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": 2.699803149606299, + "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": 5.5525, + "step": 1 + }, + { + "epoch": 0.001968503937007874, + "grad_norm": NaN, + "learning_rate": 0.0, + "loss": 6.2948, + "step": 2 + }, + { + "epoch": 0.002952755905511811, + "grad_norm": 27.109525680541992, + "learning_rate": 9.940357852882705e-10, + "loss": 4.6464, + "step": 3 + }, + { + "epoch": 0.003937007874015748, + "grad_norm": Infinity, + "learning_rate": 9.940357852882705e-10, + "loss": 5.0071, + "step": 4 + }, + { + "epoch": 0.004921259842519685, + "grad_norm": 23.50518798828125, + "learning_rate": 1.988071570576541e-09, + "loss": 5.3338, + "step": 5 + }, + { + "epoch": 0.005905511811023622, + "grad_norm": 37.24928665161133, + "learning_rate": 2.9821073558648116e-09, + "loss": 5.1998, + "step": 6 + }, + { + "epoch": 0.006889763779527559, + "grad_norm": Infinity, + "learning_rate": 2.9821073558648116e-09, + "loss": 26.7171, + "step": 7 + }, + { + "epoch": 0.007874015748031496, + "grad_norm": 45.332584381103516, + "learning_rate": 3.976143141153082e-09, + "loss": 6.8404, + "step": 8 + }, + { + "epoch": 0.008858267716535433, + "grad_norm": Infinity, + "learning_rate": 3.976143141153082e-09, + "loss": 26.8068, + "step": 9 + }, + { + "epoch": 0.00984251968503937, + "grad_norm": 37.57707595825195, + "learning_rate": 4.970178926441353e-09, + "loss": 5.7224, + "step": 10 + }, + { + "epoch": 0.010826771653543307, + "grad_norm": 38.55592727661133, + "learning_rate": 5.964214711729623e-09, + "loss": 5.9428, + "step": 11 + }, + { + "epoch": 0.011811023622047244, + "grad_norm": 44.171085357666016, + "learning_rate": 6.9582504970178946e-09, + "loss": 6.7852, + "step": 12 + }, + { + "epoch": 0.012795275590551181, + "grad_norm": 33.0239372253418, + "learning_rate": 7.952286282306164e-09, + "loss": 6.505, + "step": 13 + }, + { + "epoch": 0.013779527559055118, + "grad_norm": 29.830385208129883, + "learning_rate": 8.946322067594435e-09, + "loss": 5.4793, + "step": 14 + }, + { + "epoch": 0.014763779527559055, + "grad_norm": 37.74555206298828, + "learning_rate": 9.940357852882705e-09, + "loss": 5.8316, + "step": 15 + }, + { + "epoch": 0.015748031496062992, + "grad_norm": 26.375228881835938, + "learning_rate": 1.0934393638170978e-08, + "loss": 4.7716, + "step": 16 + }, + { + "epoch": 0.01673228346456693, + "grad_norm": 27.332616806030273, + "learning_rate": 1.1928429423459246e-08, + "loss": 4.9198, + "step": 17 + }, + { + "epoch": 0.017716535433070866, + "grad_norm": 19.702518463134766, + "learning_rate": 1.2922465208747517e-08, + "loss": 4.9205, + "step": 18 + }, + { + "epoch": 0.018700787401574805, + "grad_norm": 37.775146484375, + "learning_rate": 1.3916500994035789e-08, + "loss": 6.1949, + "step": 19 + }, + { + "epoch": 0.01968503937007874, + "grad_norm": 41.4459342956543, + "learning_rate": 1.4910536779324056e-08, + "loss": 6.935, + "step": 20 + }, + { + "epoch": 0.02066929133858268, + "grad_norm": 69.52754974365234, + "learning_rate": 1.590457256461233e-08, + "loss": 26.2611, + "step": 21 + }, + { + "epoch": 0.021653543307086614, + "grad_norm": 39.509559631347656, + "learning_rate": 1.68986083499006e-08, + "loss": 6.2476, + "step": 22 + }, + { + "epoch": 0.022637795275590553, + "grad_norm": 32.32150650024414, + "learning_rate": 1.789264413518887e-08, + "loss": 5.1881, + "step": 23 + }, + { + "epoch": 0.023622047244094488, + "grad_norm": 28.91634178161621, + "learning_rate": 1.888667992047714e-08, + "loss": 4.9198, + "step": 24 + }, + { + "epoch": 0.024606299212598427, + "grad_norm": 70.45647430419922, + "learning_rate": 1.988071570576541e-08, + "loss": 25.6461, + "step": 25 + }, + { + "epoch": 0.025590551181102362, + "grad_norm": 69.56190490722656, + "learning_rate": 2.087475149105368e-08, + "loss": 25.9935, + "step": 26 + }, + { + "epoch": 0.0265748031496063, + "grad_norm": 29.483633041381836, + "learning_rate": 2.1868787276341955e-08, + "loss": 6.194, + "step": 27 + }, + { + "epoch": 0.027559055118110236, + "grad_norm": 64.40437316894531, + "learning_rate": 2.2862823061630224e-08, + "loss": 24.9355, + "step": 28 + }, + { + "epoch": 0.028543307086614175, + "grad_norm": 29.5191650390625, + "learning_rate": 2.3856858846918493e-08, + "loss": 5.2458, + "step": 29 + }, + { + "epoch": 0.02952755905511811, + "grad_norm": 23.680418014526367, + "learning_rate": 2.4850894632206765e-08, + "loss": 5.487, + "step": 30 + }, + { + "epoch": 0.03051181102362205, + "grad_norm": 20.74715805053711, + "learning_rate": 2.5844930417495034e-08, + "loss": 4.3883, + "step": 31 + }, + { + "epoch": 0.031496062992125984, + "grad_norm": 37.96383285522461, + "learning_rate": 2.6838966202783303e-08, + "loss": 6.4174, + "step": 32 + }, + { + "epoch": 0.03248031496062992, + "grad_norm": 27.312122344970703, + "learning_rate": 2.7833001988071578e-08, + "loss": 5.9618, + "step": 33 + }, + { + "epoch": 0.03346456692913386, + "grad_norm": 27.779197692871094, + "learning_rate": 2.8827037773359847e-08, + "loss": 5.1205, + "step": 34 + }, + { + "epoch": 0.0344488188976378, + "grad_norm": 29.39383316040039, + "learning_rate": 2.982107355864811e-08, + "loss": 5.6965, + "step": 35 + }, + { + "epoch": 0.03543307086614173, + "grad_norm": 62.016685485839844, + "learning_rate": 3.081510934393639e-08, + "loss": 23.4505, + "step": 36 + }, + { + "epoch": 0.03641732283464567, + "grad_norm": 27.892751693725586, + "learning_rate": 3.180914512922466e-08, + "loss": 5.2556, + "step": 37 + }, + { + "epoch": 0.03740157480314961, + "grad_norm": 24.589818954467773, + "learning_rate": 3.280318091451293e-08, + "loss": 4.9034, + "step": 38 + }, + { + "epoch": 0.038385826771653545, + "grad_norm": 59.737709045410156, + "learning_rate": 3.37972166998012e-08, + "loss": 22.9819, + "step": 39 + }, + { + "epoch": 0.03937007874015748, + "grad_norm": 20.44304847717285, + "learning_rate": 3.479125248508947e-08, + "loss": 4.5872, + "step": 40 + }, + { + "epoch": 0.040354330708661415, + "grad_norm": 19.250782012939453, + "learning_rate": 3.578528827037774e-08, + "loss": 5.0072, + "step": 41 + }, + { + "epoch": 0.04133858267716536, + "grad_norm": 18.789878845214844, + "learning_rate": 3.6779324055666005e-08, + "loss": 4.3917, + "step": 42 + }, + { + "epoch": 0.04232283464566929, + "grad_norm": 15.848797798156738, + "learning_rate": 3.777335984095428e-08, + "loss": 4.0698, + "step": 43 + }, + { + "epoch": 0.04330708661417323, + "grad_norm": 20.03360939025879, + "learning_rate": 3.8767395626242556e-08, + "loss": 4.9849, + "step": 44 + }, + { + "epoch": 0.04429133858267716, + "grad_norm": 23.202022552490234, + "learning_rate": 3.976143141153082e-08, + "loss": 4.9534, + "step": 45 + }, + { + "epoch": 0.045275590551181105, + "grad_norm": 18.500751495361328, + "learning_rate": 4.0755467196819094e-08, + "loss": 4.7506, + "step": 46 + }, + { + "epoch": 0.04625984251968504, + "grad_norm": 15.16248607635498, + "learning_rate": 4.174950298210736e-08, + "loss": 4.6928, + "step": 47 + }, + { + "epoch": 0.047244094488188976, + "grad_norm": 51.08195114135742, + "learning_rate": 4.274353876739563e-08, + "loss": 21.0191, + "step": 48 + }, + { + "epoch": 0.04822834645669291, + "grad_norm": 20.750757217407227, + "learning_rate": 4.373757455268391e-08, + "loss": 4.5581, + "step": 49 + }, + { + "epoch": 0.04921259842519685, + "grad_norm": 26.147449493408203, + "learning_rate": 4.4731610337972176e-08, + "loss": 5.8754, + "step": 50 + }, + { + "epoch": 0.05019685039370079, + "grad_norm": 18.754613876342773, + "learning_rate": 4.572564612326045e-08, + "loss": 4.4697, + "step": 51 + }, + { + "epoch": 0.051181102362204724, + "grad_norm": 23.842952728271484, + "learning_rate": 4.6719681908548713e-08, + "loss": 5.1178, + "step": 52 + }, + { + "epoch": 0.05216535433070866, + "grad_norm": 21.579486846923828, + "learning_rate": 4.7713717693836986e-08, + "loss": 4.7223, + "step": 53 + }, + { + "epoch": 0.0531496062992126, + "grad_norm": 49.253963470458984, + "learning_rate": 4.870775347912525e-08, + "loss": 20.0171, + "step": 54 + }, + { + "epoch": 0.054133858267716536, + "grad_norm": 18.999666213989258, + "learning_rate": 4.970178926441353e-08, + "loss": 4.5361, + "step": 55 + }, + { + "epoch": 0.05511811023622047, + "grad_norm": 16.33707618713379, + "learning_rate": 5.06958250497018e-08, + "loss": 4.2409, + "step": 56 + }, + { + "epoch": 0.05610236220472441, + "grad_norm": 13.874579429626465, + "learning_rate": 5.168986083499007e-08, + "loss": 4.8144, + "step": 57 + }, + { + "epoch": 0.05708661417322835, + "grad_norm": 47.893150329589844, + "learning_rate": 5.268389662027834e-08, + "loss": 18.6889, + "step": 58 + }, + { + "epoch": 0.058070866141732284, + "grad_norm": 18.776273727416992, + "learning_rate": 5.3677932405566605e-08, + "loss": 5.1869, + "step": 59 + }, + { + "epoch": 0.05905511811023622, + "grad_norm": 14.070062637329102, + "learning_rate": 5.467196819085488e-08, + "loss": 3.9955, + "step": 60 + }, + { + "epoch": 0.060039370078740155, + "grad_norm": 16.921262741088867, + "learning_rate": 5.5666003976143156e-08, + "loss": 4.705, + "step": 61 + }, + { + "epoch": 0.0610236220472441, + "grad_norm": 10.885038375854492, + "learning_rate": 5.666003976143142e-08, + "loss": 4.0126, + "step": 62 + }, + { + "epoch": 0.06200787401574803, + "grad_norm": 11.301884651184082, + "learning_rate": 5.7654075546719694e-08, + "loss": 3.3997, + "step": 63 + }, + { + "epoch": 0.06299212598425197, + "grad_norm": 22.516740798950195, + "learning_rate": 5.864811133200796e-08, + "loss": 4.9795, + "step": 64 + }, + { + "epoch": 0.0639763779527559, + "grad_norm": 21.108901977539062, + "learning_rate": 5.964214711729623e-08, + "loss": 5.867, + "step": 65 + }, + { + "epoch": 0.06496062992125984, + "grad_norm": 13.245989799499512, + "learning_rate": 6.06361829025845e-08, + "loss": 4.2919, + "step": 66 + }, + { + "epoch": 0.06594488188976377, + "grad_norm": 10.895593643188477, + "learning_rate": 6.163021868787278e-08, + "loss": 4.4194, + "step": 67 + }, + { + "epoch": 0.06692913385826772, + "grad_norm": 13.352555274963379, + "learning_rate": 6.262425447316104e-08, + "loss": 4.2895, + "step": 68 + }, + { + "epoch": 0.06791338582677166, + "grad_norm": 40.99150085449219, + "learning_rate": 6.361829025844931e-08, + "loss": 16.8972, + "step": 69 + }, + { + "epoch": 0.0688976377952756, + "grad_norm": 11.698071479797363, + "learning_rate": 6.461232604373759e-08, + "loss": 4.6964, + "step": 70 + }, + { + "epoch": 0.06988188976377953, + "grad_norm": 11.680269241333008, + "learning_rate": 6.560636182902586e-08, + "loss": 3.9401, + "step": 71 + }, + { + "epoch": 0.07086614173228346, + "grad_norm": 12.034364700317383, + "learning_rate": 6.660039761431412e-08, + "loss": 3.9605, + "step": 72 + }, + { + "epoch": 0.0718503937007874, + "grad_norm": 11.984710693359375, + "learning_rate": 6.75944333996024e-08, + "loss": 4.1512, + "step": 73 + }, + { + "epoch": 0.07283464566929133, + "grad_norm": 40.74935531616211, + "learning_rate": 6.858846918489067e-08, + "loss": 16.3885, + "step": 74 + }, + { + "epoch": 0.07381889763779527, + "grad_norm": 11.025054931640625, + "learning_rate": 6.958250497017893e-08, + "loss": 3.7214, + "step": 75 + }, + { + "epoch": 0.07480314960629922, + "grad_norm": 11.088836669921875, + "learning_rate": 7.057654075546721e-08, + "loss": 3.6038, + "step": 76 + }, + { + "epoch": 0.07578740157480315, + "grad_norm": 11.580099105834961, + "learning_rate": 7.157057654075548e-08, + "loss": 4.1485, + "step": 77 + }, + { + "epoch": 0.07677165354330709, + "grad_norm": 10.454608917236328, + "learning_rate": 7.256461232604374e-08, + "loss": 3.7848, + "step": 78 + }, + { + "epoch": 0.07775590551181102, + "grad_norm": 10.972256660461426, + "learning_rate": 7.355864811133201e-08, + "loss": 