { + "epoch": 2.0570866141732282, + "grad_norm": 11.618494033813477, + "learning_rate": 4.557311741559135e-07, + "loss": 0.3253, + "step": 2090 + }, + { + "epoch": 2.0580708661417324, + "grad_norm": 11.252474784851074, + "learning_rate": 4.5496781305111215e-07, + "loss": 0.3904, + "step": 2091 + }, + { + "epoch": 2.059055118110236, + "grad_norm": 7.2509050369262695, + "learning_rate": 4.5420456247993354e-07, + "loss": 0.2205, + "step": 2092 + }, + { + "epoch": 2.0600393700787403, + "grad_norm": 13.61648178100586, + "learning_rate": 4.5344142424894954e-07, + "loss": 0.5558, + "step": 2093 + }, + { + "epoch": 2.061023622047244, + "grad_norm": 10.435012817382812, + "learning_rate": 4.526784001644667e-07, + "loss": 0.4491, + "step": 2094 + }, + { + "epoch": 2.062007874015748, + "grad_norm": 3.372648239135742, + "learning_rate": 4.5191549203252103e-07, + "loss": 0.0676, + "step": 2095 + }, + { + "epoch": 2.062992125984252, + "grad_norm": 9.130961418151855, + "learning_rate": 4.511527016588742e-07, + "loss": 0.3404, + "step": 2096 + }, + { + "epoch": 2.063976377952756, + "grad_norm": 9.293866157531738, + "learning_rate": 4.503900308490092e-07, + "loss": 0.2066, + "step": 2097 + }, + { + "epoch": 2.0649606299212597, + "grad_norm": 8.999009132385254, + "learning_rate": 4.496274814081259e-07, + "loss": 0.2436, + "step": 2098 + }, + { + "epoch": 2.065944881889764, + "grad_norm": 6.857397556304932, + "learning_rate": 4.4886505514113695e-07, + "loss": 0.152, + "step": 2099 + }, + { + "epoch": 2.0669291338582676, + "grad_norm": 16.2700252532959, + "learning_rate": 4.481027538526636e-07, + "loss": 0.734, + "step": 2100 + }, + { + "epoch": 2.0679133858267718, + "grad_norm": 11.695120811462402, + "learning_rate": 4.4734057934703095e-07, + "loss": 0.588, + "step": 2101 + }, + { + "epoch": 2.0688976377952755, + "grad_norm": 7.291047096252441, + "learning_rate": 4.4657853342826435e-07, + "loss": 0.2979, + "step": 2102 + }, + { + "epoch": 2.0698818897637796, + "grad_norm": 14.575162887573242, + "learning_rate": 4.458166179000848e-07, + "loss": 0.6343, + "step": 2103 + }, + { + "epoch": 2.0708661417322833, + "grad_norm": 17.440893173217773, + "learning_rate": 4.4505483456590413e-07, + "loss": 0.7228, + "step": 2104 + }, + { + "epoch": 2.0718503937007875, + "grad_norm": 11.222655296325684, + "learning_rate": 4.442931852288219e-07, + "loss": 0.2904, + "step": 2105 + }, + { + "epoch": 2.072834645669291, + "grad_norm": 1.9647241830825806, + "learning_rate": 4.435316716916202e-07, + "loss": 0.081, + "step": 2106 + }, + { + "epoch": 2.0738188976377954, + "grad_norm": 6.077908992767334, + "learning_rate": 4.4277029575675983e-07, + "loss": 0.1277, + "step": 2107 + }, + { + "epoch": 2.074803149606299, + "grad_norm": 11.017847061157227, + "learning_rate": 4.4200905922637575e-07, + "loss": 0.3885, + "step": 2108 + }, + { + "epoch": 2.0757874015748032, + "grad_norm": 9.708911895751953, + "learning_rate": 4.4124796390227313e-07, + "loss": 0.4551, + "step": 2109 + }, + { + "epoch": 2.076771653543307, + "grad_norm": 13.412023544311523, + "learning_rate": 4.4048701158592277e-07, + "loss": 0.5163, + "step": 2110 + }, + { + "epoch": 2.077755905511811, + "grad_norm": 9.517277717590332, + "learning_rate": 4.3972620407845714e-07, + "loss": 0.2665, + "step": 2111 + }, + { + "epoch": 2.078740157480315, + "grad_norm": 10.492469787597656, + "learning_rate": 4.3896554318066556e-07, + "loss": 0.2557, + "step": 2112 + }, + { + "epoch": 2.079724409448819, + "grad_norm": 10.706164360046387, + "learning_rate": 4.3820503069299077e-07, + "loss": 0.3867, + "step": 2113 + }, + { + "epoch": 2.0807086614173227, + "grad_norm": 15.469551086425781, + "learning_rate": 4.3744466841552416e-07, + "loss": 0.473, + "step": 2114 + }, + { + "epoch": 2.081692913385827, + "grad_norm": 13.868034362792969, + "learning_rate": 4.3668445814800137e-07, + "loss": 0.4647, + "step": 2115 + }, + { + "epoch": 2.0826771653543306, + "grad_norm": 8.555710792541504, + "learning_rate": 4.359244016897984e-07, + "loss": 0.185, + "step": 2116 + }, + { + "epoch": 2.0836614173228347, + "grad_norm": 13.925908088684082, + "learning_rate": 4.3516450083992714e-07, + "loss": 0.7512, + "step": 2117 + }, + { + "epoch": 2.0846456692913384, + "grad_norm": 13.838081359863281, + "learning_rate": 4.344047573970313e-07, + "loss": 0.8137, + "step": 2118 + }, + { + "epoch": 2.0856299212598426, + "grad_norm": 20.284046173095703, + "learning_rate": 4.336451731593817e-07, + "loss": 1.091, + "step": 2119 + }, + { + "epoch": 2.0866141732283463, + "grad_norm": 11.106941223144531, + "learning_rate": 4.328857499248727e-07, + "loss": 0.3861, + "step": 2120 + }, + { + "epoch": 2.0875984251968505, + "grad_norm": 9.837322235107422, + "learning_rate": 4.321264894910173e-07, + "loss": 0.4264, + "step": 2121 + }, + { + "epoch": 2.088582677165354, + "grad_norm": 12.149086952209473, + "learning_rate": 4.3136739365494334e-07, + "loss": 0.6489, + "step": 2122 + }, + { + "epoch": 2.0895669291338583, + "grad_norm": 9.458745956420898, + "learning_rate": 4.306084642133889e-07, + "loss": 0.2108, + "step": 2123 + }, + { + "epoch": 2.090551181102362, + "grad_norm": 13.98378849029541, + "learning_rate": 4.2984970296269856e-07, + "loss": 0.5081, + "step": 2124 + }, + { + "epoch": 2.091535433070866, + "grad_norm": 11.618062019348145, + "learning_rate": 4.290911116988181e-07, + "loss": 0.6273, + "step": 2125 + }, + { + "epoch": 2.09251968503937, + "grad_norm": 12.753189086914062, + "learning_rate": 4.2833269221729153e-07, + "loss": 0.3319, + "step": 2126 + }, + { + "epoch": 2.093503937007874, + "grad_norm": 11.565346717834473, + "learning_rate": 4.275744463132564e-07, + "loss": 0.3813, + "step": 2127 + }, + { + "epoch": 2.094488188976378, + "grad_norm": 13.388854026794434, + "learning_rate": 4.268163757814388e-07, + "loss": 0.736, + "step": 2128 + }, + { + "epoch": 2.095472440944882, + "grad_norm": 8.618043899536133, + "learning_rate": 4.2605848241615035e-07, + "loss": 0.3996, + "step": 2129 + }, + { + "epoch": 2.0964566929133857, + "grad_norm": 18.8110294342041, + "learning_rate": 4.2530076801128275e-07, + "loss": 0.7987, + "step": 2130 + }, + { + "epoch": 2.09744094488189, + "grad_norm": 8.883956909179688, + "learning_rate": 4.2454323436030457e-07, + "loss": 0.2163, + "step": 2131 + }, + { + "epoch": 2.0984251968503935, + "grad_norm": 11.823029518127441, + "learning_rate": 4.2378588325625627e-07, + "loss": 0.6272, + "step": 2132 + }, + { + "epoch": 2.0994094488188977, + "grad_norm": 13.031164169311523, + "learning_rate": 4.2302871649174646e-07, + "loss": 0.4459, + "step": 2133 + }, + { + "epoch": 2.1003937007874014, + "grad_norm": 11.237198829650879, + "learning_rate": 4.2227173585894724e-07, + "loss": 0.4126, + "step": 2134 + }, + { + "epoch": 2.1013779527559056, + "grad_norm": 12.452037811279297, + "learning_rate": 4.2151494314959015e-07, + "loss": 0.5132, + "step": 2135 + }, + { + "epoch": 2.1023622047244093, + "grad_norm": 5.083644866943359, + "learning_rate": 4.20758340154962e-07, + "loss": 0.0866, + "step": 2136 + }, + { + "epoch": 2.1033464566929134, + "grad_norm": 10.535178184509277, + "learning_rate": 4.200019286659007e-07, + "loss": 0.3466, + "step": 2137 + }, + { + "epoch": 2.104330708661417, + "grad_norm": 17.82977867126465, + "learning_rate": 4.192457104727905e-07, + "loss": 0.8075, + "step": 2138 + }, + { + "epoch": 2.1053149606299213, + "grad_norm": 12.288310050964355, + "learning_rate": 4.184896873655584e-07, + "loss": 0.4033, + "step": 2139 + }, + { + "epoch": 2.106299212598425, + "grad_norm": 8.79240608215332, + "learning_rate": 4.177338611336697e-07, + "loss": 0.3997, + "step": 2140 + }, + { + "epoch": 2.107283464566929, + "grad_norm": 7.214118957519531, + "learning_rate": 4.169782335661235e-07, + "loss": 0.1653, + "step": 2141 + }, + { + "epoch": 2.108267716535433, + "grad_norm": 12.292855262756348, + "learning_rate": 4.1622280645144883e-07, + "loss": 0.3839, + "step": 2142 + }, + { + "epoch": 2.108267716535433, + "eval_Qnli-dev_cosine_accuracy": 0.6875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8290761709213257, + "eval_Qnli-dev_cosine_ap": 0.6934549862457828, + "eval_Qnli-dev_cosine_f1": 0.6857142857142857, + "eval_Qnli-dev_cosine_f1_threshold": 0.8018802404403687, + "eval_Qnli-dev_cosine_precision": 0.6228373702422145, + "eval_Qnli-dev_cosine_recall": 0.7627118644067796, + "eval_Qnli-dev_dot_accuracy": 0.6875, + "eval_Qnli-dev_dot_accuracy_threshold": 628.5807495117188, + "eval_Qnli-dev_dot_ap": 0.6943891998231493, + "eval_Qnli-dev_dot_f1": 0.6871401151631478, + "eval_Qnli-dev_dot_f1_threshold": 614.128173828125, + "eval_Qnli-dev_dot_precision": 0.6280701754385964, + "eval_Qnli-dev_dot_recall": 0.7584745762711864, + "eval_Qnli-dev_euclidean_accuracy": 0.689453125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 16.34896469116211, + "eval_Qnli-dev_euclidean_ap": 0.6934492812737558, + "eval_Qnli-dev_euclidean_f1": 0.6857142857142857, + "eval_Qnli-dev_euclidean_f1_threshold": 17.400602340698242, + "eval_Qnli-dev_euclidean_precision": 0.6228373702422145, + "eval_Qnli-dev_euclidean_recall": 0.7627118644067796, + "eval_Qnli-dev_manhattan_accuracy": 0.689453125, + "eval_Qnli-dev_manhattan_accuracy_threshold": 352.53851318359375, + "eval_Qnli-dev_manhattan_ap": 0.6972539407103127, + "eval_Qnli-dev_manhattan_f1": 0.6924564796905222, + "eval_Qnli-dev_manhattan_f1_threshold": 380.4490051269531, + "eval_Qnli-dev_manhattan_precision": 0.6370106761565836, + "eval_Qnli-dev_manhattan_recall": 0.7584745762711864, + "eval_Qnli-dev_max_accuracy": 0.689453125, + "eval_Qnli-dev_max_accuracy_threshold": 628.5807495117188, + "eval_Qnli-dev_max_ap": 0.6972539407103127, + "eval_Qnli-dev_max_f1": 0.6924564796905222, + "eval_Qnli-dev_max_f1_threshold": 614.128173828125, + "eval_Qnli-dev_max_precision": 0.6370106761565836, + "eval_Qnli-dev_max_recall": 0.7627118644067796, + "eval_allNLI-dev_cosine_accuracy": 0.685546875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9349581003189087, + "eval_allNLI-dev_cosine_ap": 0.5165175342103042, + "eval_allNLI-dev_cosine_f1": 0.5597147950089127, + "eval_allNLI-dev_cosine_f1_threshold": 0.8061341047286987, + "eval_allNLI-dev_cosine_precision": 0.40463917525773196, + "eval_allNLI-dev_cosine_recall": 0.9075144508670521, + "eval_allNLI-dev_dot_accuracy": 0.685546875, + "eval_allNLI-dev_dot_accuracy_threshold": 714.799072265625, + "eval_allNLI-dev_dot_ap": 0.5174007045633133, + "eval_allNLI-dev_dot_f1": 0.5597147950089127, + "eval_allNLI-dev_dot_f1_threshold": 615.6774291992188, + "eval_allNLI-dev_dot_precision": 0.40463917525773196, + "eval_allNLI-dev_dot_recall": 0.9075144508670521, + "eval_allNLI-dev_euclidean_accuracy": 0.685546875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.972558975219727, + "eval_allNLI-dev_euclidean_ap": 0.5163551004521006, + "eval_allNLI-dev_euclidean_f1": 0.5597147950089127, + "eval_allNLI-dev_euclidean_f1_threshold": 17.21194839477539, + "eval_allNLI-dev_euclidean_precision": 0.40463917525773196, + "eval_allNLI-dev_euclidean_recall": 0.9075144508670521, + "eval_allNLI-dev_manhattan_accuracy": 0.6875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 274.15997314453125, + "eval_allNLI-dev_manhattan_ap": 0.5171478030416333, + "eval_allNLI-dev_manhattan_f1": 0.5641952983725136, + "eval_allNLI-dev_manhattan_f1_threshold": 377.43255615234375, + "eval_allNLI-dev_manhattan_precision": 0.4105263157894737, + "eval_allNLI-dev_manhattan_recall": 0.9017341040462428, + "eval_allNLI-dev_max_accuracy": 0.6875, + "eval_allNLI-dev_max_accuracy_threshold": 714.799072265625, + "eval_allNLI-dev_max_ap": 0.5174007045633133, + "eval_allNLI-dev_max_f1": 0.5641952983725136, + "eval_allNLI-dev_max_f1_threshold": 615.6774291992188, + "eval_allNLI-dev_max_precision": 0.4105263157894737, + "eval_allNLI-dev_max_recall": 0.9075144508670521, + "eval_loss": 0.758155107498169, + "eval_runtime": 50.5304, + "eval_samples_per_second": 32.931, + "eval_sequential_score": 0.6972539407103127, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.6954741487251752, + "eval_sts-test_pearson_dot": 0.6951395271561636, + "eval_sts-test_pearson_euclidean": 0.7169358064019951, + "eval_sts-test_pearson_manhattan": 0.7174687294256047, + "eval_sts-test_pearson_max": 0.7174687294256047, + "eval_sts-test_spearman_cosine": 0.7066826973192926, + "eval_sts-test_spearman_dot": 0.7061246043384803, + "eval_sts-test_spearman_euclidean": 0.7067148287249715, + "eval_sts-test_spearman_manhattan": 0.7073796982154636, + "eval_sts-test_spearman_max": 0.7073796982154636, + "step": 2142 + }, + { + "epoch": 2.109251968503937, + "grad_norm": 11.334417343139648, + "learning_rate": 4.1546758157770004e-07, + "loss": 0.3693, + "step": 2143 + }, + { + "epoch": 2.1102362204724407, + "grad_norm": 8.823077201843262, + "learning_rate": 4.1471256073245306e-07, + "loss": 0.2938, + "step": 2144 + }, + { + "epoch": 2.111220472440945, + "grad_norm": 15.313334465026855, + "learning_rate": 4.139577457028009e-07, + "loss": 0.7838, + "step": 2145 + }, + { + "epoch": 2.1122047244094486, + "grad_norm": 6.814699172973633, + "learning_rate": 4.13203138275349e-07, + "loss": 0.1539, + "step": 2146 + }, + { + "epoch": 2.1131889763779528, + "grad_norm": 10.211042404174805, + "learning_rate": 4.124487402362119e-07, + "loss": 0.2996, + "step": 2147 + }, + { + "epoch": 2.1141732283464565, + "grad_norm": 8.373467445373535, + "learning_rate": 4.116945533710084e-07, + "loss": 0.2718, + "step": 2148 + }, + { + "epoch": 2.1151574803149606, + "grad_norm": 8.609745979309082, + "learning_rate": 4.1094057946485717e-07, + "loss": 0.2482, + "step": 2149 + }, + { + "epoch": 2.1161417322834644, + "grad_norm": 11.000456809997559, + "learning_rate": 4.1018682030237324e-07, + "loss": 0.345, + "step": 2150 + }, + { + "epoch": 2.1171259842519685, + "grad_norm": 10.883407592773438, + "learning_rate": 4.0943327766766314e-07, + "loss": 0.3794, + "step": 2151 + }, + { + "epoch": 2.1181102362204722, + "grad_norm": 11.536715507507324, + "learning_rate": 4.086799533443208e-07, + "loss": 0.4432, + "step": 2152 + }, + { + "epoch": 2.1190944881889764, + "grad_norm": 13.385024070739746, + "learning_rate": 4.0792684911542373e-07, + "loss": 0.4275, + "step": 2153 + }, + { + "epoch": 2.12007874015748, + "grad_norm": 4.893040657043457, + "learning_rate": 4.071739667635281e-07, + "loss": 0.098, + "step": 2154 + }, + { + "epoch": 2.1210629921259843, + "grad_norm": 9.595856666564941, + "learning_rate": 4.064213080706652e-07, + "loss": 0.2238, + "step": 2155 + }, + { + "epoch": 2.122047244094488, + "grad_norm": 11.337559700012207, + "learning_rate": 