4.679, + "step": 79 + }, + { + "epoch": 0.07874015748031496, + "grad_norm": 42.540199279785156, + "learning_rate": 7.455268389662029e-08, + "loss": 14.9568, + "step": 80 + }, + { + "epoch": 0.0797244094488189, + "grad_norm": 11.498126029968262, + "learning_rate": 7.554671968190855e-08, + "loss": 4.1834, + "step": 81 + }, + { + "epoch": 0.08070866141732283, + "grad_norm": 11.960761070251465, + "learning_rate": 7.654075546719683e-08, + "loss": 4.6593, + "step": 82 + }, + { + "epoch": 0.08169291338582677, + "grad_norm": 10.582605361938477, + "learning_rate": 7.753479125248511e-08, + "loss": 3.5834, + "step": 83 + }, + { + "epoch": 0.08267716535433071, + "grad_norm": 10.250467300415039, + "learning_rate": 7.852882703777338e-08, + "loss": 3.9454, + "step": 84 + }, + { + "epoch": 0.08366141732283465, + "grad_norm": 13.08010482788086, + "learning_rate": 7.952286282306164e-08, + "loss": 4.1069, + "step": 85 + }, + { + "epoch": 0.08464566929133858, + "grad_norm": 11.765287399291992, + "learning_rate": 8.051689860834992e-08, + "loss": 3.8427, + "step": 86 + }, + { + "epoch": 0.08562992125984252, + "grad_norm": 11.19498062133789, + "learning_rate": 8.151093439363819e-08, + "loss": 4.1385, + "step": 87 + }, + { + "epoch": 0.08661417322834646, + "grad_norm": 9.966585159301758, + "learning_rate": 8.250497017892645e-08, + "loss": 3.678, + "step": 88 + }, + { + "epoch": 0.08759842519685039, + "grad_norm": 10.848848342895508, + "learning_rate": 8.349900596421472e-08, + "loss": 4.6092, + "step": 89 + }, + { + "epoch": 0.08858267716535433, + "grad_norm": 42.184574127197266, + "learning_rate": 8.4493041749503e-08, + "loss": 13.5522, + "step": 90 + }, + { + "epoch": 0.08956692913385826, + "grad_norm": 10.029223442077637, + "learning_rate": 8.548707753479126e-08, + "loss": 3.091, + "step": 91 + }, + { + "epoch": 0.09055118110236221, + "grad_norm": 13.168388366699219, + "learning_rate": 8.648111332007953e-08, + "loss": 4.1669, + "step": 92 + }, + { + "epoch": 0.09153543307086615, + "grad_norm": 9.128764152526855, + "learning_rate": 8.747514910536782e-08, + "loss": 3.9135, + "step": 93 + }, + { + "epoch": 0.09251968503937008, + "grad_norm": 41.40175247192383, + "learning_rate": 8.846918489065609e-08, + "loss": 13.7737, + "step": 94 + }, + { + "epoch": 0.09350393700787402, + "grad_norm": 10.502514839172363, + "learning_rate": 8.946322067594435e-08, + "loss": 3.7046, + "step": 95 + }, + { + "epoch": 0.09448818897637795, + "grad_norm": 10.247618675231934, + "learning_rate": 9.045725646123262e-08, + "loss": 4.2823, + "step": 96 + }, + { + "epoch": 0.09547244094488189, + "grad_norm": 9.737785339355469, + "learning_rate": 9.14512922465209e-08, + "loss": 3.5464, + "step": 97 + }, + { + "epoch": 0.09645669291338582, + "grad_norm": 10.351418495178223, + "learning_rate": 9.244532803180916e-08, + "loss": 4.3904, + "step": 98 + }, + { + "epoch": 0.09744094488188976, + "grad_norm": 10.671058654785156, + "learning_rate": 9.343936381709743e-08, + "loss": 4.3634, + "step": 99 + }, + { + "epoch": 0.0984251968503937, + "grad_norm": 8.69608211517334, + "learning_rate": 9.44333996023857e-08, + "loss": 3.974, + "step": 100 + }, + { + "epoch": 0.09940944881889764, + "grad_norm": 9.65053939819336, + "learning_rate": 9.542743538767397e-08, + "loss": 3.2875, + "step": 101 + }, + { + "epoch": 0.10039370078740158, + "grad_norm": 9.247035026550293, + "learning_rate": 9.642147117296224e-08, + "loss": 3.7778, + "step": 102 + }, + { + "epoch": 0.10137795275590551, + "grad_norm": 9.037029266357422, + "learning_rate": 9.74155069582505e-08, + "loss": 3.6514, + "step": 103 + }, + { + "epoch": 0.10236220472440945, + "grad_norm": 12.288069725036621, + "learning_rate": 9.840954274353878e-08, + "loss": 4.0938, + "step": 104 + }, + { + "epoch": 0.10334645669291338, + "grad_norm": 10.267972946166992, + "learning_rate": 9.940357852882706e-08, + "loss": 3.7483, + "step": 105 + }, + { + "epoch": 0.10433070866141732, + "grad_norm": 12.335816383361816, + "learning_rate": 1.0039761431411533e-07, + "loss": 3.857, + "step": 106 + }, + { + "epoch": 0.10531496062992125, + "grad_norm": 10.682831764221191, + "learning_rate": 1.013916500994036e-07, + "loss": 4.01, + "step": 107 + }, + { + "epoch": 0.1062992125984252, + "grad_norm": 9.247139930725098, + "learning_rate": 1.0238568588469187e-07, + "loss": 3.0828, + "step": 108 + }, + { + "epoch": 0.10728346456692914, + "grad_norm": 9.375234603881836, + "learning_rate": 1.0337972166998014e-07, + "loss": 3.5559, + "step": 109 + }, + { + "epoch": 0.10826771653543307, + "grad_norm": 47.49584197998047, + "learning_rate": 1.043737574552684e-07, + "loss": 13.298, + "step": 110 + }, + { + "epoch": 0.10925196850393701, + "grad_norm": 10.562832832336426, + "learning_rate": 1.0536779324055668e-07, + "loss": 3.8423, + "step": 111 + }, + { + "epoch": 0.11023622047244094, + "grad_norm": 9.938478469848633, + "learning_rate": 1.0636182902584495e-07, + "loss": 3.7067, + "step": 112 + }, + { + "epoch": 0.11122047244094488, + "grad_norm": 11.084013938903809, + "learning_rate": 1.0735586481113321e-07, + "loss": 3.6174, + "step": 113 + }, + { + "epoch": 0.11220472440944881, + "grad_norm": 11.869967460632324, + "learning_rate": 1.0834990059642149e-07, + "loss": 4.1381, + "step": 114 + }, + { + "epoch": 0.11318897637795275, + "grad_norm": 37.53839111328125, + "learning_rate": 1.0934393638170976e-07, + "loss": 5.5301, + "step": 115 + }, + { + "epoch": 0.1141732283464567, + "grad_norm": 10.753473281860352, + "learning_rate": 1.1033797216699802e-07, + "loss": 3.8449, + "step": 116 + }, + { + "epoch": 0.11515748031496063, + "grad_norm": 10.626410484313965, + "learning_rate": 1.1133200795228631e-07, + "loss": 3.2364, + "step": 117 + }, + { + "epoch": 0.11614173228346457, + "grad_norm": 11.42529296875, + "learning_rate": 1.1232604373757458e-07, + "loss": 3.875, + "step": 118 + }, + { + "epoch": 0.1171259842519685, + "grad_norm": 50.451011657714844, + "learning_rate": 1.1332007952286284e-07, + "loss": 14.9283, + "step": 119 + }, + { + "epoch": 0.11811023622047244, + "grad_norm": 13.83673095703125, + "learning_rate": 1.1431411530815111e-07, + "loss": 3.4704, + "step": 120 + }, + { + "epoch": 0.11909448818897637, + "grad_norm": 10.942705154418945, + "learning_rate": 1.1530815109343939e-07, + "loss": 3.6903, + "step": 121 + }, + { + "epoch": 0.12007874015748031, + "grad_norm": 12.80109691619873, + "learning_rate": 1.1630218687872765e-07, + "loss": 3.6518, + "step": 122 + }, + { + "epoch": 0.12106299212598425, + "grad_norm": 11.662338256835938, + "learning_rate": 1.1729622266401592e-07, + "loss": 3.6259, + "step": 123 + }, + { + "epoch": 0.1220472440944882, + "grad_norm": 53.99936294555664, + "learning_rate": 1.182902584493042e-07, + "loss": 14.6407, + "step": 124 + }, + { + "epoch": 0.12303149606299213, + "grad_norm": 9.35714340209961, + "learning_rate": 1.1928429423459245e-07, + "loss": 3.6638, + "step": 125 + }, + { + "epoch": 0.12401574803149606, + "grad_norm": 10.13734245300293, + "learning_rate": 1.2027833001988073e-07, + "loss": 3.5541, + "step": 126 + }, + { + "epoch": 0.125, + "grad_norm": 45.38465118408203, + "learning_rate": 1.21272365805169e-07, + "loss": 13.1452, + "step": 127 + }, + { + "epoch": 0.12598425196850394, + "grad_norm": 9.38457202911377, + "learning_rate": 1.222664015904573e-07, + "loss": 3.2458, + "step": 128 + }, + { + "epoch": 0.12696850393700787, + "grad_norm": 44.96736526489258, + "learning_rate": 1.2326043737574557e-07, + "loss": 12.2948, + "step": 129 + }, + { + "epoch": 0.1279527559055118, + "grad_norm": 9.680144309997559, + "learning_rate": 1.2425447316103382e-07, + "loss": 3.1747, + "step": 130 + }, + { + "epoch": 0.12893700787401574, + "grad_norm": 10.51301097869873, + "learning_rate": 1.2524850894632207e-07, + "loss": 3.2047, + "step": 131 + }, + { + "epoch": 0.12992125984251968, + "grad_norm": 10.548707008361816, + "learning_rate": 1.2624254473161035e-07, + "loss": 3.3273, + "step": 132 + }, + { + "epoch": 0.1309055118110236, + "grad_norm": 10.188769340515137, + "learning_rate": 1.2723658051689863e-07, + "loss": 3.7355, + "step": 133 + }, + { + "epoch": 0.13188976377952755, + "grad_norm": 9.908833503723145, + "learning_rate": 1.282306163021869e-07, + "loss": 3.4228, + "step": 134 + }, + { + "epoch": 0.1328740157480315, + "grad_norm": 11.374958038330078, + "learning_rate": 1.2922465208747519e-07, + "loss": 3.5149, + "step": 135 + }, + { + "epoch": 0.13385826771653545, + "grad_norm": 10.55904483795166, + "learning_rate": 1.3021868787276344e-07, + "loss": 3.1236, + "step": 136 + }, + { + "epoch": 0.13484251968503938, + "grad_norm": 55.37184524536133, + "learning_rate": 1.3121272365805172e-07, + "loss": 12.2603, + "step": 137 + }, + { + "epoch": 0.13582677165354332, + "grad_norm": 77.44969940185547, + "learning_rate": 1.3220675944333997e-07, + "loss": 12.36, + "step": 138 + }, + { + "epoch": 0.13681102362204725, + "grad_norm": 10.958036422729492, + "learning_rate": 1.3320079522862825e-07, + "loss": 3.4953, + "step": 139 + }, + { + "epoch": 0.1377952755905512, + "grad_norm": 10.853155136108398, + "learning_rate": 1.3419483101391653e-07, + "loss": 3.7484, + "step": 140 + }, + { + "epoch": 0.13877952755905512, + "grad_norm": 10.227787017822266, + "learning_rate": 1.351888667992048e-07, + "loss": 3.6234, + "step": 141 + }, + { + "epoch": 0.13976377952755906, + "grad_norm": 37.7499885559082, + "learning_rate": 1.3618290258449306e-07, + "loss": 11.2821, + "step": 142 + }, + { + "epoch": 0.140748031496063, + "grad_norm": 9.38892650604248, + "learning_rate": 1.3717693836978134e-07, + "loss": 3.0984, + "step": 143 + }, + { + "epoch": 0.14173228346456693, + "grad_norm": 10.987704277038574, + "learning_rate": 1.381709741550696e-07, + "loss": 3.5274, + "step": 144 + }, + { + "epoch": 0.14271653543307086, + "grad_norm": 11.594810485839844, + "learning_rate": 1.3916500994035787e-07, + "loss": 3.3123, + "step": 145 + }, + { + "epoch": 0.1437007874015748, + "grad_norm": 10.136295318603516, + "learning_rate": 1.4015904572564615e-07, + "loss": 3.4145, + "step": 146 + }, + { + "epoch": 0.14468503937007873, + "grad_norm": 37.52333068847656, + "learning_rate": 1.4115308151093443e-07, + "loss": 10.6589, + "step": 147 + }, + { + "epoch": 0.14566929133858267, + "grad_norm": 9.571504592895508, + "learning_rate": 1.421471172962227e-07, + "loss": 3.1544, + "step": 148 + }, + { + "epoch": 0.1466535433070866, + "grad_norm": 9.913719177246094, + "learning_rate": 1.4314115308151096e-07, + "loss": 3.5877, + "step": 149 + }, + { + "epoch": 0.14763779527559054, + "grad_norm": 10.658916473388672, + "learning_rate": 1.4413518886679924e-07, + "loss": 3.2683, + "step": 150 + }, + { + "epoch": 0.1486220472440945, + "grad_norm": 11.442712783813477, + "learning_rate": 1.451292246520875e-07, + "loss": 3.5216, + "step": 151 + }, + { + "epoch": 0.14960629921259844, + "grad_norm": 51.636688232421875, + "learning_rate": 1.4612326043737577e-07, + "loss": 10.2844, + "step": 152 + }, + { + "epoch": 0.15059055118110237, + "grad_norm": 9.432707786560059, + "learning_rate": 1.4711729622266402e-07, + "loss": 3.6084, + "step": 153 + }, + { + "epoch": 0.15059055118110237, + "eval_Qnli-dev_cosine_accuracy": 0.5859375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.9280097484588623, + "eval_Qnli-dev_cosine_ap": 0.5559868473974543, + "eval_Qnli-dev_cosine_f1": 0.629878869448183, + "eval_Qnli-dev_cosine_f1_threshold": 0.8130671977996826, + "eval_Qnli-dev_cosine_precision": 0.46153846153846156, + "eval_Qnli-dev_cosine_recall": 0.9915254237288136, + "eval_Qnli-dev_dot_accuracy": 0.5859375, + "eval_Qnli-dev_dot_accuracy_threshold": 712.7769775390625, + "eval_Qnli-dev_dot_ap": 0.5559079473320865, + "eval_Qnli-dev_dot_f1": 0.629878869448183, + "eval_Qnli-dev_dot_f1_threshold": 624.4847412109375, + "eval_Qnli-dev_dot_precision": 0.46153846153846156, + "eval_Qnli-dev_dot_recall": 0.9915254237288136, + "eval_Qnli-dev_euclidean_accuracy": 0.5859375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 10.516043663024902, + "eval_Qnli-dev_euclidean_ap": 0.5559637033258996, + "eval_Qnli-dev_euclidean_f1": 0.629878869448183, + "eval_Qnli-dev_euclidean_f1_threshold": 16.945524215698242, + "eval_Qnli-dev_euclidean_precision": 0.46153846153846156, + "eval_Qnli-dev_euclidean_recall": 0.9915254237288136, + "eval_Qnli-dev_manhattan_accuracy": 0.60546875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 191.7781982421875, + "eval_Qnli-dev_manhattan_ap": 0.5815151188080563, + "eval_Qnli-dev_manhattan_f1": 0.6331658291457287, + "eval_Qnli-dev_manhattan_f1_threshold": 215.1607666015625, + "eval_Qnli-dev_manhattan_precision": 0.5235457063711911, + "eval_Qnli-dev_manhattan_recall": 0.8008474576271186, + "eval_Qnli-dev_max_accuracy": 0.60546875, + "eval_Qnli-dev_max_accuracy_threshold": 712.7769775390625, + "eval_Qnli-dev_max_ap": 0.5815151188080563, + "eval_Qnli-dev_max_f1": 0.6331658291457287, + "eval_Qnli-dev_max_f1_threshold": 624.4847412109375, + "eval_Qnli-dev_max_precision": 0.5235457063711911, + "eval_Qnli-dev_max_recall": 0.9915254237288136, + "eval_allNLI-dev_cosine_accuracy": 0.669921875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9783592820167542, + "eval_allNLI-dev_cosine_ap": 0.41204201087169523, + "eval_allNLI-dev_cosine_f1": 0.5058479532163743, + "eval_allNLI-dev_cosine_f1_threshold": 0.8176803588867188, + "eval_allNLI-dev_cosine_precision": 0.3385518590998043, + "eval_allNLI-dev_cosine_recall": 1.0, + "eval_allNLI-dev_dot_accuracy": 0.669921875, + "eval_allNLI-dev_dot_accuracy_threshold": 751.4491577148438, + "eval_allNLI-dev_dot_ap": 0.4121114399865795, + "eval_allNLI-dev_dot_f1": 0.5058479532163743, + "eval_allNLI-dev_dot_f1_threshold": 628.0350952148438, + "eval_allNLI-dev_dot_precision": 0.3385518590998043, + "eval_allNLI-dev_dot_recall": 1.0, + "eval_allNLI-dev_euclidean_accuracy": 0.669921875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 5.765616416931152, + "eval_allNLI-dev_euclidean_ap": 0.41201664900644336, + "eval_allNLI-dev_euclidean_f1": 0.5058479532163743, + "eval_allNLI-dev_euclidean_f1_threshold": 16.7352294921875, + "eval_allNLI-dev_euclidean_precision": 0.3385518590998043, + "eval_allNLI-dev_euclidean_recall": 1.0, + "eval_allNLI-dev_manhattan_accuracy": 0.669921875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 128.160400390625, + "eval_allNLI-dev_manhattan_ap": 0.410947097259041, + "eval_allNLI-dev_manhattan_f1": 0.5071942446043165, + "eval_allNLI-dev_manhattan_f1_threshold": 212.47366333007812, + "eval_allNLI-dev_manhattan_precision": 0.3681462140992167, + "eval_allNLI-dev_manhattan_recall": 0.815028901734104, + "eval_allNLI-dev_max_accuracy": 0.669921875, + "eval_allNLI-dev_max_accuracy_threshold": 751.4491577148438, + "eval_allNLI-dev_max_ap": 0.4121114399865795, + "eval_allNLI-dev_max_f1": 0.5071942446043165, + "eval_allNLI-dev_max_f1_threshold": 628.0350952148438, + "eval_allNLI-dev_max_precision": 0.3681462140992167, + "eval_allNLI-dev_max_recall": 1.0, + "eval_loss": 3.319310426712036, + "eval_runtime": 54.6245, + "eval_samples_per_second": 30.463, + "eval_sequential_score": 0.5815151188080563, + "eval_steps_per_second": 0.128, + "eval_sts-test_pearson_cosine": 0.19018447741866837, + "eval_sts-test_pearson_dot": 0.19002495383670479, + "eval_sts-test_pearson_euclidean": 0.2209959418841952, + "eval_sts-test_pearson_manhattan": 0.24668057542280933, + "eval_sts-test_pearson_max": 0.24668057542280933, + "eval_sts-test_spearman_cosine": 0.25171604480037757, + "eval_sts-test_spearman_dot": 0.25161920307072655, + "eval_sts-test_spearman_euclidean": 0.2517444171000997, + "eval_sts-test_spearman_manhattan": 0.27227065716753845, + "eval_sts-test_spearman_max": 0.27227065716753845, + "step": 153 + }, + { + "epoch": 0.1515748031496063, + "grad_norm": 10.549654960632324, + "learning_rate": 1.4811133200795232e-07, + "loss": 3.7025, + "step": 154 + }, + { + "epoch": 0.15255905511811024, + "grad_norm": 10.626328468322754, + "learning_rate": 1.4910536779324058e-07, + "loss": 3.2625, + "step": 155 + }, + { + "epoch": 0.15354330708661418, + "grad_norm": 10.456741333007812, + "learning_rate": 1.5009940357852886e-07, + "loss": 3.0366, + "step": 156 + }, + { + "epoch": 0.1545275590551181, + "grad_norm": 38.44537353515625, + "learning_rate": 1.510934393638171e-07, + "loss": 9.183, + "step": 157 + }, + { + "epoch": 0.15551181102362205, + "grad_norm": 10.580129623413086, + "learning_rate": 1.5208747514910539e-07, + "loss": 3.5048, + "step": 158 + }, + { + "epoch": 0.15649606299212598, + "grad_norm": 42.2885627746582, + "learning_rate": 1.5308151093439367e-07, + "loss": 9.8654, + "step": 159 + }, + { + "epoch": 0.15748031496062992, + "grad_norm": 8.579136848449707, + "learning_rate": 1.5407554671968192e-07, + "loss": 3.0458, + "step": 160 + }, + { + "epoch": 0.15846456692913385, + "grad_norm": 12.40068244934082, + "learning_rate": 1.5506958250497022e-07, + "loss": 2.7602, + "step": 161 + }, + { + "epoch": 0.1594488188976378, + "grad_norm": 10.646333694458008, + "learning_rate": 1.5606361829025848e-07, + "loss": 3.4081, + "step": 162 + }, + { + "epoch": 0.16043307086614172, + "grad_norm": 12.419439315795898, + "learning_rate": 1.5705765407554675e-07, + "loss": 3.6212, + "step": 163 + }, + { + "epoch": 0.16141732283464566, + "grad_norm": 11.597086906433105, + "learning_rate": 1.58051689860835e-07, + "loss": 3.625, + "step": 164 + }, + { + "epoch": 0.1624015748031496, + "grad_norm": 15.256538391113281, + "learning_rate": 1.5904572564612329e-07, + "loss": 3.9519, + "step": 165 + }, + { + "epoch": 0.16338582677165353, + "grad_norm": 69.25948333740234, + "learning_rate": 1.6003976143141154e-07, + "loss": 9.3491, + "step": 166 + }, + { + "epoch": 0.1643700787401575, + "grad_norm": 10.0520658493042, + "learning_rate": 1.6103379721669984e-07, + "loss": 2.7756, + "step": 167 + }, + { + "epoch": 0.16535433070866143, + "grad_norm": 12.943922996520996, + "learning_rate": 1.620278330019881e-07, + "loss": 2.9741, + "step": 168 + }, + { + "epoch": 0.16633858267716536, + "grad_norm": 13.181682586669922, + "learning_rate": 1.6302186878727637e-07, + "loss": 3.6422, + "step": 169 + }, + { + "epoch": 0.1673228346456693, + "grad_norm": 11.436388969421387, + "learning_rate": 1.6401590457256465e-07, + "loss": 3.3038, + "step": 170 + }, + { + "epoch": 0.16830708661417323, + "grad_norm": 10.708134651184082, + "learning_rate": 1.650099403578529e-07, + "loss": 2.8385, + "step": 171 + }, + { + "epoch": 0.16929133858267717, + "grad_norm": 13.435418128967285, + "learning_rate": 1.6600397614314118e-07, + "loss": 3.1081, + "step": 172 + }, + { + "epoch": 0.1702755905511811, + "grad_norm": 11.510961532592773, + "learning_rate": 1.6699801192842944e-07, + "loss": 3.4009, + "step": 173 + }, + { + "epoch": 0.17125984251968504, + "grad_norm": 14.482342720031738, + "learning_rate": 1.6799204771371774e-07, + "loss": 2.7715, + "step": 174 + }, + { + "epoch": 0.17224409448818898, + "grad_norm": 11.370920181274414, + "learning_rate": 1.68986083499006e-07, + "loss": 3.1816, + "step": 175 + }, + { + "epoch": 0.1732283464566929, + "grad_norm": 14.27595329284668, + "learning_rate": 1.6998011928429427e-07, + "loss": 3.0486, + "step": 176 + }, + { + "epoch": 0.17421259842519685, + "grad_norm": 11.878148078918457, + "learning_rate": 1.7097415506958253e-07, + "loss": 2.8444, + "step": 177 + }, + { + "epoch": 0.17519685039370078, + "grad_norm": 115.54975128173828, + "learning_rate": 1.719681908548708e-07, + "loss": 12.6448, + "step": 178 + }, + { + "epoch": 0.17618110236220472, + "grad_norm": 12.08621597290039, + "learning_rate": 1.7296222664015906e-07, + "loss": 3.0867, + "step": 179 + }, + { + "epoch": 0.17716535433070865, + "grad_norm": 11.200992584228516, + "learning_rate": 1.7395626242544734e-07, + "loss": 2.8874, + "step": 180 + }, + { + "epoch": 0.1781496062992126, + "grad_norm": 10.824020385742188, + "learning_rate": 1.7495029821073564e-07, + "loss": 2.8767, + "step": 181 + }, + { + "epoch": 0.17913385826771652, + "grad_norm": 12.792362213134766, + "learning_rate": 1.759443339960239e-07, + "loss": 3.1107, + "step": 182 + }, + { + "epoch": 0.18011811023622049, + "grad_norm": 12.027408599853516, + "learning_rate": 1.7693836978131217e-07, + "loss": 2.3596, + "step": 183 + }, + { + "epoch": 0.18110236220472442, + "grad_norm": 12.46939468383789, + "learning_rate": 1.7793240556660042e-07, + "loss": 2.6352, + "step": 184 + }, + { + "epoch": 0.18208661417322836, + "grad_norm": 12.76395320892334, + "learning_rate": 1.789264413518887e-07, + "loss": 2.8008, + "step": 185 + }, + { + "epoch": 0.1830708661417323, + "grad_norm": 10.95465087890625, + "learning_rate": 1.7992047713717695e-07, + "loss": 2.7295, + "step": 186 + }, + { + "epoch": 0.18405511811023623, + "grad_norm": 75.06632995605469, + "learning_rate": 1.8091451292246523e-07, + "loss": 12.2772, + "step": 187 + }, + { + "epoch": 0.18503937007874016, + "grad_norm": 10.998300552368164, + "learning_rate": 1.819085487077535e-07, + "loss": 2.6023, + "step": 188 + }, + { + "epoch": 0.1860236220472441, + "grad_norm": 13.164251327514648, + "learning_rate": 1.829025844930418e-07, + "loss": 3.27, + "step": 189 + }, + { + "epoch": 0.18700787401574803, + "grad_norm": 12.379849433898926, + "learning_rate": 1.8389662027833004e-07, + "loss": 3.1023, + "step": 190 + }, + { + "epoch": 0.18799212598425197, + "grad_norm": 123.68094635009766, + "learning_rate": 1.8489065606361832e-07, + "loss": 15.7886, + "step": 191 + }, + { + "epoch": 0.1889763779527559, + "grad_norm": 12.211271286010742, + "learning_rate": 1.8588469184890657e-07, + "loss": 2.8468, + "step": 192 + }, + { + "epoch": 0.18996062992125984, + "grad_norm": 13.436681747436523, + "learning_rate": 1.8687872763419485e-07, + "loss": 2.1157, + "step": 193 + }, + { + "epoch": 0.19094488188976377, + "grad_norm": 71.8902816772461, + "learning_rate": 1.8787276341948313e-07, + "loss": 12.1402, + "step": 194 + }, + { + "epoch": 0.1919291338582677, + "grad_norm": 59.65647888183594, + "learning_rate": 1.888667992047714e-07, + "loss": 10.5477, + "step": 195 + }, + { + "epoch": 0.19291338582677164, + "grad_norm": 11.768324851989746, + "learning_rate": 1.898608349900597e-07, + "loss": 2.8112, + "step": 196 + }, + { + "epoch": 0.19389763779527558, + "grad_norm": 76.99214172363281, + "learning_rate": 1.9085487077534794e-07, + "loss": 11.2557, + "step": 197 + }, + { + "epoch": 0.19488188976377951, + "grad_norm": 13.789910316467285, + "learning_rate": 1.9184890656063622e-07, + "loss": 2.629, + "step": 198 + }, + { + "epoch": 0.19586614173228348, + "grad_norm": 13.719752311706543, + "learning_rate": 1.9284294234592447e-07, + "loss": 3.1383, + "step": 199 + }, + { + "epoch": 0.1968503937007874, + "grad_norm": 58.55073547363281, + "learning_rate": 1.9383697813121275e-07, + "loss": 9.7592, + "step": 200 + }, + { + "epoch": 0.19783464566929135, + "grad_norm": 10.811251640319824, + "learning_rate": 1.94831013916501e-07, + "loss": 2.9253, + "step": 201 + }, + { + "epoch": 0.19881889763779528, + "grad_norm": 13.063098907470703, + "learning_rate": 1.958250497017893e-07, + "loss": 2.9712, + "step": 202 + }, + { + "epoch": 0.19980314960629922, + "grad_norm": 12.683143615722656, + "learning_rate": 1.9681908548707756e-07, + "loss": 2.8585, + "step": 203 + }, + { + "epoch": 0.20078740157480315, + "grad_norm": 11.156440734863281, + "learning_rate": 1.9781312127236584e-07, + "loss": 3.082, + "step": 204 + }, + { + "epoch": 0.2017716535433071, + "grad_norm": 11.65755844116211, + "learning_rate": 1.9880715705765412e-07, + "loss": 3.2078, + "step": 