4.056688748183368e-07, + "loss": 0.3703, + "step": 2156 + }, + { + "epoch": 2.123031496062992, + "grad_norm": 14.757852554321289, + "learning_rate": 4.049166687875111e-07, + "loss": 0.4959, + "step": 2157 + }, + { + "epoch": 2.1240157480314963, + "grad_norm": 9.285175323486328, + "learning_rate": 4.0416469175861855e-07, + "loss": 0.3255, + "step": 2158 + }, + { + "epoch": 2.125, + "grad_norm": 8.792654037475586, + "learning_rate": 4.034129455115474e-07, + "loss": 0.2369, + "step": 2159 + }, + { + "epoch": 2.1259842519685037, + "grad_norm": 12.386228561401367, + "learning_rate": 4.0266143182563974e-07, + "loss": 0.4698, + "step": 2160 + }, + { + "epoch": 2.126968503937008, + "grad_norm": 12.109793663024902, + "learning_rate": 4.019101524796874e-07, + "loss": 0.3598, + "step": 2161 + }, + { + "epoch": 2.127952755905512, + "grad_norm": 13.906962394714355, + "learning_rate": 4.0115910925192705e-07, + "loss": 0.7223, + "step": 2162 + }, + { + "epoch": 2.1289370078740157, + "grad_norm": 7.747769832611084, + "learning_rate": 4.004083039200369e-07, + "loss": 0.1452, + "step": 2163 + }, + { + "epoch": 2.1299212598425195, + "grad_norm": 11.867863655090332, + "learning_rate": 3.99657738261132e-07, + "loss": 0.3608, + "step": 2164 + }, + { + "epoch": 2.1309055118110236, + "grad_norm": 9.891944885253906, + "learning_rate": 3.9890741405176e-07, + "loss": 0.3727, + "step": 2165 + }, + { + "epoch": 2.1318897637795278, + "grad_norm": 10.925929069519043, + "learning_rate": 3.981573330678973e-07, + "loss": 0.3437, + "step": 2166 + }, + { + "epoch": 2.1328740157480315, + "grad_norm": 11.268866539001465, + "learning_rate": 3.9740749708494405e-07, + "loss": 0.4198, + "step": 2167 + }, + { + "epoch": 2.1338582677165356, + "grad_norm": 7.4711127281188965, + "learning_rate": 3.96657907877721e-07, + "loss": 0.2847, + "step": 2168 + }, + { + "epoch": 2.1348425196850394, + "grad_norm": 15.180150985717773, + "learning_rate": 3.959085672204647e-07, + "loss": 0.7157, + "step": 2169 + }, + { + "epoch": 2.1358267716535435, + "grad_norm": 12.531563758850098, + "learning_rate": 3.951594768868233e-07, + "loss": 0.2186, + "step": 2170 + }, + { + "epoch": 2.1368110236220472, + "grad_norm": 6.124568462371826, + "learning_rate": 3.9441063864985255e-07, + "loss": 0.1063, + "step": 2171 + }, + { + "epoch": 2.1377952755905514, + "grad_norm": 12.518921852111816, + "learning_rate": 3.936620542820114e-07, + "loss": 0.4995, + "step": 2172 + }, + { + "epoch": 2.138779527559055, + "grad_norm": 9.476792335510254, + "learning_rate": 3.9291372555515794e-07, + "loss": 0.2194, + "step": 2173 + }, + { + "epoch": 2.1397637795275593, + "grad_norm": 12.813644409179688, + "learning_rate": 3.921656542405452e-07, + "loss": 0.4858, + "step": 2174 + }, + { + "epoch": 2.140748031496063, + "grad_norm": 9.723433494567871, + "learning_rate": 3.9141784210881695e-07, + "loss": 0.3456, + "step": 2175 + }, + { + "epoch": 2.141732283464567, + "grad_norm": 10.258703231811523, + "learning_rate": 3.9067029093000335e-07, + "loss": 0.4129, + "step": 2176 + }, + { + "epoch": 2.142716535433071, + "grad_norm": 11.286211013793945, + "learning_rate": 3.899230024735171e-07, + "loss": 0.4063, + "step": 2177 + }, + { + "epoch": 2.143700787401575, + "grad_norm": 8.208313941955566, + "learning_rate": 3.8917597850814885e-07, + "loss": 0.2538, + "step": 2178 + }, + { + "epoch": 2.1446850393700787, + "grad_norm": 13.45827579498291, + "learning_rate": 3.884292208020635e-07, + "loss": 0.4645, + "step": 2179 + }, + { + "epoch": 2.145669291338583, + "grad_norm": 14.17795467376709, + "learning_rate": 3.876827311227954e-07, + "loss": 0.728, + "step": 2180 + }, + { + "epoch": 2.1466535433070866, + "grad_norm": 14.115456581115723, + "learning_rate": 3.8693651123724455e-07, + "loss": 0.4584, + "step": 2181 + }, + { + "epoch": 2.1476377952755907, + "grad_norm": 8.762264251708984, + "learning_rate": 3.861905629116725e-07, + "loss": 0.4323, + "step": 2182 + }, + { + "epoch": 2.1486220472440944, + "grad_norm": 9.946696281433105, + "learning_rate": 3.854448879116981e-07, + "loss": 0.3082, + "step": 2183 + }, + { + "epoch": 2.1496062992125986, + "grad_norm": 8.062214851379395, + "learning_rate": 3.84699488002293e-07, + "loss": 0.2129, + "step": 2184 + }, + { + "epoch": 2.1505905511811023, + "grad_norm": 9.5697660446167, + "learning_rate": 3.83954364947778e-07, + "loss": 0.3056, + "step": 2185 + }, + { + "epoch": 2.1515748031496065, + "grad_norm": 8.632911682128906, + "learning_rate": 3.832095205118183e-07, + "loss": 0.2874, + "step": 2186 + }, + { + "epoch": 2.15255905511811, + "grad_norm": 11.519652366638184, + "learning_rate": 3.8246495645742006e-07, + "loss": 0.4901, + "step": 2187 + }, + { + "epoch": 2.1535433070866143, + "grad_norm": 11.75122356414795, + "learning_rate": 3.8172067454692517e-07, + "loss": 0.2411, + "step": 2188 + }, + { + "epoch": 2.154527559055118, + "grad_norm": 8.060043334960938, + "learning_rate": 3.8097667654200826e-07, + "loss": 0.3098, + "step": 2189 + }, + { + "epoch": 2.155511811023622, + "grad_norm": 13.556063652038574, + "learning_rate": 3.802329642036716e-07, + "loss": 0.6039, + "step": 2190 + }, + { + "epoch": 2.156496062992126, + "grad_norm": 7.22255277633667, + "learning_rate": 3.7948953929224155e-07, + "loss": 0.1453, + "step": 2191 + }, + { + "epoch": 2.15748031496063, + "grad_norm": 8.92556095123291, + "learning_rate": 3.787464035673641e-07, + "loss": 0.2557, + "step": 2192 + }, + { + "epoch": 2.158464566929134, + "grad_norm": 4.606111526489258, + "learning_rate": 3.780035587880007e-07, + "loss": 0.0994, + "step": 2193 + }, + { + "epoch": 2.159448818897638, + "grad_norm": 8.817214965820312, + "learning_rate": 3.7726100671242414e-07, + "loss": 0.4115, + "step": 2194 + }, + { + "epoch": 2.1604330708661417, + "grad_norm": 10.211170196533203, + "learning_rate": 3.765187490982144e-07, + "loss": 0.3638, + "step": 2195 + }, + { + "epoch": 2.161417322834646, + "grad_norm": 12.256020545959473, + "learning_rate": 3.757767877022545e-07, + "loss": 0.6126, + "step": 2196 + }, + { + "epoch": 2.1624015748031495, + "grad_norm": 13.432823181152344, + "learning_rate": 3.750351242807264e-07, + "loss": 0.6262, + "step": 2197 + }, + { + "epoch": 2.1633858267716537, + "grad_norm": 8.21986198425293, + "learning_rate": 3.7429376058910674e-07, + "loss": 0.2684, + "step": 2198 + }, + { + "epoch": 2.1643700787401574, + "grad_norm": 9.793008804321289, + "learning_rate": 3.735526983821624e-07, + "loss": 0.2969, + "step": 2199 + }, + { + "epoch": 2.1653543307086616, + "grad_norm": 14.64374828338623, + "learning_rate": 3.728119394139472e-07, + "loss": 0.549, + "step": 2200 + }, + { + "epoch": 2.1663385826771653, + "grad_norm": 13.609929084777832, + "learning_rate": 3.7207148543779677e-07, + "loss": 0.6069, + "step": 2201 + }, + { + "epoch": 2.1673228346456694, + "grad_norm": 8.831336975097656, + "learning_rate": 3.713313382063253e-07, + "loss": 0.2243, + "step": 2202 + }, + { + "epoch": 2.168307086614173, + "grad_norm": 11.816344261169434, + "learning_rate": 3.705914994714203e-07, + "loss": 0.4659, + "step": 2203 + }, + { + "epoch": 2.1692913385826773, + "grad_norm": 12.861519813537598, + "learning_rate": 3.6985197098423957e-07, + "loss": 0.5116, + "step": 2204 + }, + { + "epoch": 2.170275590551181, + "grad_norm": 10.335835456848145, + "learning_rate": 3.691127544952065e-07, + "loss": 0.34, + "step": 2205 + }, + { + "epoch": 2.171259842519685, + "grad_norm": 9.590200424194336, + "learning_rate": 3.6837385175400597e-07, + "loss": 0.4082, + "step": 2206 + }, + { + "epoch": 2.172244094488189, + "grad_norm": 6.215512752532959, + "learning_rate": 3.6763526450958005e-07, + "loss": 0.1625, + "step": 2207 + }, + { + "epoch": 2.173228346456693, + "grad_norm": 8.190068244934082, + "learning_rate": 3.668969945101245e-07, + "loss": 0.1897, + "step": 2208 + }, + { + "epoch": 2.1742125984251968, + "grad_norm": 8.764188766479492, + "learning_rate": 3.6615904350308343e-07, + "loss": 0.2351, + "step": 2209 + }, + { + "epoch": 2.175196850393701, + "grad_norm": 9.251346588134766, + "learning_rate": 3.6542141323514666e-07, + "loss": 0.3023, + "step": 2210 + }, + { + "epoch": 2.1761811023622046, + "grad_norm": 9.985815048217773, + "learning_rate": 3.6468410545224436e-07, + "loss": 0.4175, + "step": 2211 + }, + { + "epoch": 2.177165354330709, + "grad_norm": 9.384307861328125, + "learning_rate": 3.6394712189954365e-07, + "loss": 0.3834, + "step": 2212 + }, + { + "epoch": 2.1781496062992125, + "grad_norm": 8.385250091552734, + "learning_rate": 3.632104643214442e-07, + "loss": 0.2453, + "step": 2213 + }, + { + "epoch": 2.1791338582677167, + "grad_norm": 9.428893089294434, + "learning_rate": 3.62474134461574e-07, + "loss": 0.2864, + "step": 2214 + }, + { + "epoch": 2.1801181102362204, + "grad_norm": 6.5480828285217285, + "learning_rate": 3.6173813406278543e-07, + "loss": 0.1528, + "step": 2215 + }, + { + "epoch": 2.1811023622047245, + "grad_norm": 13.004796028137207, + "learning_rate": 3.610024648671511e-07, + "loss": 0.4384, + "step": 2216 + }, + { + "epoch": 2.1820866141732282, + "grad_norm": 10.215922355651855, + "learning_rate": 3.602671286159594e-07, + "loss": 0.269, + "step": 2217 + }, + { + "epoch": 2.1830708661417324, + "grad_norm": 7.948540687561035, + "learning_rate": 3.5953212704971096e-07, + "loss": 0.287, + "step": 2218 + }, + { + "epoch": 2.184055118110236, + "grad_norm": 14.778656005859375, + "learning_rate": 3.5879746190811415e-07, + "loss": 0.7013, + "step": 2219 + }, + { + "epoch": 2.1850393700787403, + "grad_norm": 8.496538162231445, + "learning_rate": 3.5806313493008104e-07, + "loss": 0.3957, + "step": 2220 + }, + { + "epoch": 2.186023622047244, + "grad_norm": 10.980823516845703, + "learning_rate": 3.573291478537231e-07, + "loss": 0.4059, + "step": 2221 + }, + { + "epoch": 2.187007874015748, + "grad_norm": 8.794949531555176, + "learning_rate": 3.565955024163477e-07, + "loss": 0.2807, + "step": 2222 + }, + { + "epoch": 2.187992125984252, + "grad_norm": 4.939118385314941, + "learning_rate": 3.5586220035445327e-07, + "loss": 0.1012, + "step": 2223 + }, + { + "epoch": 2.188976377952756, + "grad_norm": 11.635772705078125, + "learning_rate": 3.551292434037254e-07, + "loss": 0.5463, + "step": 2224 + }, + { + "epoch": 2.1899606299212597, + "grad_norm": 3.9744551181793213, + "learning_rate": 3.54396633299033e-07, + "loss": 0.0915, + "step": 2225 + }, + { + "epoch": 2.190944881889764, + "grad_norm": 7.735105037689209, + "learning_rate": 3.5366437177442394e-07, + "loss": 0.1824, + "step": 2226 + }, + { + "epoch": 2.1919291338582676, + "grad_norm": 12.265244483947754, + "learning_rate": 3.529324605631211e-07, + "loss": 0.3098, + "step": 2227 + }, + { + "epoch": 2.1929133858267718, + "grad_norm": 4.93965482711792, + "learning_rate": 3.522009013975184e-07, + "loss": 0.1096, + "step": 2228 + }, + { + "epoch": 2.1938976377952755, + "grad_norm": 8.849276542663574, + "learning_rate": 3.514696960091756e-07, + "loss": 0.2096, + "step": 2229 + }, + { + "epoch": 2.1948818897637796, + "grad_norm": 10.99960708618164, + "learning_rate": 3.5073884612881616e-07, + "loss": 0.3661, + "step": 2230 + }, + { + "epoch": 2.1958661417322833, + "grad_norm": 8.13372802734375, + "learning_rate": 3.500083534863215e-07, + "loss": 0.2762, + "step": 2231 + }, + { + "epoch": 2.1968503937007875, + "grad_norm": 8.868273735046387, + "learning_rate": 3.492782198107275e-07, + "loss": 0.2249, + "step": 2232 + }, + { + "epoch": 2.197834645669291, + "grad_norm": 7.006740093231201, + "learning_rate": 3.485484468302207e-07, + "loss": 0.0947, + "step": 2233 + }, + { + "epoch": 2.1988188976377954, + "grad_norm": 14.659770011901855, + "learning_rate": 3.4781903627213347e-07, + "loss": 0.5143, + "step": 2234 + }, + { + "epoch": 2.199803149606299, + "grad_norm": 18.32170295715332, + "learning_rate": 3.470899898629408e-07, + "loss": 0.9211, + "step": 2235 + }, + { + "epoch": 2.2007874015748032, + "grad_norm": 13.795584678649902, + "learning_rate": 3.463613093282552e-07, + "loss": 0.4362, + "step": 2236 + }, + { + "epoch": 2.201771653543307, + "grad_norm": 9.573206901550293, + "learning_rate": 3.456329963928237e-07, + "loss": 0.3663, + "step": 2237 + }, + { + "epoch": 2.202755905511811, + "grad_norm": 11.189313888549805, + "learning_rate": 3.4490505278052295e-07, + "loss": 0.2053, + "step": 2238 + }, + { + "epoch": 2.203740157480315, + "grad_norm": 15.250197410583496, + "learning_rate": 3.4417748021435563e-07, + "loss": 0.7213, + "step": 2239 + }, + { + "epoch": 2.204724409448819, + "grad_norm": 15.996099472045898, + "learning_rate": 3.4345028041644595e-07, + "loss": 0.7317, + "step": 2240 + }, + { + "epoch": 2.2057086614173227, + "grad_norm": 20.053302764892578, + "learning_rate": 3.427234551080358e-07, + "loss": 1.1896, + "step": 2241 + }, + { + "epoch": 2.206692913385827, + "grad_norm": 13.37335205078125, + "learning_rate": 3.41997006009481e-07, + "loss": 0.585, + "step": 2242 + }, + { + "epoch": 2.2076771653543306, + "grad_norm": 12.944981575012207, + "learning_rate": 3.4127093484024656e-07, + "loss": 0.6656, + "step": 2243 + }, + { + "epoch": 2.2086614173228347, + "grad_norm": 7.117852210998535, + "learning_rate": 3.4054524331890316e-07, + "loss": 0.19, + "step": 2244 + }, + { + "epoch": 2.2096456692913384, + "grad_norm": 13.761758804321289, + "learning_rate": 3.3981993316312277e-07, + "loss": 0.4805, + "step": 2245 + }, + { + "epoch": 2.2106299212598426, + "grad_norm": 9.12771224975586, + "learning_rate": 3.390950060896746e-07, + "loss": 0.3275, + "step": 2246 + }, + { + "epoch": 2.2116141732283463, + "grad_norm": 9.08574104309082, + "learning_rate": 3.3837046381442135e-07, + "loss": 0.2341, + "step": 2247 + }, + { + "epoch": 2.2125984251968505, + "grad_norm": 6.652701377868652, + "learning_rate": 3.376463080523147e-07, + "loss": 0.1437, + "step": 2248 + }, + { + "epoch": 2.213582677165354, + "grad_norm": 9.009031295776367, + "learning_rate": 3.369225405173918e-07, + "loss": 0.3254, + "step": 2249 + }, + { + "epoch": 2.2145669291338583, + "grad_norm": 7.903627872467041, + "learning_rate": 3.361991629227704e-07, + "loss": 0.252, + "step": 2250 + }, + { + "epoch": 2.215551181102362, + "grad_norm": 13.05065631866455, + "learning_rate": 3.354761769806456e-07, + "loss": 0.3776, + "step": 2251 + }, + { + "epoch": 2.216535433070866, + "grad_norm": 8.087733268737793, + "learning_rate": 3.3475358440228547e-07, + "loss": 0.1664, + "step": 2252 + }, + { + "epoch": 2.21751968503937, + "grad_norm": 10.325349807739258, + "learning_rate": 3.340313868980271e-07, + "loss": 0.3588, + "step": 2253 + }, + { + "epoch": 