205 + }, + { + "epoch": 0.20275590551181102, + "grad_norm": 11.212764739990234, + "learning_rate": 1.9980119284294237e-07, + "loss": 2.4837, + "step": 206 + }, + { + "epoch": 0.20374015748031496, + "grad_norm": 47.730979919433594, + "learning_rate": 2.0079522862823065e-07, + "loss": 7.7377, + "step": 207 + }, + { + "epoch": 0.2047244094488189, + "grad_norm": 12.969826698303223, + "learning_rate": 2.017892644135189e-07, + "loss": 2.3949, + "step": 208 + }, + { + "epoch": 0.20570866141732283, + "grad_norm": 10.695167541503906, + "learning_rate": 2.027833001988072e-07, + "loss": 2.7936, + "step": 209 + }, + { + "epoch": 0.20669291338582677, + "grad_norm": 11.578596115112305, + "learning_rate": 2.0377733598409546e-07, + "loss": 2.5481, + "step": 210 + }, + { + "epoch": 0.2076771653543307, + "grad_norm": 12.902640342712402, + "learning_rate": 2.0477137176938374e-07, + "loss": 2.8108, + "step": 211 + }, + { + "epoch": 0.20866141732283464, + "grad_norm": 13.556937217712402, + "learning_rate": 2.05765407554672e-07, + "loss": 2.6916, + "step": 212 + }, + { + "epoch": 0.20964566929133857, + "grad_norm": 14.155546188354492, + "learning_rate": 2.0675944333996027e-07, + "loss": 3.3112, + "step": 213 + }, + { + "epoch": 0.2106299212598425, + "grad_norm": 12.20670223236084, + "learning_rate": 2.0775347912524852e-07, + "loss": 2.6884, + "step": 214 + }, + { + "epoch": 0.21161417322834647, + "grad_norm": 13.744552612304688, + "learning_rate": 2.087475149105368e-07, + "loss": 2.4656, + "step": 215 + }, + { + "epoch": 0.2125984251968504, + "grad_norm": 12.515267372131348, + "learning_rate": 2.097415506958251e-07, + "loss": 2.3028, + "step": 216 + }, + { + "epoch": 0.21358267716535434, + "grad_norm": 12.110260009765625, + "learning_rate": 2.1073558648111336e-07, + "loss": 2.4515, + "step": 217 + }, + { + "epoch": 0.21456692913385828, + "grad_norm": 13.38983154296875, + "learning_rate": 2.1172962226640164e-07, + "loss": 2.7888, + "step": 218 + }, + { + "epoch": 0.2155511811023622, + "grad_norm": 12.12717342376709, + "learning_rate": 2.127236580516899e-07, + "loss": 2.7798, + "step": 219 + }, + { + "epoch": 0.21653543307086615, + "grad_norm": 11.084267616271973, + "learning_rate": 2.1371769383697817e-07, + "loss": 2.2994, + "step": 220 + }, + { + "epoch": 0.21751968503937008, + "grad_norm": 13.082998275756836, + "learning_rate": 2.1471172962226642e-07, + "loss": 2.4806, + "step": 221 + }, + { + "epoch": 0.21850393700787402, + "grad_norm": 13.084571838378906, + "learning_rate": 2.1570576540755473e-07, + "loss": 2.6355, + "step": 222 + }, + { + "epoch": 0.21948818897637795, + "grad_norm": 12.001659393310547, + "learning_rate": 2.1669980119284298e-07, + "loss": 2.4232, + "step": 223 + }, + { + "epoch": 0.2204724409448819, + "grad_norm": 12.470407485961914, + "learning_rate": 2.1769383697813126e-07, + "loss": 2.6152, + "step": 224 + }, + { + "epoch": 0.22145669291338582, + "grad_norm": 11.995153427124023, + "learning_rate": 2.186878727634195e-07, + "loss": 2.6337, + "step": 225 + }, + { + "epoch": 0.22244094488188976, + "grad_norm": 14.227544784545898, + "learning_rate": 2.196819085487078e-07, + "loss": 2.4903, + "step": 226 + }, + { + "epoch": 0.2234251968503937, + "grad_norm": 12.475507736206055, + "learning_rate": 2.2067594433399604e-07, + "loss": 2.3831, + "step": 227 + }, + { + "epoch": 0.22440944881889763, + "grad_norm": 13.871504783630371, + "learning_rate": 2.2166998011928432e-07, + "loss": 2.1024, + "step": 228 + }, + { + "epoch": 0.22539370078740156, + "grad_norm": 14.166851043701172, + "learning_rate": 2.2266401590457263e-07, + "loss": 2.7021, + "step": 229 + }, + { + "epoch": 0.2263779527559055, + "grad_norm": 13.170698165893555, + "learning_rate": 2.2365805168986088e-07, + "loss": 2.3359, + "step": 230 + }, + { + "epoch": 0.22736220472440946, + "grad_norm": 12.220829963684082, + "learning_rate": 2.2465208747514916e-07, + "loss": 1.9508, + "step": 231 + }, + { + "epoch": 0.2283464566929134, + "grad_norm": 14.62851619720459, + "learning_rate": 2.256461232604374e-07, + "loss": 2.061, + "step": 232 + }, + { + "epoch": 0.22933070866141733, + "grad_norm": 117.03069305419922, + "learning_rate": 2.266401590457257e-07, + "loss": 15.199, + "step": 233 + }, + { + "epoch": 0.23031496062992127, + "grad_norm": 111.26533508300781, + "learning_rate": 2.2763419483101394e-07, + "loss": 14.708, + "step": 234 + }, + { + "epoch": 0.2312992125984252, + "grad_norm": 107.53176879882812, + "learning_rate": 2.2862823061630222e-07, + "loss": 15.589, + "step": 235 + }, + { + "epoch": 0.23228346456692914, + "grad_norm": 11.968624114990234, + "learning_rate": 2.296222664015905e-07, + "loss": 2.4485, + "step": 236 + }, + { + "epoch": 0.23326771653543307, + "grad_norm": 14.77911376953125, + "learning_rate": 2.3061630218687878e-07, + "loss": 2.1559, + "step": 237 + }, + { + "epoch": 0.234251968503937, + "grad_norm": 17.22453498840332, + "learning_rate": 2.3161033797216703e-07, + "loss": 3.0109, + "step": 238 + }, + { + "epoch": 0.23523622047244094, + "grad_norm": 15.654526710510254, + "learning_rate": 2.326043737574553e-07, + "loss": 2.2779, + "step": 239 + }, + { + "epoch": 0.23622047244094488, + "grad_norm": 79.38912200927734, + "learning_rate": 2.335984095427436e-07, + "loss": 12.1054, + "step": 240 + }, + { + "epoch": 0.2372047244094488, + "grad_norm": 13.8206787109375, + "learning_rate": 2.3459244532803184e-07, + "loss": 2.7905, + "step": 241 + }, + { + "epoch": 0.23818897637795275, + "grad_norm": 13.186335563659668, + "learning_rate": 2.3558648111332012e-07, + "loss": 2.2739, + "step": 242 + }, + { + "epoch": 0.23917322834645668, + "grad_norm": 10.965237617492676, + "learning_rate": 2.365805168986084e-07, + "loss": 2.0147, + "step": 243 + }, + { + "epoch": 0.24015748031496062, + "grad_norm": 65.3853530883789, + "learning_rate": 2.3757455268389668e-07, + "loss": 10.8463, + "step": 244 + }, + { + "epoch": 0.24114173228346455, + "grad_norm": 11.911761283874512, + "learning_rate": 2.385685884691849e-07, + "loss": 2.4448, + "step": 245 + }, + { + "epoch": 0.2421259842519685, + "grad_norm": 11.165079116821289, + "learning_rate": 2.395626242544732e-07, + "loss": 2.1458, + "step": 246 + }, + { + "epoch": 0.24311023622047245, + "grad_norm": 83.87309265136719, + "learning_rate": 2.4055666003976146e-07, + "loss": 11.9076, + "step": 247 + }, + { + "epoch": 0.2440944881889764, + "grad_norm": 48.82333755493164, + "learning_rate": 2.4155069582504976e-07, + "loss": 4.2321, + "step": 248 + }, + { + "epoch": 0.24507874015748032, + "grad_norm": 62.72069549560547, + "learning_rate": 2.42544731610338e-07, + "loss": 9.6525, + "step": 249 + }, + { + "epoch": 0.24606299212598426, + "grad_norm": 54.45147705078125, + "learning_rate": 2.4353876739562627e-07, + "loss": 9.5191, + "step": 250 + }, + { + "epoch": 0.2470472440944882, + "grad_norm": 13.214591979980469, + "learning_rate": 2.445328031809146e-07, + "loss": 3.0668, + "step": 251 + }, + { + "epoch": 0.24803149606299213, + "grad_norm": 36.288726806640625, + "learning_rate": 2.4552683896620283e-07, + "loss": 7.0423, + "step": 252 + }, + { + "epoch": 0.24901574803149606, + "grad_norm": 11.362079620361328, + "learning_rate": 2.4652087475149113e-07, + "loss": 2.4391, + "step": 253 + }, + { + "epoch": 0.25, + "grad_norm": 11.57868766784668, + "learning_rate": 2.475149105367794e-07, + "loss": 2.0855, + "step": 254 + }, + { + "epoch": 0.25098425196850394, + "grad_norm": 10.305681228637695, + "learning_rate": 2.4850894632206764e-07, + "loss": 2.5628, + "step": 255 + }, + { + "epoch": 0.25196850393700787, + "grad_norm": 34.71120834350586, + "learning_rate": 2.495029821073559e-07, + "loss": 6.4129, + "step": 256 + }, + { + "epoch": 0.2529527559055118, + "grad_norm": 12.710351943969727, + "learning_rate": 2.5049701789264414e-07, + "loss": 3.0284, + "step": 257 + }, + { + "epoch": 0.25393700787401574, + "grad_norm": 11.934942245483398, + "learning_rate": 2.5149105367793245e-07, + "loss": 3.0393, + "step": 258 + }, + { + "epoch": 0.2549212598425197, + "grad_norm": 33.399269104003906, + "learning_rate": 2.524850894632207e-07, + "loss": 5.7205, + "step": 259 + }, + { + "epoch": 0.2559055118110236, + "grad_norm": 10.524690628051758, + "learning_rate": 2.53479125248509e-07, + "loss": 2.8864, + "step": 260 + }, + { + "epoch": 0.25688976377952755, + "grad_norm": 10.373811721801758, + "learning_rate": 2.5447316103379726e-07, + "loss": 2.1274, + "step": 261 + }, + { + "epoch": 0.2578740157480315, + "grad_norm": 10.531494140625, + "learning_rate": 2.554671968190855e-07, + "loss": 2.3253, + "step": 262 + }, + { + "epoch": 0.2588582677165354, + "grad_norm": 12.033869743347168, + "learning_rate": 2.564612326043738e-07, + "loss": 2.7868, + "step": 263 + }, + { + "epoch": 0.25984251968503935, + "grad_norm": 11.448728561401367, + "learning_rate": 2.5745526838966207e-07, + "loss": 2.5113, + "step": 264 + }, + { + "epoch": 0.2608267716535433, + "grad_norm": 31.349775314331055, + "learning_rate": 2.5844930417495037e-07, + "loss": 5.5063, + "step": 265 + }, + { + "epoch": 0.2618110236220472, + "grad_norm": 10.772862434387207, + "learning_rate": 2.5944333996023857e-07, + "loss": 2.5678, + "step": 266 + }, + { + "epoch": 0.26279527559055116, + "grad_norm": 39.487667083740234, + "learning_rate": 2.604373757455269e-07, + "loss": 5.6321, + "step": 267 + }, + { + "epoch": 0.2637795275590551, + "grad_norm": 12.496540069580078, + "learning_rate": 2.614314115308152e-07, + "loss": 2.3219, + "step": 268 + }, + { + "epoch": 0.26476377952755903, + "grad_norm": 9.853267669677734, + "learning_rate": 2.6242544731610343e-07, + "loss": 2.0576, + "step": 269 + }, + { + "epoch": 0.265748031496063, + "grad_norm": 11.759198188781738, + "learning_rate": 2.634194831013917e-07, + "loss": 2.7055, + "step": 270 + }, + { + "epoch": 0.26673228346456695, + "grad_norm": 11.300050735473633, + "learning_rate": 2.6441351888667994e-07, + "loss": 2.2286, + "step": 271 + }, + { + "epoch": 0.2677165354330709, + "grad_norm": 13.588562965393066, + "learning_rate": 2.6540755467196824e-07, + "loss": 2.6628, + "step": 272 + }, + { + "epoch": 0.2687007874015748, + "grad_norm": 37.17454528808594, + "learning_rate": 2.664015904572565e-07, + "loss": 5.2455, + "step": 273 + }, + { + "epoch": 0.26968503937007876, + "grad_norm": 12.299790382385254, + "learning_rate": 2.6739562624254475e-07, + "loss": 2.6966, + "step": 274 + }, + { + "epoch": 0.2706692913385827, + "grad_norm": 12.163739204406738, + "learning_rate": 2.6838966202783305e-07, + "loss": 2.65, + "step": 275 + }, + { + "epoch": 0.27165354330708663, + "grad_norm": 12.674065589904785, + "learning_rate": 2.693836978131213e-07, + "loss": 2.6302, + "step": 276 + }, + { + "epoch": 0.27263779527559057, + "grad_norm": 13.80542278289795, + "learning_rate": 2.703777335984096e-07, + "loss": 2.6564, + "step": 277 + }, + { + "epoch": 0.2736220472440945, + "grad_norm": 11.924317359924316, + "learning_rate": 2.7137176938369786e-07, + "loss": 2.0339, + "step": 278 + }, + { + "epoch": 0.27460629921259844, + "grad_norm": 15.281108856201172, + "learning_rate": 2.723658051689861e-07, + "loss": 2.8719, + "step": 279 + }, + { + "epoch": 0.2755905511811024, + "grad_norm": 14.091357231140137, + "learning_rate": 2.7335984095427437e-07, + "loss": 2.5937, + "step": 280 + }, + { + "epoch": 0.2765748031496063, + "grad_norm": 37.14817428588867, + "learning_rate": 2.743538767395627e-07, + "loss": 5.97, + "step": 281 + }, + { + "epoch": 0.27755905511811024, + "grad_norm": 11.206151962280273, + "learning_rate": 2.75347912524851e-07, + "loss": 2.1888, + "step": 282 + }, + { + "epoch": 0.2785433070866142, + "grad_norm": 11.060794830322266, + "learning_rate": 2.763419483101392e-07, + "loss": 2.4111, + "step": 283 + }, + { + "epoch": 0.2795275590551181, + "grad_norm": 11.33033561706543, + "learning_rate": 2.773359840954275e-07, + "loss": 2.2334, + "step": 284 + }, + { + "epoch": 0.28051181102362205, + "grad_norm": 46.83280563354492, + "learning_rate": 2.7833001988071574e-07, + "loss": 6.7992, + "step": 285 + }, + { + "epoch": 0.281496062992126, + "grad_norm": 12.653908729553223, + "learning_rate": 2.7932405566600404e-07, + "loss": 1.8102, + "step": 286 + }, + { + "epoch": 0.2824803149606299, + "grad_norm": 12.369451522827148, + "learning_rate": 2.803180914512923e-07, + "loss": 2.3559, + "step": 287 + }, + { + "epoch": 0.28346456692913385, + "grad_norm": 47.14701461791992, + "learning_rate": 2.8131212723658055e-07, + "loss": 6.0575, + "step": 288 + }, + { + "epoch": 0.2844488188976378, + "grad_norm": 11.92288875579834, + "learning_rate": 2.8230616302186885e-07, + "loss": 2.3846, + "step": 289 + }, + { + "epoch": 0.2854330708661417, + "grad_norm": 12.775434494018555, + "learning_rate": 2.833001988071571e-07, + "loss": 2.7311, + "step": 290 + }, + { + "epoch": 0.28641732283464566, + "grad_norm": 66.64407348632812, + "learning_rate": 2.842942345924454e-07, + "loss": 7.2351, + "step": 291 + }, + { + "epoch": 0.2874015748031496, + "grad_norm": 12.071573257446289, + "learning_rate": 2.852882703777336e-07, + "loss": 2.3992, + "step": 292 + }, + { + "epoch": 0.28838582677165353, + "grad_norm": 9.866507530212402, + "learning_rate": 2.862823061630219e-07, + "loss": 2.1329, + "step": 293 + }, + { + "epoch": 0.28937007874015747, + "grad_norm": 10.664796829223633, + "learning_rate": 2.8727634194831017e-07, + "loss": 2.0372, + "step": 294 + }, + { + "epoch": 0.2903543307086614, + "grad_norm": 11.258280754089355, + "learning_rate": 2.8827037773359847e-07, + "loss": 1.9959, + "step": 295 + }, + { + "epoch": 0.29133858267716534, + "grad_norm": 11.857529640197754, + "learning_rate": 2.892644135188867e-07, + "loss": 2.5762, + "step": 296 + }, + { + "epoch": 0.29232283464566927, + "grad_norm": 12.668633460998535, + "learning_rate": 2.90258449304175e-07, + "loss": 2.454, + "step": 297 + }, + { + "epoch": 0.2933070866141732, + "grad_norm": 11.908174514770508, + "learning_rate": 2.912524850894633e-07, + "loss": 2.4293, + "step": 298 + }, + { + "epoch": 0.29429133858267714, + "grad_norm": 10.250532150268555, + "learning_rate": 2.9224652087475153e-07, + "loss": 1.6916, + "step": 299 + }, + { + "epoch": 0.2952755905511811, + "grad_norm": 11.219808578491211, + "learning_rate": 2.9324055666003984e-07, + "loss": 2.0638, + "step": 300 + }, + { + "epoch": 0.296259842519685, + "grad_norm": 10.419336318969727, + "learning_rate": 2.9423459244532804e-07, + "loss": 1.967, + "step": 301 + }, + { + "epoch": 0.297244094488189, + "grad_norm": 12.338981628417969, + "learning_rate": 2.9522862823061634e-07, + "loss": 2.3572, + "step": 302 + }, + { + "epoch": 0.29822834645669294, + "grad_norm": 11.970561027526855, + "learning_rate": 2.9622266401590465e-07, + "loss": 2.1485, + "step": 303 + }, + { + "epoch": 0.2992125984251969, + "grad_norm": 52.54031753540039, + "learning_rate": 2.972166998011929e-07, + "loss": 7.7883, + "step": 304 + }, + { + "epoch": 0.3001968503937008, + "grad_norm": 11.679478645324707, + "learning_rate": 2.9821073558648115e-07, + "loss": 2.1426, + "step": 305 + }, + { + "epoch": 0.30118110236220474, + "grad_norm": 10.657000541687012, + "learning_rate": 2.992047713717694e-07, + "loss": 1.7839, + "step": 306 + }, + { + "epoch": 0.30118110236220474, + "eval_Qnli-dev_cosine_accuracy": 0.603515625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.90336012840271, + "eval_Qnli-dev_cosine_ap": 0.5993966680292693, + "eval_Qnli-dev_cosine_f1": 0.637883008356546, + "eval_Qnli-dev_cosine_f1_threshold": 0.7992255687713623, + "eval_Qnli-dev_cosine_precision": 0.475103734439834, + "eval_Qnli-dev_cosine_recall": 0.9703389830508474, + "eval_Qnli-dev_dot_accuracy": 0.603515625, + "eval_Qnli-dev_dot_accuracy_threshold": 693.1090087890625, + "eval_Qnli-dev_dot_ap": 0.5990996370946063, + "eval_Qnli-dev_dot_f1": 0.637883008356546, + "eval_Qnli-dev_dot_f1_threshold": 613.3412475585938, + "eval_Qnli-dev_dot_precision": 0.475103734439834, + "eval_Qnli-dev_dot_recall": 0.9703389830508474, + "eval_Qnli-dev_euclidean_accuracy": 0.603515625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 12.177652359008789, + "eval_Qnli-dev_euclidean_ap": 0.5994114375803649, + "eval_Qnli-dev_euclidean_f1": 0.637883008356546, + "eval_Qnli-dev_euclidean_f1_threshold": 17.55432891845703, + "eval_Qnli-dev_euclidean_precision": 0.475103734439834, + "eval_Qnli-dev_euclidean_recall": 0.9703389830508474, + "eval_Qnli-dev_manhattan_accuracy": 0.623046875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 253.44888305664062, + "eval_Qnli-dev_manhattan_ap": 0.6114921536490292, + "eval_Qnli-dev_manhattan_f1": 0.6529411764705882, + "eval_Qnli-dev_manhattan_f1_threshold": 327.55078125, + "eval_Qnli-dev_manhattan_precision": 0.5, + "eval_Qnli-dev_manhattan_recall": 0.940677966101695, + "eval_Qnli-dev_max_accuracy": 0.623046875, + "eval_Qnli-dev_max_accuracy_threshold": 693.1090087890625, + "eval_Qnli-dev_max_ap": 0.6114921536490292, + "eval_Qnli-dev_max_f1": 0.6529411764705882, + "eval_Qnli-dev_max_f1_threshold": 613.3412475585938, + "eval_Qnli-dev_max_precision": 0.5, + "eval_Qnli-dev_max_recall": 0.9703389830508474, + "eval_allNLI-dev_cosine_accuracy": 0.669921875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9743889570236206, + "eval_allNLI-dev_cosine_ap": 0.41419182158586465, + "eval_allNLI-dev_cosine_f1": 0.5081669691470054, + "eval_allNLI-dev_cosine_f1_threshold": 0.8874048590660095, + "eval_allNLI-dev_cosine_precision": 0.37037037037037035, + "eval_allNLI-dev_cosine_recall": 0.8092485549132948, + "eval_allNLI-dev_dot_accuracy": 0.669921875, + "eval_allNLI-dev_dot_accuracy_threshold": 750.77294921875, + "eval_allNLI-dev_dot_ap": 0.41489360676575493, + "eval_allNLI-dev_dot_f1": 0.5078864353312303, + "eval_allNLI-dev_dot_f1_threshold": 647.291259765625, + "eval_allNLI-dev_dot_precision": 0.3492407809110629, + "eval_allNLI-dev_dot_recall": 0.930635838150289, + "eval_allNLI-dev_euclidean_accuracy": 0.669921875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 6.270920753479004, + "eval_allNLI-dev_euclidean_ap": 0.4142019252328734, + "eval_allNLI-dev_euclidean_f1": 0.5081669691470054, + "eval_allNLI-dev_euclidean_f1_threshold": 13.145881652832031, + "eval_allNLI-dev_euclidean_precision": 0.37037037037037035, + "eval_allNLI-dev_euclidean_recall": 0.8092485549132948, + "eval_allNLI-dev_manhattan_accuracy": 0.66796875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 108.75921630859375, + "eval_allNLI-dev_manhattan_ap": 0.41121266093840314, + "eval_allNLI-dev_manhattan_f1": 0.5156794425087108, + "eval_allNLI-dev_manhattan_f1_threshold": 267.4344787597656, + "eval_allNLI-dev_manhattan_precision": 0.3690773067331671, + "eval_allNLI-dev_manhattan_recall": 0.8554913294797688, + "eval_allNLI-dev_max_accuracy": 0.669921875, + "eval_allNLI-dev_max_accuracy_threshold": 750.77294921875, + "eval_allNLI-dev_max_ap": 0.41489360676575493, + "eval_allNLI-dev_max_f1": 0.5156794425087108, + "eval_allNLI-dev_max_f1_threshold": 647.291259765625, + "eval_allNLI-dev_max_precision": 0.37037037037037035, + "eval_allNLI-dev_max_recall": 0.930635838150289, + "eval_loss": 2.382014036178589, + "eval_runtime": 54.4017, + "eval_samples_per_second": 30.587, + "eval_sequential_score": 0.6114921536490292, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.21416737125119828, + "eval_sts-test_pearson_dot": 0.2130784047531441, + "eval_sts-test_pearson_euclidean": 0.24175068780760803, + "eval_sts-test_pearson_manhattan": 0.2519623781736086, + "eval_sts-test_pearson_max": 0.2519623781736086, + "eval_sts-test_spearman_cosine": 0.2708540559881937, + "eval_sts-test_spearman_dot": 0.26943982662570265, + "eval_sts-test_spearman_euclidean": 0.2709061912077147, + "eval_sts-test_spearman_manhattan": 0.27978654201570125, + "eval_sts-test_spearman_max": 0.27978654201570125, + "step": 306 + }, + { + "epoch": 0.3021653543307087, + "grad_norm": 10.734902381896973, + "learning_rate": 3.001988071570577e-07, + "loss": 1.7923, + "step": 307 + }, + { + "epoch": 0.3031496062992126, + "grad_norm": 10.291141510009766, + "learning_rate": 3.01192842942346e-07, + "loss": 1.8325, + "step": 308 + }, + { + "epoch": 0.30413385826771655, + "grad_norm": 10.940272331237793, + "learning_rate": 3.021868787276342e-07, + "loss": 1.8853, + "step": 309 + }, + { + "epoch": 0.3051181102362205, + "grad_norm": 79.87450408935547, + "learning_rate": 3.031809145129225e-07, + "loss": 9.9739, + "step": 310 + }, + { + "epoch": 0.3061023622047244, + "grad_norm": 58.47280502319336, + "learning_rate": 3.0417495029821077e-07, + "loss": 7.9147, + "step": 311 + }, + { + "epoch": 0.30708661417322836, + "grad_norm": 11.367313385009766, + "learning_rate": 3.051689860834991e-07, + "loss": 2.2901, + "step": 312 + }, + { + "epoch": 0.3080708661417323, + "grad_norm": 11.76492691040039, + "learning_rate": 3.0616302186878733e-07, + "loss": 2.2322, + "step": 313 + }, + { + "epoch": 0.3090551181102362, + "grad_norm": 10.923507690429688, + "learning_rate": 3.071570576540756e-07, + "loss": 1.7877, + "step": 314 + }, + { + "epoch": 0.31003937007874016, + "grad_norm": 9.680418014526367, + "learning_rate": 3.0815109343936384e-07, + "loss": 1.7058, + "step": 315 + }, + { + "epoch": 0.3110236220472441, + "grad_norm": 52.868309020996094, + "learning_rate": 3.0914512922465214e-07, + "loss": 7.9885, + "step": 316 + }, + { + "epoch": 0.31200787401574803, + "grad_norm": 11.308423042297363, + "learning_rate": 3.1013916500994045e-07, + "loss": 2.0856, + "step": 317 + }, + { + "epoch": 0.31299212598425197, + "grad_norm": 13.043806076049805, + "learning_rate": 3.1113320079522865e-07, + "loss": 2.0736, + "step": 318 + }, + { + "epoch": 0.3139763779527559, + "grad_norm": 13.645163536071777, + "learning_rate": 3.1212723658051695e-07, + "loss": 2.4477, + "step": 319 + }, + { + "epoch": 0.31496062992125984, + "grad_norm": 11.428119659423828, + "learning_rate": 3.131212723658052e-07, + "loss": 2.3418, + "step": 320 + }, + { + "epoch": 0.3159448818897638, + "grad_norm": 12.059881210327148, + "learning_rate": 3.141153081510935e-07, + "loss": 2.113, + "step": 321 + }, + { + "epoch": 0.3169291338582677, + "grad_norm": 62.59613800048828, + "learning_rate": 3.151093439363818e-07, + "loss": 9.2682, + "step": 322 + }, + { + "epoch": 0.31791338582677164, + "grad_norm": 61.35698699951172, + "learning_rate": 3.1610337972167e-07, + "loss": 8.0044, + "step": 323 + }, + { + "epoch": 0.3188976377952756, + "grad_norm": 45.25782775878906, + "learning_rate": 3.170974155069583e-07, + "loss": 6.6853, + "step": 324 + }, + { + "epoch": 0.3198818897637795, + "grad_norm": 10.384099006652832, + "learning_rate": 3.1809145129224657e-07, + "loss": 1.9131, + "step": 325 + }, + { + "epoch": 0.32086614173228345, + "grad_norm": 10.898626327514648, + "learning_rate": 3.190854870775349e-07, + "loss": 1.9983, + "step": 326 + }, + { + "epoch": 0.3218503937007874, + "grad_norm": 11.038727760314941, + "learning_rate": 3.200795228628231e-07, + "loss": 2.1731, + "step": 327 + }, + { + "epoch": 0.3228346456692913, + "grad_norm": 11.289586067199707, + "learning_rate": 3.210735586481114e-07, + "loss": 1.6724, + "step": 328 + }, + { + "epoch": 0.32381889763779526, + "grad_norm": 10.893710136413574, + "learning_rate": 3.220675944333997e-07, + "loss": 1.9086, + "step": 329 + }, + { + "epoch": 0.3248031496062992, + "grad_norm": 9.912519454956055, + "learning_rate": 3.2306163021868794e-07, + "loss": 1.9539, + "step": 330 + }, + { + "epoch": 0.3257874015748031, + "grad_norm": 47.91545104980469, + "learning_rate": 3.240556660039762e-07, + "loss": 5.7975, + "step": 331 + }, + { + "epoch": 0.32677165354330706, + "grad_norm": 10.154540061950684, + "learning_rate": 3.2504970178926444e-07, + "loss": 1.8287, + "step": 332 + }, + { + "epoch": 0.327755905511811, + "grad_norm": 9.187210083007812, + "learning_rate": 3.2604373757455275e-07, + "loss": 1.8369, + "step": 333 + }, + { + "epoch": 0.328740157480315, + "grad_norm": 9.3264799118042, + "learning_rate": 3.27037773359841e-07, + "loss": 1.4491, + "step": 334 + }, + { + "epoch": 0.3297244094488189, + "grad_norm": 11.264083862304688, + "learning_rate": 3.280318091451293e-07, + "loss": 1.9481, + "step": 335 + }, + { + "epoch": 0.33070866141732286, + "grad_norm": 56.167755126953125, + "learning_rate": 3.2902584493041756e-07, + "loss": 6.3526, + "step": 336 + }, + { + "epoch": 0.3316929133858268, + "grad_norm": 11.542598724365234, + "learning_rate": 3.300198807157058e-07, + "loss": 2.2364, + "step": 337 + }, + { + "epoch": 0.33267716535433073, + "grad_norm": 9.792985916137695, + "learning_rate": 3.310139165009941e-07, + "loss": 1.7361, + "step": 338 + }, + { + "epoch": 0.33366141732283466, + "grad_norm": 11.404271125793457, + "learning_rate": 3.3200795228628237e-07, + "loss": 2.0384, + "step": 339 + }, + { + "epoch": 0.3346456692913386, + "grad_norm": 10.791166305541992, + "learning_rate": 3.330019880715706e-07, + "loss": 1.7119, + "step": 340 + }, + { + "epoch": 0.33562992125984253, + "grad_norm": 11.242233276367188, + "learning_rate": 3.3399602385685887e-07, + "loss": 2.0796, + "step": 341 + }, + { + "epoch": 0.33661417322834647, + "grad_norm": 8.66325569152832, + "learning_rate": 3.349900596421472e-07, + "loss": 1.423, + "step": 342 + }, + { + "epoch": 0.3375984251968504, + "grad_norm": 49.26806640625, + "learning_rate": 3.359840954274355e-07, + "loss": 7.3127, + "step": 343 + }, + { + "epoch": 0.33858267716535434, + "grad_norm": 12.85938549041748, + "learning_rate": 3.369781312127237e-07, + "loss": 1.7333, + "step": 344 + }, + { + "epoch": 0.3395669291338583, + "grad_norm": 11.912487030029297, + "learning_rate": 3.37972166998012e-07, + "loss": 2.0256, + "step": 345 + }, + { + "epoch": 0.3405511811023622, + "grad_norm": 13.381354331970215, + "learning_rate": 3.3896620278330024e-07, + "loss": 2.281, + "step": 346 + }, + { + "epoch": 0.34153543307086615, + "grad_norm": 11.3228178024292, + "learning_rate": 3.3996023856858855e-07, + "loss": 2.6135, + "step": 347 + }, + { + "epoch": 0.3425196850393701, + "grad_norm": 13.237906455993652, + "learning_rate": 3.409542743538768e-07, + "loss": 1.7188, + "step": 348 + }, + { + "epoch": 0.343503937007874, + "grad_norm": 51.69926071166992, + "learning_rate": 3.4194831013916505e-07, + "loss": 7.2075, + "step": 349 + }, + { + "epoch": 0.34448818897637795, + "grad_norm": 14.641998291015625, + "learning_rate": 3.4294234592445336e-07, + "loss": 2.2859, + "step": 350 + }, + { + "epoch": 0.3454724409448819, + "grad_norm": 12.565581321716309, + "learning_rate": 3.439363817097416e-07, + "loss": 2.3042, + "step": 351 + }, + { + "epoch": 0.3464566929133858, + "grad_norm": 11.288939476013184, + "learning_rate": 3.449304174950299e-07, + "loss": 1.8944, + "step": 352 + }, + { + "epoch": 0.34744094488188976, + "grad_norm": 35.340904235839844, + "learning_rate": 3.459244532803181e-07, + "loss": 3.2227, + "step": 353 + }, + { + "epoch": 0.3484251968503937, + "grad_norm": 12.464967727661133, + "learning_rate": 3.469184890656064e-07, + "loss": 1.9706, + "step": 354 + }, + { + "epoch": 0.3494094488188976, + "grad_norm": 10.959451675415039, + "learning_rate": 3.4791252485089467e-07, + "loss": 1.4694, + "step": 355 + }, + { + "epoch": 0.35039370078740156, + "grad_norm": 13.121545791625977, + "learning_rate": 3.48906560636183e-07, + "loss": 2.6948, + "step": 356 + }, + { + "epoch": 0.3513779527559055, + "grad_norm": 52.15006637573242, + "learning_rate": 3.499005964214713e-07, + "loss": 7.2386, + "step": 357 + }, + { + "epoch": 0.35236220472440943, + "grad_norm": 53.013832092285156, + "learning_rate": 3.508946322067595e-07, + "loss": 6.7146, + "step": 358 + }, + { + "epoch": 0.35334645669291337, + "grad_norm": 56.041343688964844, + "learning_rate": 3.518886679920478e-07, + "loss": 7.1446, + "step": 359 + }, + { + "epoch": 0.3543307086614173, + "grad_norm": 13.036025047302246, + "learning_rate": 3.5288270377733604e-07, + "loss": 2.1246, + "step": 360 + }, + { + "epoch": 0.35531496062992124, + "grad_norm": 12.321011543273926, + "learning_rate": 3.5387673956262434e-07, + "loss": 2.2676, + "step": 361 + }, + { + "epoch": 0.3562992125984252, + "grad_norm": 12.051976203918457, + "learning_rate": 3.5487077534791254e-07, + "loss": 1.8415, + "step": 362 + }, + { + "epoch": 0.3572834645669291, + "grad_norm": 40.2060661315918, + "learning_rate": 3.5586481113320085e-07, + "loss": 5.9398, + "step": 363 + }, + { + "epoch": 0.35826771653543305, + "grad_norm": 11.134186744689941, + "learning_rate": 3.5685884691848915e-07, + "loss": 2.022, + "step": 364 + }, + { + "epoch": 0.359251968503937, + "grad_norm": 13.448582649230957, + "learning_rate": 3.578528827037774e-07, + "loss": 2.4997, + "step": 365 + }, + { + "epoch": 0.36023622047244097, + "grad_norm": 9.906368255615234, + "learning_rate": 3.5884691848906566e-07, + "loss": 1.8627, + "step": 366 + }, + { + "epoch": 0.3612204724409449, + "grad_norm": 12.35049819946289, + "learning_rate": 3.598409542743539e-07, + "loss": 2.0428, + "step": 367 + }, + { + "epoch": 0.36220472440944884, + "grad_norm": 11.315701484680176, + "learning_rate": 3.608349900596422e-07, + "loss": 1.9337, + "step": 368 + }, + { + "epoch": 0.3631889763779528, + "grad_norm": 12.36326789855957, + "learning_rate": 3.6182902584493047e-07, + "loss": 2.2796, + "step": 369 + }, + { + "epoch": 0.3641732283464567, + "grad_norm": 12.894477844238281, + "learning_rate": 3.6282306163021877e-07, + "loss": 2.1442, + "step": 370 + }, + { + "epoch": 0.36515748031496065, + "grad_norm": 11.680022239685059, + "learning_rate": 3.63817097415507e-07, + "loss": 1.856, + "step": 371 + }, + { + "epoch": 0.3661417322834646, + "grad_norm": 35.735713958740234, + "learning_rate": 3.648111332007953e-07, + "loss": 5.1339, + "step": 372 + }, + { + "epoch": 0.3671259842519685, + "grad_norm": 14.422959327697754, + "learning_rate": 3.658051689860836e-07, + "loss": 2.5143, + "step": 373 + }, + { + "epoch": 0.36811023622047245, + "grad_norm": 10.153562545776367, + "learning_rate": 3.6679920477137184e-07, + "loss": 1.3353, + "step": 374 + }, + { + "epoch": 0.3690944881889764, + "grad_norm": 9.560186386108398, + "learning_rate": 3.677932405566601e-07, + "loss": 1.1054, + "step": 375 + }, + { + "epoch": 0.3700787401574803, + "grad_norm": 52.342227935791016, + "learning_rate": 3.6878727634194834e-07, + "loss": 5.1893, + "step": 376 + }, + { + "epoch": 0.37106299212598426, + "grad_norm": 9.996565818786621, + "learning_rate": 3.6978131212723665e-07, + "loss": 1.5849, + "step": 377 + }, + { + "epoch": 0.3720472440944882, + "grad_norm": 10.249340057373047, + "learning_rate": 3.7077534791252495e-07, + "loss": 1.7363, + "step": 378 + }, + { + "epoch": 0.37303149606299213, + "grad_norm": 54.82731628417969, + "learning_rate": 3.7176938369781315e-07, + "loss": 6.3385, + "step": 379 + }, + { + "epoch": 0.37401574803149606, + "grad_norm": 11.697839736938477, + "learning_rate": 3.7276341948310145e-07, + "loss": 1.7896, + "step": 380 + }, + { + "epoch": 0.375, + "grad_norm": 10.580652236938477, + "learning_rate": 3.737574552683897e-07, + "loss": 1.8278, + "step": 381 + }, + { + "epoch": 0.37598425196850394, + "grad_norm": 9.671079635620117, + "learning_rate": 3.74751491053678e-07, + "loss": 1.3783, + "step": 382 + }, + { + "epoch": 0.37696850393700787, + "grad_norm": 10.716227531433105, + "learning_rate": 3.7574552683896626e-07, + "loss": 1.5611, + "step": 383 + }, + { + "epoch": 0.3779527559055118, + "grad_norm": 50.10944366455078, + "learning_rate": 3.767395626242545e-07, + "loss": 6.6285, + "step": 384 + }, + { + "epoch": 0.37893700787401574, + "grad_norm": 11.59723949432373, + "learning_rate": 3.777335984095428e-07, + "loss": 1.54, + "step": 385 + }, + { + "epoch": 0.3799212598425197, + "grad_norm": 11.77585220336914, + "learning_rate": 3.787276341948311e-07, + "loss": 1.4216, + "step": 386 + }, + { + "epoch": 0.3809055118110236, + "grad_norm": 45.65110397338867, + "learning_rate": 3.797216699801194e-07, + "loss": 5.1756, + "step": 387 + }, + { + "epoch": 0.38188976377952755, + "grad_norm": 13.375621795654297, + "learning_rate": 3.807157057654076e-07, + "loss": 2.2009, + "step": 388 + }, + { + "epoch": 0.3828740157480315, + "grad_norm": 10.92442512512207, + "learning_rate": 3.817097415506959e-07, + "loss": 1.8205, + "step": 389 + }, + { + "epoch": 0.3838582677165354, + "grad_norm": 14.326327323913574, + "learning_rate": 3.8270377733598414e-07, + "loss": 2.437, + "step": 390 + }, + { + "epoch": 0.38484251968503935, + "grad_norm": 12.294797897338867, + "learning_rate": 3.8369781312127244e-07, + "loss": 1.9934, + "step": 391 + }, + { + "epoch": 0.3858267716535433, + "grad_norm": 11.739425659179688, + "learning_rate": 3.8469184890656075e-07, + "loss": 1.9723, + "step": 392 + }, + { + "epoch": 0.3868110236220472, + "grad_norm": 12.739344596862793, + "learning_rate": 3.8568588469184895e-07, + "loss": 1.7051, + "step": 393 + }, + { + "epoch": 0.38779527559055116, + "grad_norm": 11.261693954467773, + "learning_rate": 3.8667992047713725e-07, + "loss": 1.7474, + "step": 394 + }, + { + "epoch": 0.3887795275590551, + "grad_norm": 39.55622100830078, + "learning_rate": 3.876739562624255e-07, + "loss": 5.1513, + "step": 395 + }, + { + "epoch": 0.38976377952755903, + "grad_norm": 12.222193717956543, + "learning_rate": 3.886679920477138e-07, + "loss": 1.6779, + "step": 396 + }, + { + "epoch": 0.390748031496063, + "grad_norm": 9.99728012084961, + "learning_rate": 3.89662027833002e-07, + "loss": 1.4055, + "step": 397 + }, + { + "epoch": 0.39173228346456695, + "grad_norm": 11.975881576538086, + "learning_rate": 3.906560636182903e-07, + "loss": 1.6386, + "step": 398 + }, + { + "epoch": 0.3927165354330709, + "grad_norm": 10.65119457244873, + "learning_rate": 3.916500994035786e-07, + "loss": 1.2141, + "step": 399 + }, + { + "epoch": 0.3937007874015748, + "grad_norm": 45.11775588989258, + "learning_rate": 3.9264413518886687e-07, + "loss": 6.2437, + "step": 400 + }, + { + "epoch": 0.39468503937007876, + "grad_norm": 11.461906433105469, + "learning_rate": 3.936381709741551e-07, + "loss": 1.3342, + "step": 401 + }, + { + "epoch": 0.3956692913385827, + "grad_norm": 11.590656280517578, + "learning_rate": 3.946322067594434e-07, + "loss": 1.5153, + "step": 402 + }, + { + "epoch": 0.39665354330708663, + "grad_norm": 11.251949310302734, + "learning_rate": 3.956262425447317e-07, + "loss": 1.4861, + "step": 403 + }, + { + "epoch": 0.39763779527559057, + "grad_norm": 11.35289478302002, + "learning_rate": 3.9662027833001993e-07, + "loss": 1.6596, + "step": 404 + }, + { + "epoch": 0.3986220472440945, + "grad_norm": 57.254600524902344, + "learning_rate": 3.9761431411530824e-07, + "loss": 6.3475, + "step": 405 + }, + { + "epoch": 0.39960629921259844, + "grad_norm": 13.950235366821289, + "learning_rate": 3.986083499005965e-07, + "loss": 1.7669, + "step": 406 + }, + { + "epoch": 0.4005905511811024, + "grad_norm": 12.42131519317627, + "learning_rate": 3.9960238568588474e-07, + "loss": 1.5388, + "step": 407 + }, + { + "epoch": 0.4015748031496063, + "grad_norm": 11.60079288482666, + "learning_rate": 4.0059642147117305e-07, + "loss": 1.6276, + "step": 408 + }, + { + "epoch": 0.40255905511811024, + "grad_norm": 12.122014999389648, + "learning_rate": 4.015904572564613e-07, + "loss": 1.4932, + "step": 409 + }, + { + "epoch": 0.4035433070866142, + "grad_norm": 63.547462463378906, + "learning_rate": 