2.218503937007874, + "grad_norm": 7.86494255065918, + "learning_rate": 3.3330958617727213e-07, + "loss": 0.3109, + "step": 2254 + }, + { + "epoch": 2.219488188976378, + "grad_norm": 7.871631622314453, + "learning_rate": 3.325881839484833e-07, + "loss": 0.2717, + "step": 2255 + }, + { + "epoch": 2.220472440944882, + "grad_norm": 11.341468811035156, + "learning_rate": 3.318671819191802e-07, + "loss": 0.4369, + "step": 2256 + }, + { + "epoch": 2.2214566929133857, + "grad_norm": 7.715664863586426, + "learning_rate": 3.3114658179593487e-07, + "loss": 0.1796, + "step": 2257 + }, + { + "epoch": 2.22244094488189, + "grad_norm": 14.130294799804688, + "learning_rate": 3.304263852843683e-07, + "loss": 0.4716, + "step": 2258 + }, + { + "epoch": 2.2234251968503935, + "grad_norm": 10.31694507598877, + "learning_rate": 3.297065940891462e-07, + "loss": 0.404, + "step": 2259 + }, + { + "epoch": 2.2244094488188977, + "grad_norm": 10.897465705871582, + "learning_rate": 3.289872099139746e-07, + "loss": 0.501, + "step": 2260 + }, + { + "epoch": 2.2253937007874014, + "grad_norm": 12.96509075164795, + "learning_rate": 3.282682344615966e-07, + "loss": 0.5194, + "step": 2261 + }, + { + "epoch": 2.2263779527559056, + "grad_norm": 7.422422409057617, + "learning_rate": 3.275496694337874e-07, + "loss": 0.1071, + "step": 2262 + }, + { + "epoch": 2.2273622047244093, + "grad_norm": 5.680365085601807, + "learning_rate": 3.2683151653135104e-07, + "loss": 0.1748, + "step": 2263 + }, + { + "epoch": 2.2283464566929134, + "grad_norm": 8.01179027557373, + "learning_rate": 3.261137774541161e-07, + "loss": 0.2036, + "step": 2264 + }, + { + "epoch": 2.229330708661417, + "grad_norm": 11.250329971313477, + "learning_rate": 3.2539645390093155e-07, + "loss": 0.3187, + "step": 2265 + }, + { + "epoch": 2.2303149606299213, + "grad_norm": 16.47153091430664, + "learning_rate": 3.2467954756966276e-07, + "loss": 0.7576, + "step": 2266 + }, + { + "epoch": 2.231299212598425, + "grad_norm": 4.137571811676025, + "learning_rate": 3.2396306015718775e-07, + "loss": 0.0618, + "step": 2267 + }, + { + "epoch": 2.232283464566929, + "grad_norm": 12.095766067504883, + "learning_rate": 3.2324699335939273e-07, + "loss": 0.3125, + "step": 2268 + }, + { + "epoch": 2.233267716535433, + "grad_norm": 8.904653549194336, + "learning_rate": 3.225313488711686e-07, + "loss": 0.162, + "step": 2269 + }, + { + "epoch": 2.234251968503937, + "grad_norm": 15.208680152893066, + "learning_rate": 3.2181612838640673e-07, + "loss": 0.4277, + "step": 2270 + }, + { + "epoch": 2.2352362204724407, + "grad_norm": 16.477054595947266, + "learning_rate": 3.2110133359799457e-07, + "loss": 0.6385, + "step": 2271 + }, + { + "epoch": 2.236220472440945, + "grad_norm": 10.606884002685547, + "learning_rate": 3.203869661978121e-07, + "loss": 0.2628, + "step": 2272 + }, + { + "epoch": 2.2372047244094486, + "grad_norm": 12.28905963897705, + "learning_rate": 3.196730278767277e-07, + "loss": 0.5488, + "step": 2273 + }, + { + "epoch": 2.2381889763779528, + "grad_norm": 14.246539115905762, + "learning_rate": 3.189595203245944e-07, + "loss": 0.695, + "step": 2274 + }, + { + "epoch": 2.2391732283464565, + "grad_norm": 7.976707935333252, + "learning_rate": 3.182464452302452e-07, + "loss": 0.2487, + "step": 2275 + }, + { + "epoch": 2.2401574803149606, + "grad_norm": 14.724424362182617, + "learning_rate": 3.175338042814898e-07, + "loss": 0.5041, + "step": 2276 + }, + { + "epoch": 2.2411417322834644, + "grad_norm": 1.95339834690094, + "learning_rate": 3.168215991651101e-07, + "loss": 0.0307, + "step": 2277 + }, + { + "epoch": 2.2421259842519685, + "grad_norm": 9.749213218688965, + "learning_rate": 3.1610983156685664e-07, + "loss": 0.3909, + "step": 2278 + }, + { + "epoch": 2.2431102362204722, + "grad_norm": 9.426769256591797, + "learning_rate": 3.153985031714442e-07, + "loss": 0.2115, + "step": 2279 + }, + { + "epoch": 2.2440944881889764, + "grad_norm": 7.367453575134277, + "learning_rate": 3.14687615662548e-07, + "loss": 0.1853, + "step": 2280 + }, + { + "epoch": 2.2450787401574805, + "grad_norm": 7.429681777954102, + "learning_rate": 3.139771707227997e-07, + "loss": 0.224, + "step": 2281 + }, + { + "epoch": 2.2460629921259843, + "grad_norm": 9.328291893005371, + "learning_rate": 3.1326717003378365e-07, + "loss": 0.3116, + "step": 2282 + }, + { + "epoch": 2.247047244094488, + "grad_norm": 11.240273475646973, + "learning_rate": 3.1255761527603233e-07, + "loss": 0.382, + "step": 2283 + }, + { + "epoch": 2.248031496062992, + "grad_norm": 10.6027250289917, + "learning_rate": 3.1184850812902304e-07, + "loss": 0.2589, + "step": 2284 + }, + { + "epoch": 2.2490157480314963, + "grad_norm": 9.430268287658691, + "learning_rate": 3.111398502711733e-07, + "loss": 0.2255, + "step": 2285 + }, + { + "epoch": 2.25, + "grad_norm": 5.602982997894287, + "learning_rate": 3.104316433798374e-07, + "loss": 0.1913, + "step": 2286 + }, + { + "epoch": 2.2509842519685037, + "grad_norm": 8.843626022338867, + "learning_rate": 3.0972388913130215e-07, + "loss": 0.1469, + "step": 2287 + }, + { + "epoch": 2.251968503937008, + "grad_norm": 4.191643714904785, + "learning_rate": 3.0901658920078314e-07, + "loss": 0.0531, + "step": 2288 + }, + { + "epoch": 2.252952755905512, + "grad_norm": 15.751923561096191, + "learning_rate": 3.083097452624203e-07, + "loss": 0.8686, + "step": 2289 + }, + { + "epoch": 2.2539370078740157, + "grad_norm": 11.915505409240723, + "learning_rate": 3.076033589892745e-07, + "loss": 0.4408, + "step": 2290 + }, + { + "epoch": 2.2549212598425195, + "grad_norm": 8.81048583984375, + "learning_rate": 3.0689743205332345e-07, + "loss": 0.331, + "step": 2291 + }, + { + "epoch": 2.2559055118110236, + "grad_norm": 11.591867446899414, + "learning_rate": 3.061919661254571e-07, + "loss": 0.4179, + "step": 2292 + }, + { + "epoch": 2.2568897637795278, + "grad_norm": 4.445011138916016, + "learning_rate": 3.054869628754748e-07, + "loss": 0.0832, + "step": 2293 + }, + { + "epoch": 2.2578740157480315, + "grad_norm": 10.183629035949707, + "learning_rate": 3.047824239720804e-07, + "loss": 0.3035, + "step": 2294 + }, + { + "epoch": 2.258858267716535, + "grad_norm": 14.948867797851562, + "learning_rate": 3.0407835108287885e-07, + "loss": 0.6726, + "step": 2295 + }, + { + "epoch": 2.258858267716535, + "eval_Qnli-dev_cosine_accuracy": 0.693359375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7911537289619446, + "eval_Qnli-dev_cosine_ap": 0.7170058408023475, + "eval_Qnli-dev_cosine_f1": 0.6943396226415095, + "eval_Qnli-dev_cosine_f1_threshold": 0.7661037445068359, + "eval_Qnli-dev_cosine_precision": 0.6258503401360545, + "eval_Qnli-dev_cosine_recall": 0.7796610169491526, + "eval_Qnli-dev_dot_accuracy": 0.69140625, + "eval_Qnli-dev_dot_accuracy_threshold": 604.568359375, + "eval_Qnli-dev_dot_ap": 0.7180814793300055, + "eval_Qnli-dev_dot_f1": 0.6943396226415095, + "eval_Qnli-dev_dot_f1_threshold": 584.6365966796875, + "eval_Qnli-dev_dot_precision": 0.6258503401360545, + "eval_Qnli-dev_dot_recall": 0.7796610169491526, + "eval_Qnli-dev_euclidean_accuracy": 0.693359375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.86266326904297, + "eval_Qnli-dev_euclidean_ap": 0.7167111722090085, + "eval_Qnli-dev_euclidean_f1": 0.6943396226415095, + "eval_Qnli-dev_euclidean_f1_threshold": 18.894086837768555, + "eval_Qnli-dev_euclidean_precision": 0.6258503401360545, + "eval_Qnli-dev_euclidean_recall": 0.7796610169491526, + "eval_Qnli-dev_manhattan_accuracy": 0.6875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 394.7893371582031, + "eval_Qnli-dev_manhattan_ap": 0.7191161344564313, + "eval_Qnli-dev_manhattan_f1": 0.6865671641791045, + "eval_Qnli-dev_manhattan_f1_threshold": 419.76483154296875, + "eval_Qnli-dev_manhattan_precision": 0.6133333333333333, + "eval_Qnli-dev_manhattan_recall": 0.7796610169491526, + "eval_Qnli-dev_max_accuracy": 0.693359375, + "eval_Qnli-dev_max_accuracy_threshold": 604.568359375, + "eval_Qnli-dev_max_ap": 0.7191161344564313, + "eval_Qnli-dev_max_f1": 0.6943396226415095, + "eval_Qnli-dev_max_f1_threshold": 584.6365966796875, + "eval_Qnli-dev_max_precision": 0.6258503401360545, + "eval_Qnli-dev_max_recall": 0.7796610169491526, + "eval_allNLI-dev_cosine_accuracy": 0.6953125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.916883111000061, + "eval_allNLI-dev_cosine_ap": 0.5389132858087206, + "eval_allNLI-dev_cosine_f1": 0.5739514348785872, + "eval_allNLI-dev_cosine_f1_threshold": 0.8422929048538208, + "eval_allNLI-dev_cosine_precision": 0.4642857142857143, + "eval_allNLI-dev_cosine_recall": 0.7514450867052023, + "eval_allNLI-dev_dot_accuracy": 0.697265625, + "eval_allNLI-dev_dot_accuracy_threshold": 700.6588134765625, + "eval_allNLI-dev_dot_ap": 0.5386550373326439, + "eval_allNLI-dev_dot_f1": 0.5738396624472574, + "eval_allNLI-dev_dot_f1_threshold": 637.1109619140625, + "eval_allNLI-dev_dot_precision": 0.45182724252491696, + "eval_allNLI-dev_dot_recall": 0.7861271676300579, + "eval_allNLI-dev_euclidean_accuracy": 0.6953125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.268974304199219, + "eval_allNLI-dev_euclidean_ap": 0.538861613459563, + "eval_allNLI-dev_euclidean_f1": 0.5739514348785872, + "eval_allNLI-dev_euclidean_f1_threshold": 15.526180267333984, + "eval_allNLI-dev_euclidean_precision": 0.4642857142857143, + "eval_allNLI-dev_euclidean_recall": 0.7514450867052023, + "eval_allNLI-dev_manhattan_accuracy": 0.6953125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 250.47450256347656, + "eval_allNLI-dev_manhattan_ap": 0.536589052216218, + "eval_allNLI-dev_manhattan_f1": 0.5768421052631579, + "eval_allNLI-dev_manhattan_f1_threshold": 352.81256103515625, + "eval_allNLI-dev_manhattan_precision": 0.45364238410596025, + "eval_allNLI-dev_manhattan_recall": 0.791907514450867, + "eval_allNLI-dev_max_accuracy": 0.697265625, + "eval_allNLI-dev_max_accuracy_threshold": 700.6588134765625, + "eval_allNLI-dev_max_ap": 0.5389132858087206, + "eval_allNLI-dev_max_f1": 0.5768421052631579, + "eval_allNLI-dev_max_f1_threshold": 637.1109619140625, + "eval_allNLI-dev_max_precision": 0.4642857142857143, + "eval_allNLI-dev_max_recall": 0.791907514450867, + "eval_loss": 0.7565979957580566, + "eval_runtime": 50.1674, + "eval_samples_per_second": 33.169, + "eval_sequential_score": 0.7191161344564313, + "eval_steps_per_second": 0.14, + "eval_sts-test_pearson_cosine": 0.6922684863371262, + "eval_sts-test_pearson_dot": 0.6921375998703381, + "eval_sts-test_pearson_euclidean": 0.7142932019256403, + "eval_sts-test_pearson_manhattan": 0.7157248829465845, + "eval_sts-test_pearson_max": 0.7157248829465845, + "eval_sts-test_spearman_cosine": 0.703903241225534, + "eval_sts-test_spearman_dot": 0.7039481872265008, + "eval_sts-test_spearman_euclidean": 0.704003394548528, + "eval_sts-test_spearman_manhattan": 0.7048857551693822, + "eval_sts-test_spearman_max": 0.7048857551693822, + "step": 2295 + }, + { + "epoch": 2.2598425196850394, + "grad_norm": 4.843102931976318, + "learning_rate": 3.03374745874372e-07, + "loss": 0.1025, + "step": 2296 + }, + { + "epoch": 2.2608267716535435, + "grad_norm": 12.132405281066895, + "learning_rate": 3.0267161001195493e-07, + "loss": 0.3284, + "step": 2297 + }, + { + "epoch": 2.2618110236220472, + "grad_norm": 7.466048717498779, + "learning_rate": 3.019689451599114e-07, + "loss": 0.1646, + "step": 2298 + }, + { + "epoch": 2.262795275590551, + "grad_norm": 14.641806602478027, + "learning_rate": 3.0126675298141056e-07, + "loss": 0.5941, + "step": 2299 + }, + { + "epoch": 2.263779527559055, + "grad_norm": 10.814557075500488, + "learning_rate": 3.005650351385029e-07, + "loss": 0.2353, + "step": 2300 + }, + { + "epoch": 2.2647637795275593, + "grad_norm": 12.832300186157227, + "learning_rate": 2.998637932921159e-07, + "loss": 0.4221, + "step": 2301 + }, + { + "epoch": 2.265748031496063, + "grad_norm": 5.720351696014404, + "learning_rate": 2.991630291020506e-07, + "loss": 0.1688, + "step": 2302 + }, + { + "epoch": 2.266732283464567, + "grad_norm": 17.98371696472168, + "learning_rate": 2.9846274422697734e-07, + "loss": 0.7758, + "step": 2303 + }, + { + "epoch": 2.267716535433071, + "grad_norm": 14.842743873596191, + "learning_rate": 2.9776294032443184e-07, + "loss": 0.5717, + "step": 2304 + }, + { + "epoch": 2.268700787401575, + "grad_norm": 8.881214141845703, + "learning_rate": 2.9706361905081164e-07, + "loss": 0.2188, + "step": 2305 + }, + { + "epoch": 2.2696850393700787, + "grad_norm": 7.366771697998047, + "learning_rate": 2.9636478206137174e-07, + "loss": 0.2082, + "step": 2306 + }, + { + "epoch": 2.270669291338583, + "grad_norm": 9.068266868591309, + "learning_rate": 2.95666431010221e-07, + "loss": 0.171, + "step": 2307 + }, + { + "epoch": 2.2716535433070866, + "grad_norm": 8.417882919311523, + "learning_rate": 2.949685675503179e-07, + "loss": 0.2378, + "step": 2308 + }, + { + "epoch": 2.2726377952755907, + "grad_norm": 10.107744216918945, + "learning_rate": 2.9427119333346706e-07, + "loss": 0.3472, + "step": 2309 + }, + { + "epoch": 2.2736220472440944, + "grad_norm": 6.765593528747559, + "learning_rate": 2.935743100103149e-07, + "loss": 0.1636, + "step": 2310 + }, + { + "epoch": 2.2746062992125986, + "grad_norm": 8.884689331054688, + "learning_rate": 2.9287791923034593e-07, + "loss": 0.3438, + "step": 2311 + }, + { + "epoch": 2.2755905511811023, + "grad_norm": 6.2753825187683105, + "learning_rate": 2.921820226418793e-07, + "loss": 0.1726, + "step": 2312 + }, + { + "epoch": 2.2765748031496065, + "grad_norm": 3.9449446201324463, + "learning_rate": 2.9148662189206344e-07, + "loss": 0.0537, + "step": 2313 + }, + { + "epoch": 2.27755905511811, + "grad_norm": 12.908712387084961, + "learning_rate": 2.907917186268739e-07, + "loss": 0.5469, + "step": 2314 + }, + { + "epoch": 2.2785433070866143, + "grad_norm": 8.223078727722168, + "learning_rate": 2.900973144911088e-07, + "loss": 0.2189, + "step": 2315 + }, + { + "epoch": 2.279527559055118, + "grad_norm": 16.003049850463867, + "learning_rate": 2.894034111283843e-07, + "loss": 0.8666, + "step": 2316 + }, + { + "epoch": 2.280511811023622, + "grad_norm": 10.45720386505127, + "learning_rate": 2.8871001018113183e-07, + "loss": 0.2347, + "step": 2317 + }, + { + "epoch": 2.281496062992126, + "grad_norm": 9.887803077697754, + "learning_rate": 2.880171132905932e-07, + "loss": 0.2223, + "step": 2318 + }, + { + "epoch": 2.28248031496063, + "grad_norm": 16.266801834106445, + "learning_rate": 2.8732472209681733e-07, + "loss": 0.4252, + "step": 2319 + }, + { + "epoch": 2.283464566929134, + "grad_norm": 8.588350296020508, + "learning_rate": 2.866328382386562e-07, + "loss": 0.2339, + "step": 2320 + }, + { + "epoch": 2.284448818897638, + "grad_norm": 7.157125949859619, + "learning_rate": 2.859414633537609e-07, + "loss": 0.2267, + "step": 2321 + }, + { + "epoch": 2.2854330708661417, + "grad_norm": 15.318318367004395, + "learning_rate": 2.8525059907857786e-07, + "loss": 0.7197, + "step": 2322 + }, + { + "epoch": 2.286417322834646, + "grad_norm": 8.042208671569824, + "learning_rate": 2.8456024704834483e-07, + "loss": 0.1621, + "step": 2323 + }, + { + "epoch": 2.2874015748031495, + "grad_norm": 9.61860179901123, + "learning_rate": 2.8387040889708715e-07, + "loss": 0.2266, + "step": 2324 + }, + { + "epoch": 2.2883858267716537, + "grad_norm": 6.121631145477295, + "learning_rate": 2.8318108625761395e-07, + "loss": 0.1613, + "step": 2325 + }, + { + "epoch": 2.2893700787401574, + "grad_norm": 13.954069137573242, + "learning_rate": 2.8249228076151396e-07, + "loss": 0.629, + "step": 2326 + }, + { + "epoch": 2.2903543307086616, + "grad_norm": 10.77175521850586, + "learning_rate": 2.81803994039152e-07, + "loss": 0.2812, + "step": 2327 + }, + { + "epoch": 2.2913385826771653, + "grad_norm": 13.205397605895996, + "learning_rate": 2.8111622771966523e-07, + "loss": 0.5929, + "step": 2328 + }, + { + "epoch": 2.2923228346456694, + "grad_norm": 10.837924003601074, + "learning_rate": 2.804289834309583e-07, + "loss": 0.4375, + "step": 2329 + }, + { + "epoch": 2.293307086614173, + "grad_norm": 15.504312515258789, + "learning_rate": 2.797422627997009e-07, + "loss": 0.6752, + "step": 2330 + }, + { + "epoch": 2.2942913385826773, + "grad_norm": 8.454350471496582, + "learning_rate": 2.7905606745132307e-07, + "loss": 0.1592, + "step": 2331 + }, + { + "epoch": 2.295275590551181, + "grad_norm": 9.216939926147461, + "learning_rate": 2.7837039901001155e-07, + "loss": 0.3056, + "step": 2332 + }, + { + "epoch": 2.296259842519685, + "grad_norm": 12.32398509979248, + "learning_rate": 2.776852590987061e-07, + "loss": 0.4122, + "step": 2333 + }, + { + "epoch": 2.297244094488189, + "grad_norm": 11.984452247619629, + "learning_rate": 2.7700064933909474e-07, + "loss": 0.4048, + "step": 2334 + }, + { + "epoch": 2.298228346456693, + "grad_norm": 15.117887496948242, + "learning_rate": 2.7631657135161156e-07, + "loss": 0.5066, + "step": 2335 + }, + { + "epoch": 2.2992125984251968, + "grad_norm": 11.230385780334473, + "learning_rate": 2.756330267554316e-07, + "loss": 0.5119, + "step": 2336 + }, + { + "epoch": 2.300196850393701, + "grad_norm": 16.905982971191406, + "learning_rate": 2.749500171684673e-07, + "loss": 0.6747, + "step": 2337 + }, + { + "epoch": 2.3011811023622046, + "grad_norm": 4.9277472496032715, + "learning_rate": 2.7426754420736485e-07, + "loss": 0.0582, + "step": 2338 + }, + { + "epoch": 2.302165354330709, + "grad_norm": 11.048531532287598, + "learning_rate": 2.735856094875003e-07, + "loss": 0.2521, + "step": 2339 + }, + { + "epoch": 2.3031496062992125, + "grad_norm": 3.776963472366333, + "learning_rate": 2.7290421462297574e-07, + "loss": 0.0696, + "step": 2340 + }, + { + "epoch": 2.3041338582677167, + "grad_norm": 13.674766540527344, + "learning_rate": 2.7222336122661535e-07, + "loss": 0.5675, + "step": 2341 + }, + { + "epoch": 2.3051181102362204, + "grad_norm": 9.75919246673584, + "learning_rate": 2.715430509099618e-07, + "loss": 0.1925, + "step": 2342 + }, + { + "epoch": 2.3061023622047245, + "grad_norm": 11.845928192138672, + "learning_rate": 2.7086328528327215e-07, + "loss": 0.5663, + "step": 2343 + }, + { + "epoch": 2.3070866141732282, + "grad_norm": 10.715153694152832, + "learning_rate": 2.7018406595551434e-07, + "loss": 0.4268, + "step": 2344 + }, + { + "epoch": 2.3080708661417324, + "grad_norm": 13.037115097045898, + "learning_rate": 2.695053945343633e-07, + "loss": 0.65, + "step": 2345 + }, + { + "epoch": 2.309055118110236, + "grad_norm": 7.811085224151611, + "learning_rate": 2.6882727262619696e-07, + "loss": 0.2004, + "step": 2346 + }, + { + "epoch": 2.3100393700787403, + "grad_norm": 13.735000610351562, + "learning_rate": 2.681497018360927e-07, + "loss": 0.3069, + "step": 2347 + }, + { + "epoch": 2.311023622047244, + "grad_norm": 10.576314926147461, + "learning_rate": 2.6747268376782334e-07, + "loss": 0.3747, + "step": 2348 + }, + { + "epoch": 2.312007874015748, + "grad_norm": 10.604143142700195, + "learning_rate": 2.6679622002385354e-07, + "loss": 0.4447, + "step": 2349 + }, + { + "epoch": 2.312992125984252, + "grad_norm": 15.008903503417969, + "learning_rate": 2.6612031220533574e-07, + "loss": 0.7273, + "step": 2350 + }, + { + "epoch": 2.313976377952756, + "grad_norm": 13.090288162231445, + "learning_rate": 2.654449619121067e-07, + "loss": 0.3773, + "step": 2351 + }, + { + "epoch": 2.3149606299212597, + "grad_norm": 11.25436019897461, + "learning_rate": 2.647701707426835e-07, + "loss": 0.5278, + "step": 2352 + }, + { + "epoch": 2.315944881889764, + "grad_norm": 7.501079082489014, + "learning_rate": 2.6409594029425986e-07, + "loss": 0.2348, + "step": 2353 + }, + { + "epoch": 2.3169291338582676, + "grad_norm": 10.106515884399414, + "learning_rate": 2.6342227216270205e-07, + "loss": 0.4442, + "step": 2354 + }, + { + "epoch": 2.3179133858267718, + "grad_norm": 10.640914916992188, + "learning_rate": 2.6274916794254554e-07, + "loss": 0.3665, + "step": 2355 + }, + { + "epoch": 2.3188976377952755, + "grad_norm": 12.370466232299805, + "learning_rate": 2.6207662922699125e-07, + "loss": 0.3878, + "step": 2356 + }, + { + "epoch": 2.3198818897637796, + "grad_norm": 9.007702827453613, + "learning_rate": 2.6140465760790126e-07, + "loss": 0.1487, + "step": 2357 + }, + { + "epoch": 2.3208661417322833, + "grad_norm": 7.668280601501465, + "learning_rate": 2.6073325467579553e-07, + "loss": 0.1847, + "step": 2358 + }, + { + "epoch": 2.3218503937007875, + "grad_norm": 8.930326461791992, + "learning_rate": 2.60062422019848e-07, + "loss": 0.2042, + "step": 2359 + }, + { + "epoch": 2.322834645669291, + "grad_norm": 6.747073173522949, + "learning_rate": 2.5939216122788267e-07, + "loss": 0.1178, + "step": 2360 + }, + { + "epoch": 2.3238188976377954, + "grad_norm": 5.486789226531982, + "learning_rate": 2.5872247388637e-07, + "loss": 0.1065, + "step": 2361 + }, + { + "epoch": 2.324803149606299, + "grad_norm": 12.062655448913574, + "learning_rate": 2.5805336158042327e-07, + "loss": 0.5725, + "step": 2362 + }, + { + "epoch": 2.3257874015748032, + "grad_norm": 9.281482696533203, + "learning_rate": 2.573848258937945e-07, + "loss": 0.2564, + "step": 2363 + }, + { + "epoch": 2.326771653543307, + "grad_norm": 5.837134838104248, + "learning_rate": 2.56716868408871e-07, + "loss": 0.1425, + "step": 2364 + }, + { + "epoch": 2.327755905511811, + "grad_norm": 10.866689682006836, + "learning_rate": 2.5604949070667156e-07, + "loss": 0.4065, + "step": 2365 + }, + { + "epoch": 2.328740157480315, + "grad_norm": 8.223211288452148, + "learning_rate": 2.553826943668423e-07, + "loss": 0.1472, + "step": 2366 + }, + { + "epoch": 2.329724409448819, + "grad_norm": 11.248306274414062, + "learning_rate": 2.5471648096765365e-07, + "loss": 0.3658, + "step": 2367 + }, + { + "epoch": 2.3307086614173227, + "grad_norm": 13.366337776184082, + "learning_rate": 2.5405085208599623e-07, + "loss": 0.4533, + "step": 2368 + }, + { + "epoch": 2.331692913385827, + "grad_norm": 6.944033145904541, + "learning_rate": 2.53385809297377e-07, + "loss": 0.2669, + "step": 2369 + }, + { + "epoch": 2.3326771653543306, + "grad_norm": 11.905457496643066, + "learning_rate": 2.527213541759156e-07, + "loss": 0.4721, + "step": 2370 + }, + { + "epoch": 2.3336614173228347, + "grad_norm": 10.900627136230469, + "learning_rate": 2.520574882943411e-07, + "loss": 0.5703, + "step": 2371 + }, + { + "epoch": 2.3346456692913384, + "grad_norm": 7.861691474914551, + "learning_rate": 2.5139421322398734e-07, + "loss": 0.2449, + "step": 2372 + }, + { + "epoch": 2.3356299212598426, + "grad_norm": 10.354506492614746, + "learning_rate": 2.5073153053479025e-07, + "loss": 0.3974, + "step": 2373 + }, + { + "epoch": 2.3366141732283463, + "grad_norm": 8.82080364227295, + "learning_rate": 2.5006944179528326e-07, + "loss": 0.2303, + "step": 2374 + }, + { + "epoch": 2.3375984251968505, + "grad_norm": 13.145618438720703, + "learning_rate": 2.494079485725939e-07, + "loss": 0.542, + "step": 2375 + }, + { + "epoch": 2.338582677165354, + "grad_norm": 7.5811543464660645, + "learning_rate": 2.4874705243244044e-07, + "loss": 0.3991, + "step": 2376 + }, + { + "epoch": 2.3395669291338583, + "grad_norm": 10.05690860748291, + "learning_rate": 2.480867549391278e-07, + "loss": 0.4184, + "step": 2377 + }, + { + "epoch": 2.340551181102362, + "grad_norm": 10.880393981933594, + "learning_rate": 2.4742705765554384e-07, + "loss": 0.3636, + "step": 2378 + }, + { + "epoch": 2.341535433070866, + "grad_norm": 12.805682182312012, + "learning_rate": 2.467679621431558e-07, + "loss": 0.5021, + "step": 2379 + }, + { + "epoch": 2.34251968503937, + "grad_norm": 6.25147819519043, + "learning_rate": 2.461094699620066e-07, + "loss": 0.1835, + "step": 2380 + }, + { + "epoch": 2.343503937007874, + "grad_norm": 10.553853034973145, + "learning_rate": 2.454515826707111e-07, + "loss": 0.3925, + "step": 2381 + }, + { + "epoch": 2.344488188976378, + "grad_norm": 9.985482215881348, + "learning_rate": 2.4479430182645237e-07, + "loss": 0.4167, + "step": 2382 + }, + { + "epoch": 2.345472440944882, + "grad_norm": 6.239148139953613, + "learning_rate": 2.441376289849781e-07, + "loss": 0.2309, + "step": 2383 + }, + { + "epoch": 2.3464566929133857, + "grad_norm": 16.266908645629883, + "learning_rate": 2.4348156570059685e-07, + "loss": 0.7152, + "step": 2384 + }, + { + "epoch": 2.34744094488189, + "grad_norm": 12.050719261169434, + "learning_rate": 2.428261135261743e-07, + "loss": 0.313, + "step": 2385 + }, + { + "epoch": 2.3484251968503935, + "grad_norm": 14.419236183166504, + "learning_rate": 2.4217127401312993e-07, + "loss": 0.6408, + "step": 2386 + }, + { + "epoch": 2.3494094488188977, + "grad_norm": 9.486001014709473, + "learning_rate": 2.4151704871143276e-07, + "loss": 0.4033, + "step": 2387 + }, + { + "epoch": 2.3503937007874014, + "grad_norm": 11.379007339477539, + "learning_rate": 2.4086343916959827e-07, + "loss": 0.545, + "step": 2388 + }, + { + "epoch": 2.3513779527559056, + "grad_norm": 4.567531585693359, + "learning_rate": 2.402104469346844e-07, + "loss": 0.0758, + "step": 2389 + }, + { + "epoch": 2.3523622047244093, + "grad_norm": 7.687051296234131, + "learning_rate": 2.3955807355228767e-07, + "loss": 0.361, + "step": 2390 + }, + { + "epoch": 2.3533464566929134, + "grad_norm": 11.952143669128418, + "learning_rate": 2.3890632056654027e-07, + "loss": 0.3222, + "step": 2391 + }, + { + "epoch": 2.354330708661417, + "grad_norm": 7.990842342376709, + "learning_rate": 2.3825518952010577e-07, + "loss": 0.2374, + "step": 2392 + }, + { + "epoch": 2.3553149606299213, + "grad_norm": 11.30343246459961, + "learning_rate": 2.3760468195417546e-07, + "loss": 0.3697, + "step": 2393 + }, + { + "epoch": 2.356299212598425, + "grad_norm": 11.191856384277344, + "learning_rate": 2.3695479940846513e-07, + "loss": 0.3201, + "step": 2394 + }, + { + "epoch": 2.357283464566929, + "grad_norm": 13.76095199584961, + "learning_rate": 2.363055434212112e-07, + "loss": 0.4985, + "step": 2395 + }, + { + "epoch": 2.358267716535433, + "grad_norm": 9.870569229125977, + "learning_rate": 2.3565691552916664e-07, + "loss": 0.4632, + "step": 2396 + }, + { + "epoch": 2.359251968503937, + "grad_norm": 7.816884994506836, + "learning_rate": 2.3500891726759822e-07, + "loss": 0.184, + "step": 2397 + }, + { + "epoch": 2.3602362204724407, + "grad_norm": 16.132492065429688, + "learning_rate": 2.3436155017028224e-07, + "loss": 0.8459, + "step": 2398 + }, + { + "epoch": 2.361220472440945, + "grad_norm": 8.283998489379883, + "learning_rate": 2.3371481576950107e-07, + "loss": 0.2389, + "step": 2399 + }, + { + "epoch": 2.362204724409449, + "grad_norm": 15.900023460388184, + "learning_rate": 2.3306871559603962e-07, + "loss": 0.6347, + "step": 2400 + }, + { + "epoch": 2.3631889763779528, + "grad_norm": 17.185317993164062, + "learning_rate": 2.324232511791814e-07, + "loss": 0.8377, + "step": 2401 + }, + { + "epoch": 2.3641732283464565, + "grad_norm": 13.698606491088867, + "learning_rate": 2.3177842404670548e-07, + "loss": 0.4673, + "step": 2402 + }, + { + "epoch": 2.3651574803149606, + "grad_norm": 13.84771728515625, + "learning_rate": 2.3113423572488192e-07, + "loss": 0.6065, + "step": 2403 + }, + { + "epoch": 2.366141732283465, + "grad_norm": 11.880446434020996, + "learning_rate": 2.304906877384692e-07, + "loss": 0.6109, + "step": 2404 + }, + { + "epoch": 2.3671259842519685, + "grad_norm": 12.377808570861816, + "learning_rate": 2.298477816107103e-07, + "loss": 0.4965, + "step": 2405 + }, + { + "epoch": 2.3681102362204722, + "grad_norm": 9.263788223266602, + "learning_rate": 2.292055188633285e-07, + "loss": 0.4042, + "step": 2406 + }, + { + "epoch": 2.3690944881889764, + "grad_norm": 5.993063926696777, + "learning_rate": 2.2856390101652452e-07, + "loss": 0.2272, + "step": 2407 + }, + { + "epoch": 2.3700787401574805, + "grad_norm": 3.624645709991455, + "learning_rate": 2.2792292958897266e-07, + "loss": 0.1061, + "step": 2408 + }, + { + "epoch": 2.3710629921259843, + "grad_norm": 6.131748676300049, + "learning_rate": 2.2728260609781702e-07, + "loss": 0.251, + "step": 2409 + }, + { + "epoch": 2.372047244094488, + "grad_norm": 9.094381332397461, + "learning_rate": 2.2664293205866817e-07, + "loss": 0.3711, + "step": 2410 + }, + { + "epoch": 2.373031496062992, + "grad_norm": 12.853039741516113, + "learning_rate": 2.2600390898559944e-07, + "loss": 0.4646, + "step": 2411 + }, + { + "epoch": 2.3740157480314963, + "grad_norm": 14.42044448852539, + "learning_rate": 2.2536553839114338e-07, + "loss": 0.7357, + "step": 2412 + }, + { + "epoch": 2.375, + "grad_norm": 10.636192321777344, + "learning_rate": 2.2472782178628807e-07, + "loss": 0.3643, + "step": 2413 + }, + { + "epoch": 2.3759842519685037, + "grad_norm": 8.486954689025879, + "learning_rate": 2.2409076068047373e-07, + "loss": 0.2327, + "step": 2414 + }, + { + "epoch": 2.376968503937008, + "grad_norm": 7.130517482757568, + "learning_rate": 2.2345435658158908e-07, + "loss": 0.1543, + "step": 2415 + }, + { + "epoch": 2.377952755905512, + "grad_norm": 10.911663055419922, + "learning_rate": 2.2281861099596783e-07, + "loss": 0.3892, + "step": 2416 + }, + { + "epoch": 2.3789370078740157, + "grad_norm": 10.254852294921875, + "learning_rate": 2.2218352542838447e-07, + "loss": 0.2659, + "step": 2417 + }, + { + "epoch": 