4.0258449304174955e-07, + "loss": 7.097, + "step": 410 + }, + { + "epoch": 0.4045275590551181, + "grad_norm": 11.58280086517334, + "learning_rate": 4.035785288270378e-07, + "loss": 1.4621, + "step": 411 + }, + { + "epoch": 0.40551181102362205, + "grad_norm": 44.8221549987793, + "learning_rate": 4.045725646123261e-07, + "loss": 5.6547, + "step": 412 + }, + { + "epoch": 0.406496062992126, + "grad_norm": 44.49589157104492, + "learning_rate": 4.055666003976144e-07, + "loss": 5.7338, + "step": 413 + }, + { + "epoch": 0.4074803149606299, + "grad_norm": 11.479046821594238, + "learning_rate": 4.065606361829026e-07, + "loss": 1.2834, + "step": 414 + }, + { + "epoch": 0.40846456692913385, + "grad_norm": 12.855639457702637, + "learning_rate": 4.075546719681909e-07, + "loss": 1.8426, + "step": 415 + }, + { + "epoch": 0.4094488188976378, + "grad_norm": 14.07402229309082, + "learning_rate": 4.085487077534792e-07, + "loss": 1.6923, + "step": 416 + }, + { + "epoch": 0.4104330708661417, + "grad_norm": 11.612604141235352, + "learning_rate": 4.095427435387675e-07, + "loss": 1.7412, + "step": 417 + }, + { + "epoch": 0.41141732283464566, + "grad_norm": 13.344563484191895, + "learning_rate": 4.1053677932405573e-07, + "loss": 1.5589, + "step": 418 + }, + { + "epoch": 0.4124015748031496, + "grad_norm": 11.442570686340332, + "learning_rate": 4.11530815109344e-07, + "loss": 1.2435, + "step": 419 + }, + { + "epoch": 0.41338582677165353, + "grad_norm": 12.59207820892334, + "learning_rate": 4.125248508946323e-07, + "loss": 1.9214, + "step": 420 + }, + { + "epoch": 0.41437007874015747, + "grad_norm": 10.514676094055176, + "learning_rate": 4.1351888667992054e-07, + "loss": 1.432, + "step": 421 + }, + { + "epoch": 0.4153543307086614, + "grad_norm": 11.39521598815918, + "learning_rate": 4.1451292246520885e-07, + "loss": 1.643, + "step": 422 + }, + { + "epoch": 0.41633858267716534, + "grad_norm": 12.807600021362305, + "learning_rate": 4.1550695825049705e-07, + "loss": 1.7221, + "step": 423 + }, + { + "epoch": 0.41732283464566927, + "grad_norm": 11.254819869995117, + "learning_rate": 4.1650099403578535e-07, + "loss": 1.6566, + "step": 424 + }, + { + "epoch": 0.4183070866141732, + "grad_norm": 47.7083625793457, + "learning_rate": 4.174950298210736e-07, + "loss": 5.7164, + "step": 425 + }, + { + "epoch": 0.41929133858267714, + "grad_norm": 10.80250072479248, + "learning_rate": 4.184890656063619e-07, + "loss": 1.3455, + "step": 426 + }, + { + "epoch": 0.4202755905511811, + "grad_norm": 10.916194915771484, + "learning_rate": 4.194831013916502e-07, + "loss": 1.2572, + "step": 427 + }, + { + "epoch": 0.421259842519685, + "grad_norm": 11.888566017150879, + "learning_rate": 4.204771371769384e-07, + "loss": 1.3673, + "step": 428 + }, + { + "epoch": 0.422244094488189, + "grad_norm": 37.930355072021484, + "learning_rate": 4.214711729622267e-07, + "loss": 4.6922, + "step": 429 + }, + { + "epoch": 0.42322834645669294, + "grad_norm": 13.136495590209961, + "learning_rate": 4.2246520874751497e-07, + "loss": 1.5377, + "step": 430 + }, + { + "epoch": 0.4242125984251969, + "grad_norm": 50.90897750854492, + "learning_rate": 4.234592445328033e-07, + "loss": 6.0487, + "step": 431 + }, + { + "epoch": 0.4251968503937008, + "grad_norm": 57.34842300415039, + "learning_rate": 4.244532803180915e-07, + "loss": 6.3802, + "step": 432 + }, + { + "epoch": 0.42618110236220474, + "grad_norm": 44.35916519165039, + "learning_rate": 4.254473161033798e-07, + "loss": 5.7241, + "step": 433 + }, + { + "epoch": 0.4271653543307087, + "grad_norm": 39.626949310302734, + "learning_rate": 4.264413518886681e-07, + "loss": 4.4259, + "step": 434 + }, + { + "epoch": 0.4281496062992126, + "grad_norm": 11.860918045043945, + "learning_rate": 4.2743538767395634e-07, + "loss": 1.4789, + "step": 435 + }, + { + "epoch": 0.42913385826771655, + "grad_norm": 13.54674243927002, + "learning_rate": 4.284294234592446e-07, + "loss": 2.1121, + "step": 436 + }, + { + "epoch": 0.4301181102362205, + "grad_norm": 10.985451698303223, + "learning_rate": 4.2942345924453284e-07, + "loss": 1.5906, + "step": 437 + }, + { + "epoch": 0.4311023622047244, + "grad_norm": 12.662684440612793, + "learning_rate": 4.3041749502982115e-07, + "loss": 1.6671, + "step": 438 + }, + { + "epoch": 0.43208661417322836, + "grad_norm": 19.172922134399414, + "learning_rate": 4.3141153081510945e-07, + "loss": 1.5505, + "step": 439 + }, + { + "epoch": 0.4330708661417323, + "grad_norm": 28.78730583190918, + "learning_rate": 4.324055666003977e-07, + "loss": 3.6253, + "step": 440 + }, + { + "epoch": 0.4340551181102362, + "grad_norm": 12.55512523651123, + "learning_rate": 4.3339960238568596e-07, + "loss": 2.1088, + "step": 441 + }, + { + "epoch": 0.43503937007874016, + "grad_norm": 10.92598819732666, + "learning_rate": 4.343936381709742e-07, + "loss": 1.8861, + "step": 442 + }, + { + "epoch": 0.4360236220472441, + "grad_norm": 10.909286499023438, + "learning_rate": 4.353876739562625e-07, + "loss": 1.377, + "step": 443 + }, + { + "epoch": 0.43700787401574803, + "grad_norm": 11.649571418762207, + "learning_rate": 4.3638170974155077e-07, + "loss": 1.7832, + "step": 444 + }, + { + "epoch": 0.43799212598425197, + "grad_norm": 12.003958702087402, + "learning_rate": 4.37375745526839e-07, + "loss": 2.1039, + "step": 445 + }, + { + "epoch": 0.4389763779527559, + "grad_norm": 11.032736778259277, + "learning_rate": 4.383697813121273e-07, + "loss": 1.5845, + "step": 446 + }, + { + "epoch": 0.43996062992125984, + "grad_norm": 10.186923027038574, + "learning_rate": 4.393638170974156e-07, + "loss": 1.5359, + "step": 447 + }, + { + "epoch": 0.4409448818897638, + "grad_norm": 12.455483436584473, + "learning_rate": 4.403578528827039e-07, + "loss": 2.0012, + "step": 448 + }, + { + "epoch": 0.4419291338582677, + "grad_norm": 29.561321258544922, + "learning_rate": 4.413518886679921e-07, + "loss": 3.5535, + "step": 449 + }, + { + "epoch": 0.44291338582677164, + "grad_norm": 13.571199417114258, + "learning_rate": 4.423459244532804e-07, + "loss": 1.8301, + "step": 450 + }, + { + "epoch": 0.4438976377952756, + "grad_norm": 12.335179328918457, + "learning_rate": 4.4333996023856864e-07, + "loss": 1.5625, + "step": 451 + }, + { + "epoch": 0.4448818897637795, + "grad_norm": 14.14587116241455, + "learning_rate": 4.4433399602385695e-07, + "loss": 1.9825, + "step": 452 + }, + { + "epoch": 0.44586614173228345, + "grad_norm": 11.227643013000488, + "learning_rate": 4.4532803180914525e-07, + "loss": 1.404, + "step": 453 + }, + { + "epoch": 0.4468503937007874, + "grad_norm": 11.070554733276367, + "learning_rate": 4.4632206759443345e-07, + "loss": 1.1998, + "step": 454 + }, + { + "epoch": 0.4478346456692913, + "grad_norm": 35.71782302856445, + "learning_rate": 4.4731610337972176e-07, + "loss": 3.9972, + "step": 455 + }, + { + "epoch": 0.44881889763779526, + "grad_norm": 13.148333549499512, + "learning_rate": 4.4831013916501e-07, + "loss": 2.01, + "step": 456 + }, + { + "epoch": 0.4498031496062992, + "grad_norm": 11.549042701721191, + "learning_rate": 4.493041749502983e-07, + "loss": 1.3631, + "step": 457 + }, + { + "epoch": 0.4507874015748031, + "grad_norm": 12.581208229064941, + "learning_rate": 4.502982107355865e-07, + "loss": 1.3897, + "step": 458 + }, + { + "epoch": 0.45177165354330706, + "grad_norm": 34.60251235961914, + "learning_rate": 4.512922465208748e-07, + "loss": 3.3806, + "step": 459 + }, + { + "epoch": 0.45177165354330706, + "eval_Qnli-dev_cosine_accuracy": 0.623046875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.9098526239395142, + "eval_Qnli-dev_cosine_ap": 0.6152473349893568, + "eval_Qnli-dev_cosine_f1": 0.6391184573002756, + "eval_Qnli-dev_cosine_f1_threshold": 0.7495293617248535, + "eval_Qnli-dev_cosine_precision": 0.47346938775510206, + "eval_Qnli-dev_cosine_recall": 0.9830508474576272, + "eval_Qnli-dev_dot_accuracy": 0.625, + "eval_Qnli-dev_dot_accuracy_threshold": 698.5355224609375, + "eval_Qnli-dev_dot_ap": 0.6174855915379028, + "eval_Qnli-dev_dot_f1": 0.6392318244170097, + "eval_Qnli-dev_dot_f1_threshold": 571.7469482421875, + "eval_Qnli-dev_dot_precision": 0.4726166328600406, + "eval_Qnli-dev_dot_recall": 0.9872881355932204, + "eval_Qnli-dev_euclidean_accuracy": 0.623046875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 11.752104759216309, + "eval_Qnli-dev_euclidean_ap": 0.6152099088218845, + "eval_Qnli-dev_euclidean_f1": 0.6392318244170097, + "eval_Qnli-dev_euclidean_f1_threshold": 19.713125228881836, + "eval_Qnli-dev_euclidean_precision": 0.4726166328600406, + "eval_Qnli-dev_euclidean_recall": 0.9872881355932204, + "eval_Qnli-dev_manhattan_accuracy": 0.62890625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 269.57421875, + "eval_Qnli-dev_manhattan_ap": 0.6311673051130458, + "eval_Qnli-dev_manhattan_f1": 0.6490683229813664, + "eval_Qnli-dev_manhattan_f1_threshold": 338.50018310546875, + "eval_Qnli-dev_manhattan_precision": 0.5122549019607843, + "eval_Qnli-dev_manhattan_recall": 0.885593220338983, + "eval_Qnli-dev_max_accuracy": 0.62890625, + "eval_Qnli-dev_max_accuracy_threshold": 698.5355224609375, + "eval_Qnli-dev_max_ap": 0.6311673051130458, + "eval_Qnli-dev_max_f1": 0.6490683229813664, + "eval_Qnli-dev_max_f1_threshold": 571.7469482421875, + "eval_Qnli-dev_max_precision": 0.5122549019607843, + "eval_Qnli-dev_max_recall": 0.9872881355932204, + "eval_allNLI-dev_cosine_accuracy": 0.673828125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9571062922477722, + "eval_allNLI-dev_cosine_ap": 0.43412785256494396, + "eval_allNLI-dev_cosine_f1": 0.5173116089613035, + "eval_allNLI-dev_cosine_f1_threshold": 0.8916290998458862, + "eval_allNLI-dev_cosine_precision": 0.39937106918238996, + "eval_allNLI-dev_cosine_recall": 0.7341040462427746, + "eval_allNLI-dev_dot_accuracy": 0.671875, + "eval_allNLI-dev_dot_accuracy_threshold": 734.5493774414062, + "eval_allNLI-dev_dot_ap": 0.43404768564943036, + "eval_allNLI-dev_dot_f1": 0.515748031496063, + "eval_allNLI-dev_dot_f1_threshold": 681.0838012695312, + "eval_allNLI-dev_dot_precision": 0.39104477611940297, + "eval_allNLI-dev_dot_recall": 0.7572254335260116, + "eval_allNLI-dev_euclidean_accuracy": 0.673828125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 8.112812042236328, + "eval_allNLI-dev_euclidean_ap": 0.4341037986713034, + "eval_allNLI-dev_euclidean_f1": 0.515212981744422, + "eval_allNLI-dev_euclidean_f1_threshold": 12.893339157104492, + "eval_allNLI-dev_euclidean_precision": 0.396875, + "eval_allNLI-dev_euclidean_recall": 0.7341040462427746, + "eval_allNLI-dev_manhattan_accuracy": 0.671875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 172.47096252441406, + "eval_allNLI-dev_manhattan_ap": 0.4307335026703495, + "eval_allNLI-dev_manhattan_f1": 0.5261044176706827, + "eval_allNLI-dev_manhattan_f1_threshold": 258.1781005859375, + "eval_allNLI-dev_manhattan_precision": 0.40307692307692305, + "eval_allNLI-dev_manhattan_recall": 0.7572254335260116, + "eval_allNLI-dev_max_accuracy": 0.673828125, + "eval_allNLI-dev_max_accuracy_threshold": 734.5493774414062, + "eval_allNLI-dev_max_ap": 0.43412785256494396, + "eval_allNLI-dev_max_f1": 0.5261044176706827, + "eval_allNLI-dev_max_f1_threshold": 681.0838012695312, + "eval_allNLI-dev_max_precision": 0.40307692307692305, + "eval_allNLI-dev_max_recall": 0.7572254335260116, + "eval_loss": 1.9933717250823975, + "eval_runtime": 54.2062, + "eval_samples_per_second": 30.698, + "eval_sequential_score": 0.6311673051130458, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.2805005187117122, + "eval_sts-test_pearson_dot": 0.2779761844381874, + "eval_sts-test_pearson_euclidean": 0.3103382429908865, + "eval_sts-test_pearson_manhattan": 0.3211130291412454, + "eval_sts-test_pearson_max": 0.3211130291412454, + "eval_sts-test_spearman_cosine": 