2.3799212598425195, + "grad_norm": 12.970120429992676, + "learning_rate": 2.215491013820519e-07, + "loss": 0.5829, + "step": 2418 + }, + { + "epoch": 2.3809055118110236, + "grad_norm": 8.290464401245117, + "learning_rate": 2.2091534035861703e-07, + "loss": 0.2263, + "step": 2419 + }, + { + "epoch": 2.3818897637795278, + "grad_norm": 12.914335250854492, + "learning_rate": 2.2028224385815726e-07, + "loss": 0.4357, + "step": 2420 + }, + { + "epoch": 2.3828740157480315, + "grad_norm": 9.889439582824707, + "learning_rate": 2.1964981337917744e-07, + "loss": 0.3646, + "step": 2421 + }, + { + "epoch": 2.383858267716535, + "grad_norm": 11.510004997253418, + "learning_rate": 2.1901805041860567e-07, + "loss": 0.3156, + "step": 2422 + }, + { + "epoch": 2.3848425196850394, + "grad_norm": 9.496753692626953, + "learning_rate": 2.1838695647179024e-07, + "loss": 0.3001, + "step": 2423 + }, + { + "epoch": 2.3858267716535435, + "grad_norm": 10.354239463806152, + "learning_rate": 2.1775653303249587e-07, + "loss": 0.2207, + "step": 2424 + }, + { + "epoch": 2.3868110236220472, + "grad_norm": 6.860008239746094, + "learning_rate": 2.1712678159290023e-07, + "loss": 0.201, + "step": 2425 + }, + { + "epoch": 2.387795275590551, + "grad_norm": 12.281497955322266, + "learning_rate": 2.164977036435904e-07, + "loss": 0.5521, + "step": 2426 + }, + { + "epoch": 2.388779527559055, + "grad_norm": 8.849814414978027, + "learning_rate": 2.1586930067355932e-07, + "loss": 0.3396, + "step": 2427 + }, + { + "epoch": 2.3897637795275593, + "grad_norm": 10.87320327758789, + "learning_rate": 2.152415741702023e-07, + "loss": 0.3528, + "step": 2428 + }, + { + "epoch": 2.390748031496063, + "grad_norm": 13.564628601074219, + "learning_rate": 2.146145256193136e-07, + "loss": 0.4771, + "step": 2429 + }, + { + "epoch": 2.391732283464567, + "grad_norm": 7.737844467163086, + "learning_rate": 2.1398815650508263e-07, + "loss": 0.2577, + "step": 2430 + }, + { + "epoch": 2.392716535433071, + "grad_norm": 3.8913357257843018, + "learning_rate": 2.1336246831009075e-07, + "loss": 0.0595, + "step": 2431 + }, + { + "epoch": 2.393700787401575, + "grad_norm": 10.990683555603027, + "learning_rate": 2.1273746251530752e-07, + "loss": 0.5141, + "step": 2432 + }, + { + "epoch": 2.3946850393700787, + "grad_norm": 11.935064315795898, + "learning_rate": 2.1211314060008755e-07, + "loss": 0.3178, + "step": 2433 + }, + { + "epoch": 2.395669291338583, + "grad_norm": 4.393062114715576, + "learning_rate": 2.1148950404216617e-07, + "loss": 0.0863, + "step": 2434 + }, + { + "epoch": 2.3966535433070866, + "grad_norm": 9.64417552947998, + "learning_rate": 2.10866554317657e-07, + "loss": 0.2858, + "step": 2435 + }, + { + "epoch": 2.3976377952755907, + "grad_norm": 13.084728240966797, + "learning_rate": 2.1024429290104796e-07, + "loss": 0.6671, + "step": 2436 + }, + { + "epoch": 2.3986220472440944, + "grad_norm": 7.769230842590332, + "learning_rate": 2.0962272126519756e-07, + "loss": 0.1614, + "step": 2437 + }, + { + "epoch": 2.3996062992125986, + "grad_norm": 9.682682037353516, + "learning_rate": 2.0900184088133173e-07, + "loss": 0.4314, + "step": 2438 + }, + { + "epoch": 2.4005905511811023, + "grad_norm": 8.374194145202637, + "learning_rate": 2.0838165321904045e-07, + "loss": 0.1796, + "step": 2439 + }, + { + "epoch": 2.4015748031496065, + "grad_norm": 6.3454179763793945, + "learning_rate": 2.0776215974627359e-07, + "loss": 0.1504, + "step": 2440 + }, + { + "epoch": 2.40255905511811, + "grad_norm": 11.132549285888672, + "learning_rate": 2.0714336192933824e-07, + "loss": 0.352, + "step": 2441 + }, + { + "epoch": 2.4035433070866143, + "grad_norm": 10.09714126586914, + "learning_rate": 2.0652526123289488e-07, + "loss": 0.2414, + "step": 2442 + }, + { + "epoch": 2.404527559055118, + "grad_norm": 4.490203857421875, + "learning_rate": 2.0590785911995406e-07, + "loss": 0.0617, + "step": 2443 + }, + { + "epoch": 2.405511811023622, + "grad_norm": 11.918438911437988, + "learning_rate": 2.052911570518725e-07, + "loss": 0.5506, + "step": 2444 + }, + { + "epoch": 2.406496062992126, + "grad_norm": 5.9415669441223145, + "learning_rate": 2.046751564883503e-07, + "loss": 0.1015, + "step": 2445 + }, + { + "epoch": 2.40748031496063, + "grad_norm": 7.4299635887146, + "learning_rate": 2.040598588874268e-07, + "loss": 0.1732, + "step": 2446 + }, + { + "epoch": 2.408464566929134, + "grad_norm": 7.314566135406494, + "learning_rate": 2.0344526570547774e-07, + "loss": 0.1889, + "step": 2447 + }, + { + "epoch": 2.409448818897638, + "grad_norm": 13.567390441894531, + "learning_rate": 2.0283137839721143e-07, + "loss": 0.3959, + "step": 2448 + }, + { + "epoch": 2.409448818897638, + "eval_Qnli-dev_cosine_accuracy": 0.693359375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7806241512298584, + "eval_Qnli-dev_cosine_ap": 0.7038595672714784, + "eval_Qnli-dev_cosine_f1": 0.708994708994709, + "eval_Qnli-dev_cosine_f1_threshold": 0.7480103969573975, + "eval_Qnli-dev_cosine_precision": 0.6072507552870091, + "eval_Qnli-dev_cosine_recall": 0.8516949152542372, + "eval_Qnli-dev_dot_accuracy": 0.69140625, + "eval_Qnli-dev_dot_accuracy_threshold": 596.2394409179688, + "eval_Qnli-dev_dot_ap": 0.7043489524595209, + "eval_Qnli-dev_dot_f1": 0.708994708994709, + "eval_Qnli-dev_dot_f1_threshold": 571.1837768554688, + "eval_Qnli-dev_dot_precision": 0.6072507552870091, + "eval_Qnli-dev_dot_recall": 0.8516949152542372, + "eval_Qnli-dev_euclidean_accuracy": 0.693359375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 18.30592155456543, + "eval_Qnli-dev_euclidean_ap": 0.7038272341564227, + "eval_Qnli-dev_euclidean_f1": 0.7065026362038664, + "eval_Qnli-dev_euclidean_f1_threshold": 19.632598876953125, + "eval_Qnli-dev_euclidean_precision": 0.6036036036036037, + "eval_Qnli-dev_euclidean_recall": 0.8516949152542372, + "eval_Qnli-dev_manhattan_accuracy": 0.6875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 404.1275634765625, + "eval_Qnli-dev_manhattan_ap": 0.7042847168068205, + "eval_Qnli-dev_manhattan_f1": 0.7033747779751333, + "eval_Qnli-dev_manhattan_f1_threshold": 433.2981872558594, + "eval_Qnli-dev_manhattan_precision": 0.6055045871559633, + "eval_Qnli-dev_manhattan_recall": 0.8389830508474576, + "eval_Qnli-dev_max_accuracy": 0.693359375, + "eval_Qnli-dev_max_accuracy_threshold": 596.2394409179688, + "eval_Qnli-dev_max_ap": 0.7043489524595209, + "eval_Qnli-dev_max_f1": 0.708994708994709, + "eval_Qnli-dev_max_f1_threshold": 571.1837768554688, + "eval_Qnli-dev_max_precision": 0.6072507552870091, + "eval_Qnli-dev_max_recall": 0.8516949152542372, + "eval_allNLI-dev_cosine_accuracy": 0.693359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.8861348032951355, + "eval_allNLI-dev_cosine_ap": 0.5354190469375049, + "eval_allNLI-dev_cosine_f1": 0.5840336134453781, + "eval_allNLI-dev_cosine_f1_threshold": 0.8200911283493042, + "eval_allNLI-dev_cosine_precision": 0.45874587458745875, + "eval_allNLI-dev_cosine_recall": 0.8034682080924855, + "eval_allNLI-dev_dot_accuracy": 0.69140625, + "eval_allNLI-dev_dot_accuracy_threshold": 700.5333251953125, + "eval_allNLI-dev_dot_ap": 0.5350174558876482, + "eval_allNLI-dev_dot_f1": 0.5840336134453781, + "eval_allNLI-dev_dot_f1_threshold": 626.7640380859375, + "eval_allNLI-dev_dot_precision": 0.45874587458745875, + "eval_allNLI-dev_dot_recall": 0.8034682080924855, + "eval_allNLI-dev_euclidean_accuracy": 0.693359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 13.186843872070312, + "eval_allNLI-dev_euclidean_ap": 0.5355646710263617, + "eval_allNLI-dev_euclidean_f1": 0.5840336134453781, + "eval_allNLI-dev_euclidean_f1_threshold": 16.584110260009766, + "eval_allNLI-dev_euclidean_precision": 0.45874587458745875, + "eval_allNLI-dev_euclidean_recall": 0.8034682080924855, + "eval_allNLI-dev_manhattan_accuracy": 0.693359375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 285.8642883300781, + "eval_allNLI-dev_manhattan_ap": 0.5370566276463048, + "eval_allNLI-dev_manhattan_f1": 0.5864978902953586, + "eval_allNLI-dev_manhattan_f1_threshold": 363.96441650390625, + "eval_allNLI-dev_manhattan_precision": 0.46179401993355484, + "eval_allNLI-dev_manhattan_recall": 0.8034682080924855, + "eval_allNLI-dev_max_accuracy": 0.693359375, + "eval_allNLI-dev_max_accuracy_threshold": 700.5333251953125, + "eval_allNLI-dev_max_ap": 0.5370566276463048, + "eval_allNLI-dev_max_f1": 0.5864978902953586, + "eval_allNLI-dev_max_f1_threshold": 626.7640380859375, + "eval_allNLI-dev_max_precision": 0.46179401993355484, + "eval_allNLI-dev_max_recall": 0.8034682080924855, + "eval_loss": 0.733245849609375, + "eval_runtime": 49.9523, + "eval_samples_per_second": 33.312, + "eval_sequential_score": 0.7043489524595209, + "eval_steps_per_second": 0.14, + "eval_sts-test_pearson_cosine": 0.6951996224089858, + "eval_sts-test_pearson_dot": 0.6949603001699672, + "eval_sts-test_pearson_euclidean": 0.7194989281187023, + "eval_sts-test_pearson_manhattan": 0.7204021387472773, + "eval_sts-test_pearson_max": 0.7204021387472773, + "eval_sts-test_spearman_cosine": 0.7103732443437184, + "eval_sts-test_spearman_dot": 0.7097478007145122, + "eval_sts-test_spearman_euclidean": 0.7102974357069214, + "eval_sts-test_spearman_manhattan": 0.7108715943905731, + "eval_sts-test_spearman_max": 0.7108715943905731, + "step": 2448 + }, + { + "epoch": 2.4104330708661417, + "grad_norm": 14.817706108093262, + "learning_rate": 2.022181984156654e-07, + "loss": 0.5107, + "step": 2449 + }, + { + "epoch": 2.411417322834646, + "grad_norm": 11.785439491271973, + "learning_rate": 2.0160572721220312e-07, + "loss": 0.3173, + "step": 2450 + }, + { + "epoch": 2.4124015748031495, + "grad_norm": 7.95774507522583, + "learning_rate": 2.0099396623651012e-07, + "loss": 0.2245, + "step": 2451 + }, + { + "epoch": 2.4133858267716537, + "grad_norm": 10.812651634216309, + "learning_rate": 2.0038291693659113e-07, + "loss": 0.2664, + "step": 2452 + }, + { + "epoch": 2.4143700787401574, + "grad_norm": 9.724847793579102, + "learning_rate": 1.997725807587663e-07, + "loss": 0.2888, + "step": 2453 + }, + { + "epoch": 2.4153543307086616, + "grad_norm": 12.10487174987793, + "learning_rate": 1.991629591476679e-07, + "loss": 0.3439, + "step": 2454 + }, + { + "epoch": 2.4163385826771653, + "grad_norm": 12.252102851867676, + "learning_rate": 1.9855405354623664e-07, + "loss": 0.261, + "step": 2455 + }, + { + "epoch": 2.4173228346456694, + "grad_norm": 11.291241645812988, + "learning_rate": 1.9794586539571887e-07, + "loss": 0.511, + "step": 2456 + }, + { + "epoch": 2.418307086614173, + "grad_norm": 8.785077095031738, + "learning_rate": 1.9733839613566244e-07, + "loss": 0.1052, + "step": 2457 + }, + { + "epoch": 2.4192913385826773, + "grad_norm": 5.6941609382629395, + "learning_rate": 1.9673164720391387e-07, + "loss": 0.1136, + "step": 2458 + }, + { + "epoch": 2.420275590551181, + "grad_norm": 7.104002952575684, + "learning_rate": 1.9612562003661423e-07, + "loss": 0.2537, + "step": 2459 + }, + { + "epoch": 2.421259842519685, + "grad_norm": 6.955623626708984, + "learning_rate": 1.9552031606819685e-07, + "loss": 0.1049, + "step": 2460 + }, + { + "epoch": 2.422244094488189, + "grad_norm": 10.139701843261719, + "learning_rate": 1.9491573673138287e-07, + "loss": 0.2425, + "step": 2461 + }, + { + "epoch": 2.423228346456693, + "grad_norm": 7.826995372772217, + "learning_rate": 1.9431188345717844e-07, + "loss": 0.1777, + "step": 2462 + }, + { + "epoch": 2.4242125984251968, + "grad_norm": 13.749302864074707, + "learning_rate": 1.9370875767487104e-07, + "loss": 0.4612, + "step": 2463 + }, + { + "epoch": 2.425196850393701, + "grad_norm": 13.596500396728516, + "learning_rate": 1.9310636081202636e-07, + "loss": 0.5801, + "step": 2464 + }, + { + "epoch": 2.4261811023622046, + "grad_norm": 14.912703514099121, + "learning_rate": 1.925046942944847e-07, + "loss": 0.6488, + "step": 2465 + }, + { + "epoch": 2.427165354330709, + "grad_norm": 10.965988159179688, + "learning_rate": 1.919037595463576e-07, + "loss": 0.2711, + "step": 2466 + }, + { + "epoch": 2.4281496062992125, + "grad_norm": 6.907632350921631, + "learning_rate": 1.9130355799002467e-07, + "loss": 0.1004, + "step": 2467 + }, + { + "epoch": 2.4291338582677167, + "grad_norm": 9.669986724853516, + "learning_rate": 1.9070409104613e-07, + "loss": 0.379, + "step": 2468 + }, + { + "epoch": 2.4301181102362204, + "grad_norm": 12.249451637268066, + "learning_rate": 1.90105360133579e-07, + "loss": 0.637, + "step": 2469 + }, + { + "epoch": 2.4311023622047245, + "grad_norm": 10.606658935546875, + "learning_rate": 1.89507366669535e-07, + "loss": 0.2704, + "step": 2470 + }, + { + "epoch": 2.4320866141732282, + "grad_norm": 10.840896606445312, + "learning_rate": 1.8891011206941535e-07, + "loss": 0.3991, + "step": 2471 + }, + { + "epoch": 2.4330708661417324, + "grad_norm": 12.712931632995605, + "learning_rate": 1.883135977468891e-07, + "loss": 0.5034, + "step": 2472 + }, + { + "epoch": 2.434055118110236, + "grad_norm": 10.296034812927246, + "learning_rate": 1.8771782511387285e-07, + "loss": 0.3201, + "step": 2473 + }, + { + "epoch": 2.4350393700787403, + "grad_norm": 6.260189533233643, + "learning_rate": 1.8712279558052777e-07, + "loss": 0.1225, + "step": 2474 + }, + { + "epoch": 2.436023622047244, + "grad_norm": 4.607402324676514, + "learning_rate": 1.8652851055525605e-07, + "loss": 0.0579, + "step": 2475 + }, + { + "epoch": 2.437007874015748, + "grad_norm": 11.457910537719727, + "learning_rate": 1.859349714446977e-07, + "loss": 0.3999, + "step": 2476 + }, + { + "epoch": 2.437992125984252, + "grad_norm": 10.963780403137207, + "learning_rate": 1.853421796537273e-07, + "loss": 0.4022, + "step": 2477 + }, + { + "epoch": 2.438976377952756, + "grad_norm": 4.134235382080078, + "learning_rate": 1.8475013658545054e-07, + "loss": 0.0807, + "step": 2478 + }, + { + "epoch": 2.4399606299212597, + "grad_norm": 2.986137866973877, + "learning_rate": 1.841588436412006e-07, + "loss": 0.0458, + "step": 2479 + }, + { + "epoch": 2.440944881889764, + "grad_norm": 7.944431781768799, + "learning_rate": 1.8356830222053546e-07, + "loss": 0.1684, + "step": 2480 + }, + { + "epoch": 2.4419291338582676, + "grad_norm": 7.481739521026611, + "learning_rate": 1.8297851372123444e-07, + "loss": 0.1653, + "step": 2481 + }, + { + "epoch": 2.4429133858267718, + "grad_norm": 12.850192070007324, + "learning_rate": 1.8238947953929436e-07, + "loss": 0.5445, + "step": 2482 + }, + { + "epoch": 2.4438976377952755, + "grad_norm": 8.56425666809082, + "learning_rate": 1.8180120106892685e-07, + "loss": 0.2467, + "step": 2483 + }, + { + "epoch": 2.4448818897637796, + "grad_norm": 9.05689525604248, + "learning_rate": 1.8121367970255475e-07, + "loss": 0.2696, + "step": 2484 + }, + { + "epoch": 2.4458661417322833, + "grad_norm": 11.202281951904297, + "learning_rate": 1.806269168308089e-07, + "loss": 0.3576, + "step": 2485 + }, + { + "epoch": 2.4468503937007875, + "grad_norm": 10.96251106262207, + "learning_rate": 1.800409138425248e-07, + "loss": 0.3456, + "step": 2486 + }, + { + "epoch": 2.447834645669291, + "grad_norm": 10.072763442993164, + "learning_rate": 1.794556721247393e-07, + "loss": 0.2766, + "step": 2487 + }, + { + "epoch": 2.4488188976377954, + "grad_norm": 9.944024085998535, + "learning_rate": 1.7887119306268743e-07, + "loss": 0.2978, + "step": 2488 + }, + { + "epoch": 2.449803149606299, + "grad_norm": 8.820456504821777, + "learning_rate": 1.7828747803979914e-07, + "loss": 0.1997, + "step": 2489 + }, + { + "epoch": 2.4507874015748032, + "grad_norm": 8.660542488098145, + "learning_rate": 1.7770452843769567e-07, + "loss": 0.2677, + "step": 2490 + }, + { + "epoch": 2.451771653543307, + "grad_norm": 8.645709037780762, + "learning_rate": 1.771223456361868e-07, + "loss": 0.2722, + "step": 2491 + }, + { + "epoch": 2.452755905511811, + "grad_norm": 6.167365074157715, + "learning_rate": 1.765409310132673e-07, + "loss": 0.1399, + "step": 2492 + }, + { + "epoch": 2.453740157480315, + "grad_norm": 18.60035514831543, + "learning_rate": 1.7596028594511364e-07, + "loss": 0.5965, + "step": 2493 + }, + { + "epoch": 2.454724409448819, + "grad_norm": 10.543310165405273, + "learning_rate": 1.753804118060808e-07, + "loss": 0.3029, + "step": 2494 + }, + { + "epoch": 2.4557086614173227, + "grad_norm": 8.006535530090332, + "learning_rate": 1.7480130996869905e-07, + "loss": 0.2144, + "step": 2495 + }, + { + "epoch": 2.456692913385827, + "grad_norm": 8.3842191696167, + "learning_rate": 1.7422298180367063e-07, + "loss": 0.216, + "step": 2496 + }, + { + "epoch": 2.4576771653543306, + "grad_norm": 7.863168239593506, + "learning_rate": 1.736454286798665e-07, + "loss": 0.1844, + "step": 2497 + }, + { + "epoch": 2.4586614173228347, + "grad_norm": 10.605033874511719, + "learning_rate": 1.730686519643233e-07, + "loss": 0.3645, + "step": 2498 + }, + { + "epoch": 2.4596456692913384, + "grad_norm": 4.173581123352051, + "learning_rate": 1.7249265302223998e-07, + "loss": 0.0918, + "step": 2499 + }, + { + "epoch": 2.4606299212598426, + "grad_norm": 11.259767532348633, + "learning_rate": 1.7191743321697403e-07, + "loss": 0.5575, + "step": 2500 + }, + { + "epoch": 2.4616141732283463, + "grad_norm": 4.349879264831543, + "learning_rate": 1.713429939100393e-07, + "loss": 0.0907, + "step": 2501 + }, + { + "epoch": 2.4625984251968505, + "grad_norm": 7.431346893310547, + "learning_rate": 1.7076933646110222e-07, + "loss": 0.1534, + "step": 2502 + }, + { + "epoch": 2.463582677165354, + "grad_norm": 9.7450590133667, + "learning_rate": 1.7019646222797834e-07, + "loss": 0.2732, + "step": 2503 + }, + { + "epoch": 2.4645669291338583, + "grad_norm": 10.753069877624512, + "learning_rate": 1.696243725666296e-07, + "loss": 0.3203, + "step": 2504 + }, + { + "epoch": 2.465551181102362, + "grad_norm": 10.13337230682373, + "learning_rate": 1.6905306883116077e-07, + "loss": 0.3072, + "step": 2505 + }, + { + "epoch": 2.466535433070866, + "grad_norm": 7.750153541564941, + "learning_rate": 1.684825523738165e-07, + "loss": 0.2088, + "step": 2506 + }, + { + "epoch": 2.46751968503937, + "grad_norm": 8.731947898864746, + "learning_rate": 1.67912824544978e-07, + "loss": 0.2522, + "step": 2507 + }, + { + "epoch": 2.468503937007874, + "grad_norm": 7.932433128356934, + "learning_rate": 1.6734388669315944e-07, + "loss": 0.2774, + "step": 2508 + }, + { + "epoch": 2.469488188976378, + "grad_norm": 8.18900203704834, + "learning_rate": 1.667757401650057e-07, + "loss": 0.1527, + "step": 2509 + }, + { + "epoch": 2.470472440944882, + "grad_norm": 6.190526485443115, + "learning_rate": 1.6620838630528833e-07, + "loss": 0.1992, + "step": 2510 + }, + { + "epoch": 2.4714566929133857, + "grad_norm": 10.24229621887207, + "learning_rate": 1.6564182645690283e-07, + "loss": 0.2822, + "step": 2511 + }, + { + "epoch": 2.47244094488189, + "grad_norm": 9.039300918579102, + "learning_rate": 1.650760619608653e-07, + "loss": 0.2217, + "step": 2512 + }, + { + "epoch": 2.4734251968503935, + "grad_norm": 9.909263610839844, + "learning_rate": 1.645110941563092e-07, + "loss": 0.2587, + "step": 2513 + }, + { + "epoch": 2.4744094488188977, + "grad_norm": 14.455375671386719, + "learning_rate": 1.639469243804822e-07, + "loss": 0.4147, + "step": 2514 + }, + { + "epoch": 2.4753937007874014, + "grad_norm": 1.4771108627319336, + "learning_rate": 1.6338355396874332e-07, + "loss": 0.0197, + "step": 2515 + }, + { + "epoch": 2.4763779527559056, + "grad_norm": 9.873431205749512, + "learning_rate": 1.6282098425455936e-07, + "loss": 0.3847, + "step": 2516 + }, + { + "epoch": 2.4773622047244093, + "grad_norm": 2.7959773540496826, + "learning_rate": 1.622592165695019e-07, + "loss": 0.0507, + "step": 2517 + }, + { + "epoch": 2.4783464566929134, + "grad_norm": 14.620316505432129, + "learning_rate": 1.616982522432443e-07, + "loss": 0.4266, + "step": 2518 + }, + { + "epoch": 2.479330708661417, + "grad_norm": 13.626422882080078, + "learning_rate": 1.6113809260355824e-07, + "loss": 0.5208, + "step": 2519 + }, + { + "epoch": 2.4803149606299213, + "grad_norm": 8.045567512512207, + "learning_rate": 1.6057873897631103e-07, + "loss": 0.2054, + "step": 2520 + }, + { + "epoch": 2.481299212598425, + "grad_norm": 11.742626190185547, + "learning_rate": 1.600201926854616e-07, + "loss": 0.3809, + "step": 2521 + }, + { + "epoch": 2.482283464566929, + "grad_norm": 3.622823715209961, + "learning_rate": 1.5946245505305863e-07, + "loss": 0.0522, + "step": 2522 + }, + { + "epoch": 2.483267716535433, + "grad_norm": 7.850180149078369, + "learning_rate": 1.5890552739923647e-07, + "loss": 0.1534, + "step": 2523 + }, + { + "epoch": 2.484251968503937, + "grad_norm": 10.612723350524902, + "learning_rate": 1.5834941104221217e-07, + "loss": 0.4365, + "step": 2524 + }, + { + "epoch": 2.4852362204724407, + "grad_norm": 10.891499519348145, + "learning_rate": 1.577941072982827e-07, + "loss": 0.5218, + "step": 2525 + }, + { + "epoch": 2.486220472440945, + "grad_norm": 9.140625, + "learning_rate": 1.5723961748182163e-07, + "loss": 0.2824, + "step": 2526 + }, + { + "epoch": 2.487204724409449, + "grad_norm": 9.8253173828125, + "learning_rate": 1.5668594290527568e-07, + "loss": 0.4253, + "step": 2527 + }, + { + "epoch": 2.4881889763779528, + "grad_norm": 7.857529640197754, + "learning_rate": 1.5613308487916242e-07, + "loss": 0.2438, + "step": 2528 + }, + { + "epoch": 2.4891732283464565, + "grad_norm": 7.818708896636963, + "learning_rate": 1.5558104471206626e-07, + "loss": 0.1249, + "step": 2529 + }, + { + "epoch": 2.4901574803149606, + "grad_norm": 18.685569763183594, + "learning_rate": 1.5502982371063602e-07, + "loss": 0.7602, + "step": 2530 + }, + { + "epoch": 2.491141732283465, + "grad_norm": 7.651521682739258, + "learning_rate": 1.5447942317958161e-07, + "loss": 0.1701, + "step": 2531 + }, + { + "epoch": 2.4921259842519685, + "grad_norm": 12.079002380371094, + "learning_rate": 1.539298444216708e-07, + "loss": 0.3371, + "step": 2532 + }, + { + "epoch": 2.4931102362204722, + "grad_norm": 13.053180694580078, + "learning_rate": 1.5338108873772638e-07, + "loss": 0.6008, + "step": 2533 + }, + { + "epoch": 2.4940944881889764, + "grad_norm": 13.562995910644531, + "learning_rate": 1.5283315742662288e-07, + "loss": 0.3341, + "step": 2534 + }, + { + "epoch": 2.4950787401574805, + "grad_norm": 8.421311378479004, + "learning_rate": 1.5228605178528368e-07, + "loss": 0.23, + "step": 2535 + }, + { + "epoch": 2.4960629921259843, + "grad_norm": 4.978429794311523, + "learning_rate": 1.5173977310867775e-07, + "loss": 0.0932, + "step": 2536 + }, + { + "epoch": 2.497047244094488, + "grad_norm": 14.153863906860352, + "learning_rate": 1.5119432268981675e-07, + "loss": 0.474, + "step": 2537 + }, + { + "epoch": 2.498031496062992, + "grad_norm": 8.590943336486816, + "learning_rate": 1.5064970181975204e-07, + "loss": 0.2473, + "step": 2538 + }, + { + "epoch": 2.4990157480314963, + "grad_norm": 6.933433532714844, + "learning_rate": 1.5010591178757092e-07, + "loss": 0.3215, + "step": 2539 + }, + { + "epoch": 2.5, + "grad_norm": 11.423782348632812, + "learning_rate": 1.4956295388039484e-07, + "loss": 0.3671, + "step": 2540 + }, + { + "epoch": 2.5009842519685037, + "grad_norm": 9.358104705810547, + "learning_rate": 1.4902082938337527e-07, + "loss": 0.2851, + "step": 2541 + }, + { + "epoch": 2.501968503937008, + "grad_norm": 8.418267250061035, + "learning_rate": 1.4847953957969112e-07, + "loss": 0.1457, + "step": 2542 + }, + { + "epoch": 2.502952755905512, + "grad_norm": 11.22967529296875, + "learning_rate": 1.479390857505457e-07, + "loss": 0.3991, + "step": 2543 + }, + { + "epoch": 2.5039370078740157, + "grad_norm": 11.164673805236816, + "learning_rate": 1.4739946917516362e-07, + "loss": 0.3166, + "step": 2544 + }, + { + "epoch": 2.5049212598425195, + "grad_norm": 10.916154861450195, + "learning_rate": 1.4686069113078744e-07, + "loss": 0.4918, + "step": 2545 + }, + { + "epoch": 2.5059055118110236, + "grad_norm": 4.393553733825684, + "learning_rate": 1.4632275289267535e-07, + "loss": 0.0712, + "step": 2546 + }, + { + "epoch": 2.5068897637795278, + "grad_norm": 8.63255500793457, + "learning_rate": 1.457856557340976e-07, + "loss": 0.2493, + "step": 2547 + }, + { + "epoch": 2.5078740157480315, + "grad_norm": 1.1968053579330444, + "learning_rate": 1.4524940092633376e-07, + "loss": 0.0203, + "step": 2548 + }, + { + "epoch": 2.508858267716535, + "grad_norm": 12.2052640914917, + "learning_rate": 1.447139897386694e-07, + "loss": 0.4734, + "step": 2549 + }, + { + "epoch": 2.5098425196850394, + "grad_norm": 12.872119903564453, + "learning_rate": 1.4417942343839348e-07, + "loss": 0.4039, + "step": 2550 + }, + { + "epoch": 2.5108267716535435, + "grad_norm": 10.173605918884277, + "learning_rate": 1.4364570329079504e-07, + "loss": 0.2449, + "step": 2551 + }, + { + "epoch": 2.5118110236220472, + "grad_norm": 13.367107391357422, + "learning_rate": 1.4311283055916042e-07, + "loss": 0.4475, + "step": 2552 + }, + { + "epoch": 2.512795275590551, + "grad_norm": 13.992383003234863, + "learning_rate": 1.4258080650477008e-07, + "loss": 0.484, + "step": 2553 + }, + { + "epoch": 2.513779527559055, + "grad_norm": 8.90416145324707, + "learning_rate": 1.4204963238689582e-07, + "loss": 0.1488, + "step": 2554 + }, + { + "epoch": 2.5147637795275593, + "grad_norm": 10.0111083984375, + "learning_rate": 1.4151930946279746e-07, + "loss": 0.4524, + "step": 2555 + }, + { + "epoch": 2.515748031496063, + "grad_norm": 3.560452938079834, + "learning_rate": 1.4098983898772037e-07, + "loss": 0.068, + "step": 2556 + }, + { + "epoch": 2.5167322834645667, + "grad_norm": 11.944955825805664, + "learning_rate": 1.404612222148921e-07, + "loss": 0.3445, + "step": 2557 + }, + { + "epoch": 2.517716535433071, + "grad_norm": 11.267038345336914, + "learning_rate": 1.3993346039551943e-07, + "loss": 0.3387, + "step": 2558 + }, + { + "epoch": 2.518700787401575, + "grad_norm": 11.368459701538086, + "learning_rate": 1.3940655477878572e-07, + "loss": 0.4325, + "step": 2559 + }, + { + "epoch": 2.5196850393700787, + "grad_norm": 11.744010925292969, + "learning_rate": 1.3888050661184755e-07, + "loss": 0.3661, + "step": 2560 + }, + { + "epoch": 2.5206692913385824, + "grad_norm": 11.511897087097168, + "learning_rate": 1.3835531713983213e-07, + "loss": 0.3591, + "step": 2561 + }, + { + "epoch": 2.5216535433070866, + "grad_norm": 6.421428203582764, + "learning_rate": 1.3783098760583417e-07, + "loss": 0.1244, + "step": 2562 + }, + { + "epoch": 2.5226377952755907, + "grad_norm": 9.582514762878418, + "learning_rate": 1.3730751925091263e-07, + "loss": 0.1832, + "step": 2563 + }, + { + "epoch": 2.5236220472440944, + "grad_norm": 6.133522033691406, + "learning_rate": 1.3678491331408854e-07, + "loss": 0.0931, + "step": 2564 + }, + { + "epoch": 2.524606299212598, + "grad_norm": 12.343705177307129, + "learning_rate": 1.3626317103234146e-07, + "loss": 0.2654, + "step": 2565 + }, + { + "epoch": 2.5255905511811023, + "grad_norm": 5.640179634094238, + "learning_rate": 1.3574229364060663e-07, + "loss": 0.0773, + "step": 2566 + }, + { + "epoch": 2.5265748031496065, + "grad_norm": 12.201597213745117, + "learning_rate": 1.3522228237177238e-07, + "loss": 0.6802, + "step": 2567 + }, + { + "epoch": 2.52755905511811, + "grad_norm": 5.942319393157959, + "learning_rate": 1.347031384566767e-07, + "loss": 0.0989, + "step": 2568 + }, + { + "epoch": 2.528543307086614, + "grad_norm": 11.994881629943848, + "learning_rate": 1.3418486312410484e-07, + "loss": 0.3931, + "step": 2569 + }, + { + "epoch": 2.529527559055118, + "grad_norm": 4.916299343109131, + "learning_rate": 1.3366745760078604e-07, + "loss": 0.0766, + "step": 2570 + }, + { + "epoch": 2.530511811023622, + "grad_norm": 13.462963104248047, + "learning_rate": 1.3315092311139072e-07, + "loss": 0.4033, + "step": 2571 + }, + { + "epoch": 2.531496062992126, + "grad_norm": 3.174868106842041, + "learning_rate": 1.3263526087852766e-07, + "loss": 0.0667, + "step": 2572 + }, + { + "epoch": 2.53248031496063, + "grad_norm": 4.65784215927124, + "learning_rate": 1.321204721227411e-07, + "loss": 0.0691, + "step": 2573 + }, + { + "epoch": 2.533464566929134, + "grad_norm": 10.906597137451172, + "learning_rate": 1.316065580625077e-07, + "loss": 0.4759, + "step": 2574 + }, + { + "epoch": 2.534448818897638, + "grad_norm": 10.565220832824707, + "learning_rate": 1.3109351991423402e-07, + "loss": 0.2557, + "step": 2575 + }, + { + "epoch": 2.5354330708661417, + "grad_norm": 3.8907289505004883, + "learning_rate": 1.3058135889225286e-07, + "loss": 0.0529, + "step": 2576 + }, + { + "epoch": 2.536417322834646, + "grad_norm": 10.592839241027832, + "learning_rate": 1.3007007620882145e-07, + "loss": 0.4901, + "step": 2577 + }, + { + "epoch": 2.5374015748031495, + "grad_norm": 11.24435043334961, + "learning_rate": 1.2955967307411777e-07, + "loss": 0.3706, + "step": 2578 + }, + { + "epoch": 2.5383858267716537, + "grad_norm": 13.005301475524902, + "learning_rate": 1.290501506962381e-07, + "loss": 0.4532, + "step": 2579 + }, + { + "epoch": 2.5393700787401574, + "grad_norm": 8.798903465270996, + "learning_rate": 1.2854151028119375e-07, + "loss": 0.2033, + "step": 2580 + }, + { + "epoch": 2.5403543307086616, + "grad_norm": 8.366865158081055, + "learning_rate": 1.2803375303290884e-07, + "loss": 0.1717, + "step": 2581 + }, + { + "epoch": 2.5413385826771653, + "grad_norm": 11.546663284301758, + "learning_rate": 1.2752688015321684e-07, + "loss": 0.4391, + "step": 2582 + }, + { + "epoch": 2.5423228346456694, + "grad_norm": 10.409428596496582, + "learning_rate": 1.2702089284185817e-07, + "loss": 0.3789, + "step": 2583 + }, + { + "epoch": 2.543307086614173, + "grad_norm": 10.069719314575195, + "learning_rate": 1.265157922964767e-07, + "loss": 0.1746, + "step": 2584 + }, + { + "epoch": 2.5442913385826773, + "grad_norm": 7.2144341468811035, + "learning_rate": 1.2601157971261788e-07, + "loss": 0.1431, + "step": 2585 + }, + { + "epoch": 2.545275590551181, + "grad_norm": 4.573433876037598, + "learning_rate": 1.255082562837252e-07, + "loss": 0.076, + "step": 2586 + }, + { + "epoch": 2.546259842519685, + "grad_norm": 14.100467681884766, + "learning_rate": 1.250058232011375e-07, + "loss": 0.6025, + "step": 2587 + }, + { + "epoch": 2.547244094488189, + "grad_norm": 8.085832595825195, + "learning_rate": 1.2450428165408641e-07, + "loss": 0.177, + "step": 2588 + }, + { + "epoch": 2.548228346456693, + "grad_norm": 11.929106712341309, + "learning_rate": 1.2400363282969322e-07, + "loss": 0.4074, + "step": 2589 + }, + { + "epoch": 2.5492125984251968, + "grad_norm": 6.334018230438232, + "learning_rate": 1.2350387791296617e-07, + "loss": 0.1275, + "step": 2590 + }, + { + "epoch": 2.550196850393701, + "grad_norm": 10.3558931350708, + "learning_rate": 1.2300501808679776e-07, + "loss": 0.3648, + "step": 2591 + }, + { + "epoch": 2.5511811023622046, + "grad_norm": 7.203764915466309, + "learning_rate": 1.2250705453196177e-07, + "loss": 0.1349, + "step": 2592 + }, + { + "epoch": 2.552165354330709, + "grad_norm": 4.817153453826904, + "learning_rate": 1.2200998842711062e-07, + "loss": 0.0604, + "step": 2593 + }, + { + "epoch": 2.5531496062992125, + "grad_norm": 4.7993245124816895, + "learning_rate": 1.2151382094877245e-07, + "loss": 0.0767, + "step": 2594 + }, + { + "epoch": 2.5541338582677167, + "grad_norm": 9.652447700500488, + "learning_rate": 1.2101855327134849e-07, + "loss": 0.1883, + "step": 2595 + }, + { + "epoch": 2.5551181102362204, + "grad_norm": 11.595111846923828, + "learning_rate": 1.2052418656711007e-07, + "loss": 0.5121, + "step": 2596 + }, + { + "epoch": 2.5561023622047245, + "grad_norm": 12.807509422302246, + "learning_rate": 1.200307220061961e-07, + "loss": 0.3336, + "step": 2597 + }, + { + "epoch": 2.5570866141732282, + "grad_norm": 11.648643493652344, + "learning_rate": 1.1953816075661008e-07, + "loss": 0.3166, + "step": 2598 + }, + { + "epoch": 2.5580708661417324, + "grad_norm": 8.615187644958496, + "learning_rate": 1.1904650398421742e-07, + "loss": 0.1686, + "step": 2599 + }, + { + "epoch": 2.559055118110236, + "grad_norm": 8.56308364868164, + "learning_rate": 1.185557528527428e-07, + "loss": 0.2503, + "step": 2600 + }, + { + "epoch": 2.5600393700787403, + "grad_norm": 9.235340118408203, + "learning_rate": 1.1806590852376707e-07, + "loss": 0.3077, + "step": 2601 + }, + { + "epoch": 2.5600393700787403, + "eval_Qnli-dev_cosine_accuracy": 0.69140625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.759227991104126, + "eval_Qnli-dev_cosine_ap": 0.7062433813191578, + "eval_Qnli-dev_cosine_f1": 0.6965174129353233, + "eval_Qnli-dev_cosine_f1_threshold": 0.693097710609436, + "eval_Qnli-dev_cosine_precision": 0.5722070844686649, + "eval_Qnli-dev_cosine_recall": 0.8898305084745762, + "eval_Qnli-dev_dot_accuracy": 0.693359375, + "eval_Qnli-dev_dot_accuracy_threshold": 579.80322265625, + "eval_Qnli-dev_dot_ap": 0.7072678644366573, + "eval_Qnli-dev_dot_f1": 0.6965174129353233, + "eval_Qnli-dev_dot_f1_threshold": 529.2435302734375, + "eval_Qnli-dev_dot_precision": 0.5722070844686649, + "eval_Qnli-dev_dot_recall": 0.8898305084745762, + "eval_Qnli-dev_euclidean_accuracy": 0.693359375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 19.179542541503906, + "eval_Qnli-dev_euclidean_ap": 0.70607517424392, + "eval_Qnli-dev_euclidean_f1": 0.6966666666666668, + "eval_Qnli-dev_euclidean_f1_threshold": 21.525915145874023, + "eval_Qnli-dev_euclidean_precision": 0.5741758241758241, + "eval_Qnli-dev_euclidean_recall": 0.885593220338983, + "eval_Qnli-dev_manhattan_accuracy": 0.689453125, + "eval_Qnli-dev_manhattan_accuracy_threshold": 422.5657958984375, + "eval_Qnli-dev_manhattan_ap": 0.7087825215845417, + "eval_Qnli-dev_manhattan_f1": 0.696817420435511, + "eval_Qnli-dev_manhattan_f1_threshold": 473.9744873046875, + "eval_Qnli-dev_manhattan_precision": 0.5761772853185596, + "eval_Qnli-dev_manhattan_recall": 0.8813559322033898, + "eval_Qnli-dev_max_accuracy": 0.693359375, + "eval_Qnli-dev_max_accuracy_threshold": 579.80322265625, + "eval_Qnli-dev_max_ap": 0.7087825215845417, + "eval_Qnli-dev_max_f1": 0.696817420435511, + "eval_Qnli-dev_max_f1_threshold": 529.2435302734375, + "eval_Qnli-dev_max_precision": 0.5761772853185596, + "eval_Qnli-dev_max_recall": 0.8898305084745762, + "eval_allNLI-dev_cosine_accuracy": 0.697265625, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9267362356185913, + "eval_allNLI-dev_cosine_ap": 0.5331947256130429, + "eval_allNLI-dev_cosine_f1": 0.5614754098360656, + "eval_allNLI-dev_cosine_f1_threshold": 0.8032419681549072, + "eval_allNLI-dev_cosine_precision": 0.43492063492063493, + "eval_allNLI-dev_cosine_recall": 0.791907514450867, + "eval_allNLI-dev_dot_accuracy": 0.697265625, + "eval_allNLI-dev_dot_accuracy_threshold": 708.6821899414062, + "eval_allNLI-dev_dot_ap": 0.5339001290546905, + "eval_allNLI-dev_dot_f1": 0.5635593220338982, + "eval_allNLI-dev_dot_f1_threshold": 618.6771850585938, + "eval_allNLI-dev_dot_precision": 0.44481605351170567, + "eval_allNLI-dev_dot_recall": 0.7687861271676301, + "eval_allNLI-dev_euclidean_accuracy": 0.697265625, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.585262298583984, + "eval_allNLI-dev_euclidean_ap": 0.5332561135588656, + "eval_allNLI-dev_euclidean_f1": 0.5614754098360656, + "eval_allNLI-dev_euclidean_f1_threshold": 17.341644287109375, + "eval_allNLI-dev_euclidean_precision": 0.43492063492063493, + "eval_allNLI-dev_euclidean_recall": 0.791907514450867, + "eval_allNLI-dev_manhattan_accuracy": 0.69921875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 255.64784240722656, + "eval_allNLI-dev_manhattan_ap": 0.5346969020570975, + "eval_allNLI-dev_manhattan_f1": 0.5630252100840336, + "eval_allNLI-dev_manhattan_f1_threshold": 379.50360107421875, + "eval_allNLI-dev_manhattan_precision": 0.44224422442244227, + "eval_allNLI-dev_manhattan_recall": 0.7745664739884393, + "eval_allNLI-dev_max_accuracy": 0.69921875, + "eval_allNLI-dev_max_accuracy_threshold": 708.6821899414062, + "eval_allNLI-dev_max_ap": 0.5346969020570975, + "eval_allNLI-dev_max_f1": 0.5635593220338982, + "eval_allNLI-dev_max_f1_threshold": 618.6771850585938, + "eval_allNLI-dev_max_precision": 0.44481605351170567, + "eval_allNLI-dev_max_recall": 0.791907514450867, + "eval_loss": 0.715917706489563, + "eval_runtime": 50.3184, + "eval_samples_per_second": 33.069, + "eval_sequential_score": 0.7087825215845417, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.6828491388363669, + "eval_sts-test_pearson_dot": 0.6825898367630883, + "eval_sts-test_pearson_euclidean": 0.7059464426277313, + "eval_sts-test_pearson_manhattan": 0.7071353983847126, + "eval_sts-test_pearson_max": 0.7071353983847126, + "eval_sts-test_spearman_cosine": 0.6963346393914523, + "eval_sts-test_spearman_dot": 0.6960475600496264, + "eval_sts-test_spearman_euclidean": 0.69632877697621, + "eval_sts-test_spearman_manhattan": 0.697445857963062, + "eval_sts-test_spearman_max": 0.697445857963062, + "step": 2601 + }, + { + "epoch": 2.561023622047244, + "grad_norm": 8.847234725952148, + "learning_rate": 1.1757697215672498e-07, + "loss": 0.1409, + "step": 2602 + }, + { + "epoch": 2.562007874015748, + "grad_norm": 14.99544906616211, + "learning_rate": 1.17088944908902e-07, + "loss": 0.661, + "step": 2603 + }, + { + "epoch": 2.562992125984252, + "grad_norm": 12.105046272277832, + "learning_rate": 1.1660182793543199e-07, + "loss": 0.4729, + "step": 2604 + }, + { + "epoch": 2.563976377952756, + "grad_norm": 11.425689697265625, + "learning_rate": 1.161156223892938e-07, + "loss": 0.2749, + "step": 2605 + }, + { + "epoch": 2.5649606299212597, + "grad_norm": 6.868392467498779, + "learning_rate": 1.1563032942130937e-07, + "loss": 0.1253, + "step": 2606 + }, + { + "epoch": 2.565944881889764, + "grad_norm": 9.729452133178711, + "learning_rate": 1.1514595018014061e-07, + "loss": 0.2679, + "step": 2607 + }, + { + "epoch": 2.5669291338582676, + "grad_norm": 14.86729621887207, + "learning_rate": 1.1466248581228649e-07, + "loss": 0.368, + "step": 2608 + }, + { + "epoch": 2.5679133858267718, + "grad_norm": 5.127080917358398, + "learning_rate": 1.1417993746208068e-07, + "loss": 0.0666, + "step": 2609 + }, + { + "epoch": 2.5688976377952755, + "grad_norm": 6.748661994934082, + "learning_rate": 1.136983062716886e-07, + "loss": 0.1593, + "step": 2610 + }, + { + "epoch": 2.5698818897637796, + "grad_norm": 12.44377326965332, + "learning_rate": 1.1321759338110483e-07, + "loss": 0.4291, + "step": 2611 + }, + { + "epoch": 2.5708661417322833, + "grad_norm": 8.08018684387207, + "learning_rate": 1.1273779992815057e-07, + "loss": 0.1222, + "step": 2612 + }, + { + "epoch": 2.5718503937007875, + "grad_norm": 8.787413597106934, + "learning_rate": 1.1225892704847027e-07, + "loss": 0.2681, + "step": 2613 + }, + { + "epoch": 2.572834645669291, + "grad_norm": 14.689745903015137, + "learning_rate": 1.1178097587552988e-07, + "loss": 0.4171, + "step": 2614 + }, + { + "epoch": 2.5738188976377954, + "grad_norm": 10.991671562194824, + "learning_rate": 1.113039475406136e-07, + "loss": 0.3418, + "step": 2615 + }, + { + "epoch": 2.574803149606299, + "grad_norm": 10.120105743408203, + "learning_rate": 1.1082784317282129e-07, + "loss": 0.2284, + "step": 2616 + }, + { + "epoch": 2.5757874015748032, + "grad_norm": 14.081969261169434, + "learning_rate": 1.1035266389906588e-07, + "loss": 0.4991, + "step": 2617 + }, + { + "epoch": 2.576771653543307, + "grad_norm": 14.420641899108887, + "learning_rate": 1.0987841084407059e-07, + "loss": 0.2468, + "step": 2618 + }, + { + "epoch": 2.577755905511811, + "grad_norm": 7.674907684326172, + "learning_rate": 1.0940508513036641e-07, + "loss": 0.1355, + "step": 2619 + }, + { + "epoch": 2.578740157480315, + "grad_norm": 9.165249824523926, + "learning_rate": 1.0893268787828929e-07, + "loss": 0.2884, + "step": 2620 + }, + { + "epoch": 2.579724409448819, + "grad_norm": 7.51609992980957, + "learning_rate": 1.0846122020597759e-07, + "loss": 0.1583, + "step": 2621 + }, + { + "epoch": 2.5807086614173227, + "grad_norm": 10.051922798156738, + "learning_rate": 1.0799068322936935e-07, + "loss": 0.2402, + "step": 2622 + }, + { + "epoch": 2.581692913385827, + "grad_norm": 13.24838638305664, + "learning_rate": 1.0752107806219986e-07, + "loss": 0.7007, + "step": 2623 + }, + { + "epoch": 2.5826771653543306, + "grad_norm": 11.158019065856934, + "learning_rate": 1.0705240581599881e-07, + "loss": 0.4321, + "step": 2624 + }, + { + "epoch": 2.5836614173228347, + "grad_norm": 8.564787864685059, + "learning_rate": 1.0658466760008749e-07, + "loss": 0.1727, + "step": 2625 + }, + { + "epoch": 2.5846456692913384, + "grad_norm": 10.755471229553223, + "learning_rate": 1.0611786452157668e-07, + "loss": 0.2322, + "step": 2626 + }, + { + "epoch": 2.5856299212598426, + "grad_norm": 3.40462589263916, + "learning_rate": 1.0565199768536355e-07, + "loss": 0.0606, + "step": 2627 + }, + { + "epoch": 2.5866141732283463, + "grad_norm": 7.5750274658203125, + "learning_rate": 1.0518706819412948e-07, + "loss": 0.1847, + "step": 2628 + }, + { + "epoch": 2.5875984251968505, + "grad_norm": 6.758688926696777, + "learning_rate": 1.0472307714833692e-07, + "loss": 0.1452, + "step": 2629 + }, + { + "epoch": 2.588582677165354, + "grad_norm": 6.559057235717773, + "learning_rate": 1.042600256462273e-07, + "loss": 0.1579, + "step": 2630 + }, + { + "epoch": 2.5895669291338583, + "grad_norm": 13.190975189208984, + "learning_rate": 1.0379791478381805e-07, + "loss": 0.4174, + "step": 2631 + }, + { + "epoch": 2.590551181102362, + "grad_norm": 10.964961051940918, + "learning_rate": 1.0333674565490014e-07, + "loss": 0.356, + "step": 2632 + }, + { + "epoch": 2.591535433070866, + "grad_norm": 9.847637176513672, + "learning_rate": 1.0287651935103572e-07, + "loss": 0.179, + "step": 2633 + }, + { + "epoch": 2.59251968503937, + "grad_norm": 6.125359058380127, + "learning_rate": 1.024172369615551e-07, + "loss": 0.2442, + "step": 2634 + }, + { + "epoch": 2.593503937007874, + "grad_norm": 12.130481719970703, + "learning_rate": 1.0195889957355454e-07, + "loss": 0.4187, + "step": 2635 + }, + { + "epoch": 2.594488188976378, + "grad_norm": 8.817065238952637, + "learning_rate": 1.0150150827189337e-07, + "loss": 0.2197, + "step": 2636 + }, + { + "epoch": 2.595472440944882, + "grad_norm": 9.638463973999023, + "learning_rate": 1.0104506413919173e-07, + "loss": 0.1489, + "step": 2637 + }, + { + "epoch": 2.596456692913386, + "grad_norm": 8.398432731628418, + "learning_rate": 1.0058956825582792e-07, + "loss": 0.3529, + "step": 2638 + }, + { + "epoch": 2.59744094488189, + "grad_norm": 10.151667594909668, + "learning_rate": 1.0013502169993552e-07, + "loss": 0.5286, + "step": 2639 + }, + { + "epoch": 2.5984251968503935, + "grad_norm": 9.8574800491333, + "learning_rate": 9.96814255474014e-08, + "loss": 0.3814, + "step": 2640 + }, + { + "epoch": 2.5994094488188977, + "grad_norm": 7.26898193359375, + "learning_rate": 9.922878087186273e-08, + "loss": 0.1471, + "step": 2641 + }, + { + "epoch": 2.600393700787402, + "grad_norm": 7.405521869659424, + "learning_rate": 9.877708874470455e-08, + "loss": 0.1494, + "step": 2642 + }, + { + "epoch": 2.6013779527559056, + "grad_norm": 11.82184886932373, + "learning_rate": 9.832635023505755e-08, + "loss": 0.4368, + "step": 2643 + }, + { + "epoch": 2.6023622047244093, + "grad_norm": 13.330612182617188, + "learning_rate": 9.787656640979466e-08, + "loss": 0.4112, + "step": 2644 + }, + { + "epoch": 2.6033464566929134, + "grad_norm": 10.119935035705566, + "learning_rate": 9.74277383335297e-08, + "loss": 0.3248, + "step": 2645 + }, + { + "epoch": 2.6043307086614176, + "grad_norm": 5.714234828948975, + "learning_rate": 9.697986706861403e-08, + "loss": 0.1433, + "step": 2646 + }, + { + "epoch": 2.6053149606299213, + "grad_norm": 8.152800559997559, + "learning_rate": 9.653295367513442e-08, + "loss": 0.1752, + "step": 2647 + }, + { + "epoch": 2.606299212598425, + "grad_norm": 6.399628162384033, + "learning_rate": 9.608699921091029e-08, + "loss": 0.1991, + "step": 2648 + }, + { + "epoch": 2.607283464566929, + "grad_norm": 9.352363586425781, + "learning_rate": 9.564200473149156e-08, + "loss": 0.208, + "step": 2649 + }, + { + "epoch": 2.6082677165354333, + "grad_norm": 10.281725883483887, + "learning_rate": 9.519797129015536e-08, + "loss": 0.2823, + "step": 2650 + }, + { + "epoch": 2.609251968503937, + "grad_norm": 10.279464721679688, + "learning_rate": 9.475489993790467e-08, + "loss": 0.5224, + "step": 2651 + }, + { + "epoch": 2.6102362204724407, + "grad_norm": 19.64162254333496, + "learning_rate": 9.431279172346506e-08, + "loss": 0.4913, + "step": 2652 + }, + { + "epoch": 2.611220472440945, + "grad_norm": 17.06022834777832, + "learning_rate": 9.387164769328237e-08, + "loss": 0.7894, + "step": 2653 + }, + { + "epoch": 2.612204724409449, + "grad_norm": 10.321503639221191, + "learning_rate": 9.343146889152028e-08, + "loss": 0.2381, + "step": 2654 + }, + { + "epoch": 2.6131889763779528, + "grad_norm": 3.353727102279663, + "learning_rate": 9.299225636005791e-08, + "loss": 0.1165, + "step": 2655 + }, + { + "epoch": 2.6141732283464565, + "grad_norm": 7.8739333152771, + "learning_rate": 9.25540111384872e-08, + "loss": 0.1246, + "step": 2656 + }, + { + "epoch": 2.6151574803149606, + "grad_norm": 4.163465976715088, + "learning_rate": 9.211673426411052e-08, + "loss": 0.0566, + "step": 2657 + }, + { + "epoch": 2.616141732283465, + "grad_norm": 2.981975555419922, + "learning_rate": 9.168042677193818e-08, + "loss": 0.048, + "step": 2658 + }, + { + "epoch": 2.6171259842519685, + "grad_norm": 9.967344284057617, + "learning_rate": 9.124508969468606e-08, + "loss": 0.4056, + "step": 2659 + }, + { + "epoch": 2.6181102362204722, + "grad_norm": 6.563957214355469, + "learning_rate": 9.08107240627732e-08, + "loss": 0.2078, + "step": 2660 + }, + { + "epoch": 2.6190944881889764, + "grad_norm": 6.7945966720581055, + "learning_rate": 9.037733090431912e-08, + "loss": 0.119, + "step": 2661 + }, + { + "epoch": 2.6200787401574805, + "grad_norm": 16.327863693237305, + "learning_rate": 8.994491124514166e-08, + "loss": 0.764, + "step": 2662 + }, + { + "epoch": 2.6210629921259843, + "grad_norm": 15.108491897583008, + "learning_rate": 8.951346610875441e-08, + "loss": 0.5038, + "step": 2663 + }, + { + "epoch": 2.622047244094488, + "grad_norm": 5.67038631439209, + "learning_rate": 8.908299651636436e-08, + "loss": 0.1146, + "step": 2664 + }, + { + "epoch": 2.623031496062992, + "grad_norm": 8.750030517578125, + "learning_rate": 8.865350348686934e-08, + "loss": 0.3114, + "step": 2665 + }, + { + "epoch": 2.6240157480314963, + "grad_norm": 7.805241107940674, + "learning_rate": 8.822498803685595e-08, + "loss": 0.178, + "step": 2666 + }, + { + "epoch": 2.625, + "grad_norm": 2.330080986022949, + "learning_rate": 8.779745118059643e-08, + "loss": 0.0308, + "step": 2667 + }, + { + "epoch": 2.6259842519685037, + "grad_norm": 10.4607572555542, + "learning_rate": 8.737089393004722e-08, + "loss": 0.3818, + "step": 2668 + }, + { + "epoch": 2.626968503937008, + "grad_norm": 6.982802391052246, + "learning_rate": 8.694531729484584e-08, + "loss": 0.1605, + "step": 2669 + }, + { + "epoch": 2.627952755905512, + "grad_norm": 5.182496070861816, + "learning_rate": 8.65207222823089e-08, + "loss": 0.0779, + "step": 2670 + }, + { + "epoch": 2.6289370078740157, + "grad_norm": 9.875541687011719, + "learning_rate": 8.609710989742948e-08, + "loss": 0.2747, + "step": 2671 + }, + { + "epoch": 2.6299212598425195, + "grad_norm": 9.010579109191895, + "learning_rate": 8.567448114287475e-08, + "loss": 0.2404, + "step": 2672 + }, + { + "epoch": 2.6309055118110236, + "grad_norm": 6.015598297119141, + "learning_rate": 8.525283701898387e-08, + "loss": 0.0797, + "step": 2673 + }, + { + "epoch": 2.6318897637795278, + "grad_norm": 7.764693260192871, + "learning_rate": 8.483217852376527e-08, + "loss": 0.205, + "step": 2674 + }, + { + "epoch": 2.6328740157480315, + "grad_norm": 5.811929225921631, + "learning_rate": 8.441250665289457e-08, + "loss": 0.1063, + "step": 2675 + }, + { + "epoch": 2.633858267716535, + "grad_norm": 5.1248087882995605, + "learning_rate": 8.399382239971196e-08, + "loss": 0.0866, + "step": 2676 + }, + { + "epoch": 2.6348425196850394, + "grad_norm": 13.782855033874512, + "learning_rate": 8.357612675522011e-08, + "loss": 0.5038, + "step": 2677 + }, + { + "epoch": 2.6358267716535435, + "grad_norm": 9.879081726074219, + "learning_rate": 8.315942070808176e-08, + "loss": 0.2404, + "step": 2678 + }, + { + "epoch": 2.6368110236220472, + "grad_norm": 5.677542686462402, + "learning_rate": 8.274370524461712e-08, + "loss": 0.0738, + "step": 2679 + }, + { + "epoch": 2.637795275590551, + "grad_norm": 5.569581031799316, + "learning_rate": 8.232898134880213e-08, + "loss": 0.1597, + "step": 2680 + }, + { + "epoch": 2.638779527559055, + "grad_norm": 15.741230010986328, + "learning_rate": 8.19152500022651e-08, + "loss": 0.4871, + "step": 2681 + }, + { + "epoch": 2.6397637795275593, + "grad_norm": 13.591595649719238, + "learning_rate": 8.150251218428557e-08, + "loss": 0.4956, + "step": 2682 + }, + { + "epoch": 2.640748031496063, + "grad_norm": 7.691888332366943, + "learning_rate": 8.109076887179138e-08, + "loss": 0.1012, + "step": 2683 + }, + { + "epoch": 2.6417322834645667, + "grad_norm": 12.29000473022461, + "learning_rate": 8.068002103935624e-08, + "loss": 0.2641, + "step": 2684 + }, + { + "epoch": 2.642716535433071, + "grad_norm": 6.156375408172607, + "learning_rate": 8.027026965919774e-08, + "loss": 0.1694, + "step": 2685 + }, + { + "epoch": 2.643700787401575, + "grad_norm": 14.0615873336792, + "learning_rate": 7.986151570117498e-08, + "loss": 0.4639, + "step": 2686 + }, + { + "epoch": 2.6446850393700787, + "grad_norm": 9.258757591247559, + "learning_rate": 7.945376013278619e-08, + "loss": 0.1697, + "step": 2687 + }, + { + "epoch": 2.6456692913385824, + "grad_norm": 12.940256118774414, + "learning_rate": 7.904700391916614e-08, + "loss": 0.296, + "step": 2688 + }, + { + "epoch": 2.6466535433070866, + "grad_norm": 5.00270938873291, + "learning_rate": 7.864124802308461e-08, + "loss": 0.075, + "step": 2689 + }, + { + "epoch": 2.6476377952755907, + "grad_norm": 7.078151702880859, + "learning_rate": 7.823649340494358e-08, + "loss": 0.1418, + "step": 2690 + }, + { + "epoch": 2.6486220472440944, + "grad_norm": 11.775737762451172, + "learning_rate": 7.783274102277492e-08, + "loss": 0.3814, + "step": 2691 + }, + { + "epoch": 2.649606299212598, + "grad_norm": 14.035433769226074, + "learning_rate": 7.742999183223844e-08, + "loss": 0.4045, + "step": 2692 + }, + { + "epoch": 2.6505905511811023, + "grad_norm": 7.129059791564941, + "learning_rate": 7.702824678661923e-08, + "loss": 0.134, + "step": 2693 + }, + { + "epoch": 2.6515748031496065, + "grad_norm": 10.034149169921875, + "learning_rate": 7.662750683682593e-08, + "loss": 0.2096, + "step": 2694 + }, + { + "epoch": 2.65255905511811, + "grad_norm": 9.954490661621094, + "learning_rate": 7.622777293138787e-08, + "loss": 0.3959, + "step": 2695 + }, + { + "epoch": 2.653543307086614, + "grad_norm": 12.642629623413086, + "learning_rate": 7.582904601645333e-08, + "loss": 0.3948, + "step": 2696 + }, + { + "epoch": 2.654527559055118, + "grad_norm": 6.058542728424072, + "learning_rate": 7.543132703578697e-08, + "loss": 0.1326, + "step": 2697 + }, + { + "epoch": 2.655511811023622, + "grad_norm": 12.721632957458496, + "learning_rate": 7.503461693076783e-08, + "loss": 0.4219, + "step": 2698 + }, + { + "epoch": 2.656496062992126, + "grad_norm": 9.701395034790039, + "learning_rate": 7.463891664038691e-08, + "loss": 0.2354, + "step": 2699 + }, + { + "epoch": 2.65748031496063, + "grad_norm": 7.81613302230835, + "learning_rate": 7.424422710124503e-08, + "loss": 0.0761, + "step": 2700 + }, + { + "epoch": 2.658464566929134, + "grad_norm": 6.695240497589111, + "learning_rate": 7.385054924755081e-08, + "loss": 0.136, + "step": 2701 + }, + { + "epoch": 2.659448818897638, + "grad_norm": 7.1679582595825195, + "learning_rate": 7.345788401111793e-08, + "loss": 0.097, + "step": 2702 + }, + { + "epoch": 2.6604330708661417, + "grad_norm": 13.151243209838867, + "learning_rate": 7.306623232136358e-08, + "loss": 0.7051, + "step": 2703 + }, + { + "epoch": 2.661417322834646, + "grad_norm": 4.02302360534668, + "learning_rate": 7.26755951053058e-08, + "loss": 0.0625, + "step": 2704 + }, + { + "epoch": 2.6624015748031495, + "grad_norm": 15.299405097961426, + "learning_rate": 7.228597328756136e-08, + "loss": 0.4963, + "step": 2705 + }, + { + "epoch": 2.6633858267716537, + "grad_norm": 10.938726425170898, + "learning_rate": 7.189736779034375e-08, + "loss": 0.264, + "step": 2706 + }, + { + "epoch": 2.6643700787401574, + "grad_norm": 3.407548189163208, + "learning_rate": 7.150977953346082e-08, + "loss": 0.0416, + "step": 2707 + }, + { + "epoch": 2.6653543307086616, + "grad_norm": 3.334954261779785, + "learning_rate": 7.112320943431287e-08, + "loss": 0.0484, + "step": 2708 + }, + { + "epoch": 2.6663385826771653, + "grad_norm": 6.820202350616455, + "learning_rate": 7.07376584078898e-08, + "loss": 0.0973, + "step": 2709 + }, + { + "epoch": 2.6673228346456694, + "grad_norm": 8.80772590637207, + "learning_rate": 7.035312736676993e-08, + "loss": 0.1573, + "step": 2710 + }, + { + "epoch": 2.668307086614173, + "grad_norm": 9.831692695617676, + "learning_rate": 6.996961722111701e-08, + "loss": 0.3998, + "step": 2711 + }, + { + "epoch": 2.6692913385826773, + "grad_norm": 7.1447434425354, + "learning_rate": 6.958712887867856e-08, + "loss": 0.1709, + "step": 2712 + }, + { + "epoch": 2.670275590551181, + "grad_norm": 9.20654296875, + "learning_rate": 6.920566324478348e-08, + "loss": 0.2564, + "step": 2713 + }, + { + "epoch": 2.671259842519685, + "grad_norm": 6.765684127807617, + "learning_rate": 6.882522122234002e-08, + "loss": 0.1327, + "step": 2714 + }, + { + "epoch": 2.672244094488189, + "grad_norm": 7.66633939743042, + "learning_rate": 6.844580371183347e-08, + "loss": 0.3002, + "step": 2715 + }, + { + "epoch": 2.673228346456693, + "grad_norm": 12.599959373474121, + "learning_rate": 6.806741161132428e-08, + "loss": 0.3431, + "step": 2716 + }, + { + "epoch": 2.6742125984251968, + "grad_norm": 6.302804470062256, + "learning_rate": 6.769004581644596e-08, + "loss": 0.1228, + "step": 2717 + }, + { + "epoch": 2.675196850393701, + "grad_norm": 4.744940757751465, + "learning_rate": 6.73137072204023e-08, + "loss": 0.0956, + "step": 2718 + }, + { + "epoch": 2.6761811023622046, + "grad_norm": 6.26704216003418, + "learning_rate": 6.693839671396629e-08, + "loss": 0.1675, + "step": 2719 + }, + { + "epoch": 2.677165354330709, + "grad_norm": 7.81545877456665, + "learning_rate": 6.656411518547725e-08, + "loss": 0.2229, + "step": 2720 + }, + { + "epoch": 2.6781496062992125, + "grad_norm": 18.28624725341797, + "learning_rate": 6.619086352083902e-08, + "loss": 0.7096, + "step": 2721 + }, + { + "epoch": 2.6791338582677167, + "grad_norm": 6.871081352233887, + "learning_rate": 6.581864260351775e-08, + "loss": 0.1175, + "step": 2722 + }, + { + "epoch": 2.6801181102362204, + "grad_norm": 10.073604583740234, + "learning_rate": 6.544745331454003e-08, + "loss": 0.2552, + "step": 2723 + }, + { + "epoch": 2.6811023622047245, + "grad_norm": 11.920567512512207, + "learning_rate": 6.50772965324904e-08, + "loss": 0.2744, + "step": 2724 + }, + { + "epoch": 2.6820866141732282, + "grad_norm": 12.86720085144043, + "learning_rate": 6.470817313350971e-08, + "loss": 0.5938, + "step": 2725 + }, + { + "epoch": 2.6830708661417324, + "grad_norm": 13.177947998046875, + "learning_rate": 6.434008399129279e-08, + "loss": 0.3794, + "step": 2726 + }, + { + "epoch": 2.684055118110236, + "grad_norm": 11.583633422851562, + "learning_rate": 6.397302997708641e-08, + "loss": 0.1605, + "step": 2727 + }, + { + "epoch": 2.6850393700787403, + "grad_norm": 16.39134979248047, + "learning_rate": 6.360701195968723e-08, + "loss": 0.565, + "step": 2728 + }, + { + "epoch": 2.686023622047244, + "grad_norm": 6.332181453704834, + "learning_rate": 6.324203080543998e-08, + "loss": 0.0512, + "step": 2729 + }, + { + "epoch": 2.687007874015748, + "grad_norm": 8.339241027832031, + "learning_rate": 6.287808737823473e-08, + "loss": 0.1905, + "step": 2730 + }, + { + "epoch": 2.687992125984252, + "grad_norm": 9.738916397094727, + "learning_rate": 6.251518253950574e-08, + "loss": 0.1798, + "step": 2731 + }, + { + "epoch": 2.688976377952756, + "grad_norm": 9.08163070678711, + "learning_rate": 6.215331714822883e-08, + "loss": 0.2548, + "step": 2732 + }, + { + "epoch": 2.6899606299212597, + "grad_norm": 8.07443618774414, + "learning_rate": 6.179249206091951e-08, + "loss": 0.1854, + "step": 2733 + }, + { + "epoch": 2.690944881889764, + "grad_norm": 9.08901596069336, + "learning_rate": 6.143270813163099e-08, + "loss": 0.2079, + "step": 2734 + }, + { + "epoch": 2.6919291338582676, + "grad_norm": 13.154987335205078, + "learning_rate": 6.107396621195201e-08, + "loss": 0.3843, + "step": 2735 + }, + { + "epoch": 2.6929133858267718, + "grad_norm": 6.714513778686523, + "learning_rate": 6.071626715100503e-08, + "loss": 0.1716, + "step": 2736 + }, + { + "epoch": 2.6938976377952755, + "grad_norm": 10.547233581542969, + "learning_rate": 6.035961179544409e-08, + "loss": 0.2087, + "step": 2737 + }, + { + "epoch": 2.6948818897637796, + "grad_norm": 17.66176414489746, + "learning_rate": 6.000400098945283e-08, + "loss": 0.4848, + "step": 2738 + }, + { + "epoch": 2.6958661417322833, + "grad_norm": 12.70285415649414, + "learning_rate": 5.96494355747425e-08, + "loss": 0.2722, + "step": 2739 + }, + { + "epoch": 2.6968503937007875, + "grad_norm": 11.402365684509277, + "learning_rate": 5.9295916390550034e-08, + "loss": 0.2543, + "step": 2740 + }, + { + "epoch": 2.697834645669291, + "grad_norm": 14.697494506835938, + "learning_rate": 5.89434442736358e-08, + "loss": 0.5198, + "step": 2741 + }, + { + "epoch": 2.6988188976377954, + "grad_norm": 8.330411911010742, + "learning_rate": 5.859202005828208e-08, + "loss": 0.1446, + "step": 2742 + }, + { + "epoch": 2.699803149606299, + "grad_norm": 13.814299583435059, + "learning_rate": 5.824164457629061e-08, + "loss": 0.4176, + "step": 2743 + }, + { + "epoch": 2.7007874015748032, + "grad_norm": 4.575950622558594, + "learning_rate": 5.789231865698098e-08, + "loss": 0.0883, + "step": 2744 + }, + { + "epoch": 2.701771653543307, + "grad_norm": 11.087409973144531, + "learning_rate": 5.754404312718839e-08, + "loss": 0.3302, + "step": 2745 + } + ], + "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 +}