0.3370077871450756, + "eval_sts-test_spearman_dot": 0.3341743610300792, + "eval_sts-test_spearman_euclidean": 0.3370845355583619, + "eval_sts-test_spearman_manhattan": 0.3502758201274433, + "eval_sts-test_spearman_max": 0.3502758201274433, + "step": 459 + }, + { + "epoch": 0.452755905511811, + "grad_norm": 13.650534629821777, + "learning_rate": 4.522862823061631e-07, + "loss": 1.9031, + "step": 460 + }, + { + "epoch": 0.453740157480315, + "grad_norm": 15.621660232543945, + "learning_rate": 4.532803180914514e-07, + "loss": 2.1993, + "step": 461 + }, + { + "epoch": 0.4547244094488189, + "grad_norm": 12.056524276733398, + "learning_rate": 4.542743538767397e-07, + "loss": 1.4149, + "step": 462 + }, + { + "epoch": 0.45570866141732286, + "grad_norm": 12.095454216003418, + "learning_rate": 4.552683896620279e-07, + "loss": 1.2881, + "step": 463 + }, + { + "epoch": 0.4566929133858268, + "grad_norm": 42.66931915283203, + "learning_rate": 4.562624254473162e-07, + "loss": 5.0921, + "step": 464 + }, + { + "epoch": 0.45767716535433073, + "grad_norm": 38.241668701171875, + "learning_rate": 4.5725646123260444e-07, + "loss": 3.8827, + "step": 465 + }, + { + "epoch": 0.45866141732283466, + "grad_norm": 14.144487380981445, + "learning_rate": 4.5825049701789274e-07, + "loss": 1.5278, + "step": 466 + }, + { + "epoch": 0.4596456692913386, + "grad_norm": 13.078521728515625, + "learning_rate": 4.59244532803181e-07, + "loss": 1.5084, + "step": 467 + }, + { + "epoch": 0.46062992125984253, + "grad_norm": 12.216127395629883, + "learning_rate": 4.6023856858846925e-07, + "loss": 1.5728, + "step": 468 + }, + { + "epoch": 0.46161417322834647, + "grad_norm": 12.086899757385254, + "learning_rate": 4.6123260437375755e-07, + "loss": 1.3861, + "step": 469 + }, + { + "epoch": 0.4625984251968504, + "grad_norm": 12.2554349899292, + "learning_rate": 4.622266401590458e-07, + "loss": 1.2923, + "step": 470 + }, + { + "epoch": 0.46358267716535434, + "grad_norm": 14.57932186126709, + "learning_rate": 4.6322067594433406e-07, + "loss": 1.8902, + "step": 471 + }, + { + "epoch": 0.4645669291338583, + "grad_norm": 13.89941692352295, + "learning_rate": 4.642147117296223e-07, + "loss": 1.7272, + "step": 472 + }, + { + "epoch": 0.4655511811023622, + "grad_norm": 13.036828994750977, + "learning_rate": 4.652087475149106e-07, + "loss": 1.292, + "step": 473 + }, + { + "epoch": 0.46653543307086615, + "grad_norm": 10.900352478027344, + "learning_rate": 4.662027833001989e-07, + "loss": 1.2791, + "step": 474 + }, + { + "epoch": 0.4675196850393701, + "grad_norm": 11.765789031982422, + "learning_rate": 4.671968190854872e-07, + "loss": 1.8328, + "step": 475 + }, + { + "epoch": 0.468503937007874, + "grad_norm": 15.065398216247559, + "learning_rate": 4.681908548707754e-07, + "loss": 1.5654, + "step": 476 + }, + { + "epoch": 0.46948818897637795, + "grad_norm": 11.131512641906738, + "learning_rate": 4.691848906560637e-07, + "loss": 0.865, + "step": 477 + }, + { + "epoch": 0.4704724409448819, + "grad_norm": 11.469169616699219, + "learning_rate": 4.70178926441352e-07, + "loss": 1.47, + "step": 478 + }, + { + "epoch": 0.4714566929133858, + "grad_norm": 13.804464340209961, + "learning_rate": 4.7117296222664024e-07, + "loss": 1.7305, + "step": 479 + }, + { + "epoch": 0.47244094488188976, + "grad_norm": 26.671695709228516, + "learning_rate": 4.721669980119285e-07, + "loss": 2.009, + "step": 480 + }, + { + "epoch": 0.4734251968503937, + "grad_norm": 11.814390182495117, + "learning_rate": 4.731610337972168e-07, + "loss": 1.4719, + "step": 481 + }, + { + "epoch": 0.4744094488188976, + "grad_norm": 13.533790588378906, + "learning_rate": 4.7415506958250505e-07, + "loss": 1.6774, + "step": 482 + }, + { + "epoch": 0.47539370078740156, + "grad_norm": 37.1170768737793, + "learning_rate": 4.7514910536779335e-07, + "loss": 3.9534, + "step": 483 + }, + { + "epoch": 0.4763779527559055, + "grad_norm": 13.464319229125977, + "learning_rate": 4.7614314115308155e-07, + "loss": 1.7503, + "step": 484 + }, + { + "epoch": 0.47736220472440943, + "grad_norm": 12.962063789367676, + "learning_rate": 4.771371769383698e-07, + "loss": 1.4652, + "step": 485 + }, + { + "epoch": 0.47834645669291337, + "grad_norm": 11.213044166564941, + "learning_rate": 4.781312127236581e-07, + "loss": 1.4057, + "step": 486 + }, + { + "epoch": 0.4793307086614173, + "grad_norm": 12.803650856018066, + "learning_rate": 4.791252485089464e-07, + "loss": 1.7381, + "step": 487 + }, + { + "epoch": 0.48031496062992124, + "grad_norm": 42.93519592285156, + "learning_rate": 4.801192842942347e-07, + "loss": 3.1813, + "step": 488 + }, + { + "epoch": 0.4812992125984252, + "grad_norm": 14.303778648376465, + "learning_rate": 4.811133200795229e-07, + "loss": 1.6575, + "step": 489 + }, + { + "epoch": 0.4822834645669291, + "grad_norm": 11.379716873168945, + "learning_rate": 4.821073558648112e-07, + "loss": 1.2314, + "step": 490 + }, + { + "epoch": 0.48326771653543305, + "grad_norm": 63.755924224853516, + "learning_rate": 4.831013916500995e-07, + "loss": 5.9239, + "step": 491 + }, + { + "epoch": 0.484251968503937, + "grad_norm": 15.35229778289795, + "learning_rate": 4.840954274353877e-07, + "loss": 2.1689, + "step": 492 + }, + { + "epoch": 0.48523622047244097, + "grad_norm": 11.58899974822998, + "learning_rate": 4.85089463220676e-07, + "loss": 1.5536, + "step": 493 + }, + { + "epoch": 0.4862204724409449, + "grad_norm": 42.49075698852539, + "learning_rate": 4.860834990059642e-07, + "loss": 4.2167, + "step": 494 + }, + { + "epoch": 0.48720472440944884, + "grad_norm": 11.30247974395752, + "learning_rate": 4.870775347912525e-07, + "loss": 1.3489, + "step": 495 + }, + { + "epoch": 0.4881889763779528, + "grad_norm": 12.173792839050293, + "learning_rate": 4.880715705765408e-07, + "loss": 1.6137, + "step": 496 + }, + { + "epoch": 0.4891732283464567, + "grad_norm": 10.37173080444336, + "learning_rate": 4.890656063618291e-07, + "loss": 1.3183, + "step": 497 + }, + { + "epoch": 0.49015748031496065, + "grad_norm": 35.126182556152344, + "learning_rate": 4.900596421471173e-07, + "loss": 3.9189, + "step": 498 + }, + { + "epoch": 0.4911417322834646, + "grad_norm": 30.792539596557617, + "learning_rate": 4.910536779324057e-07, + "loss": 3.003, + "step": 499 + }, + { + "epoch": 0.4921259842519685, + "grad_norm": 9.988119125366211, + "learning_rate": 4.92047713717694e-07, + "loss": 1.3985, + "step": 500 + }, + { + "epoch": 0.49311023622047245, + "grad_norm": 10.691671371459961, + "learning_rate": 4.930417495029823e-07, + "loss": 1.1173, + "step": 501 + }, + { + "epoch": 0.4940944881889764, + "grad_norm": 9.386334419250488, + "learning_rate": 4.940357852882705e-07, + "loss": 0.9892, + "step": 502 + }, + { + "epoch": 0.4950787401574803, + "grad_norm": 13.273354530334473, + "learning_rate": 4.950298210735588e-07, + "loss": 1.6817, + "step": 503 + }, + { + "epoch": 0.49606299212598426, + "grad_norm": 9.470836639404297, + "learning_rate": 4.96023856858847e-07, + "loss": 1.1026, + "step": 504 + }, + { + "epoch": 0.4970472440944882, + "grad_norm": 27.343177795410156, + "learning_rate": 4.970178926441353e-07, + "loss": 3.0063, + "step": 505 + }, + { + "epoch": 0.49803149606299213, + "grad_norm": 12.136895179748535, + "learning_rate": 4.980119284294235e-07, + "loss": 1.4292, + "step": 506 + }, + { + "epoch": 0.49901574803149606, + "grad_norm": 13.389548301696777, + "learning_rate": 4.990059642147118e-07, + "loss": 1.8554, + "step": 507 + }, + { + "epoch": 0.5, + "grad_norm": 10.676069259643555, + "learning_rate": 5.000000000000001e-07, + "loss": 1.419, + "step": 508 + }, + { + "epoch": 0.5009842519685039, + "grad_norm": 11.071086883544922, + "learning_rate": 5.009940357852883e-07, + "loss": 1.2399, + "step": 509 + }, + { + "epoch": 0.5019685039370079, + "grad_norm": 10.330881118774414, + "learning_rate": 5.019880715705767e-07, + "loss": 1.2939, + "step": 510 + }, + { + "epoch": 0.5029527559055118, + "grad_norm": 10.792057037353516, + "learning_rate": 5.029821073558649e-07, + "loss": 1.4505, + "step": 511 + }, + { + "epoch": 0.5039370078740157, + "grad_norm": 11.177830696105957, + "learning_rate": 5.039761431411532e-07, + "loss": 1.0514, + "step": 512 + }, + { + "epoch": 0.5049212598425197, + "grad_norm": 10.643841743469238, + "learning_rate": 5.049701789264414e-07, + "loss": 1.2327, + "step": 513 + }, + { + "epoch": 0.5059055118110236, + "grad_norm": 11.909382820129395, + "learning_rate": 5.059642147117297e-07, + "loss": 1.3414, + "step": 514 + }, + { + "epoch": 0.5068897637795275, + "grad_norm": 10.281572341918945, + "learning_rate": 5.06958250497018e-07, + "loss": 1.3721, + "step": 515 + }, + { + "epoch": 0.5078740157480315, + "grad_norm": 53.167823791503906, + "learning_rate": 5.079522862823062e-07, + "loss": 5.4099, + "step": 516 + }, + { + "epoch": 0.5088582677165354, + "grad_norm": 10.074481010437012, + "learning_rate": 5.089463220675945e-07, + "loss": 1.0425, + "step": 517 + }, + { + "epoch": 0.5098425196850394, + "grad_norm": 47.982383728027344, + "learning_rate": 5.099403578528828e-07, + "loss": 4.6786, + "step": 518 + }, + { + "epoch": 0.5108267716535433, + "grad_norm": 10.051337242126465, + "learning_rate": 5.10934393638171e-07, + "loss": 0.9301, + "step": 519 + }, + { + "epoch": 0.5118110236220472, + "grad_norm": 11.556605339050293, + "learning_rate": 5.119284294234593e-07, + "loss": 1.3494, + "step": 520 + }, + { + "epoch": 0.5127952755905512, + "grad_norm": 11.6151762008667, + "learning_rate": 5.129224652087476e-07, + "loss": 1.158, + "step": 521 + }, + { + "epoch": 0.5137795275590551, + "grad_norm": 9.960102081298828, + "learning_rate": 5.139165009940358e-07, + "loss": 1.0346, + "step": 522 + }, + { + "epoch": 0.514763779527559, + "grad_norm": 10.822325706481934, + "learning_rate": 5.149105367793241e-07, + "loss": 1.3054, + "step": 523 + }, + { + "epoch": 0.515748031496063, + "grad_norm": 11.666457176208496, + "learning_rate": 5.159045725646124e-07, + "loss": 1.1146, + "step": 524 + }, + { + "epoch": 0.5167322834645669, + "grad_norm": 12.345129013061523, + "learning_rate": 5.168986083499007e-07, + "loss": 1.1232, + "step": 525 + }, + { + "epoch": 0.5177165354330708, + "grad_norm": 15.775688171386719, + "learning_rate": 5.178926441351889e-07, + "loss": 1.9019, + "step": 526 + }, + { + "epoch": 0.5187007874015748, + "grad_norm": 10.290173530578613, + "learning_rate": 5.188866799204771e-07, + "loss": 0.7564, + "step": 527 + }, + { + "epoch": 0.5196850393700787, + "grad_norm": 12.738898277282715, + "learning_rate": 5.198807157057656e-07, + "loss": 1.2245, + "step": 528 + }, + { + "epoch": 0.5206692913385826, + "grad_norm": 14.781623840332031, + "learning_rate": 5.208747514910538e-07, + "loss": 1.2729, + "step": 529 + }, + { + "epoch": 0.5216535433070866, + "grad_norm": 12.573616981506348, + "learning_rate": 5.21868787276342e-07, + "loss": 1.3485, + "step": 530 + }, + { + "epoch": 0.5226377952755905, + "grad_norm": 14.437003135681152, + "learning_rate": 5.228628230616304e-07, + "loss": 1.3507, + "step": 531 + }, + { + "epoch": 0.5236220472440944, + "grad_norm": 13.429966926574707, + "learning_rate": 5.238568588469186e-07, + "loss": 1.1656, + "step": 532 + }, + { + "epoch": 0.5246062992125984, + "grad_norm": 56.66192626953125, + "learning_rate": 5.248508946322069e-07, + "loss": 5.7052, + "step": 533 + }, + { + "epoch": 0.5255905511811023, + "grad_norm": 13.437881469726562, + "learning_rate": 5.258449304174951e-07, + "loss": 1.2702, + "step": 534 + }, + { + "epoch": 0.5265748031496063, + "grad_norm": 14.13820743560791, + "learning_rate": 5.268389662027834e-07, + "loss": 1.5765, + "step": 535 + }, + { + "epoch": 0.5275590551181102, + "grad_norm": 12.186126708984375, + "learning_rate": 5.278330019880717e-07, + "loss": 1.1904, + "step": 536 + }, + { + "epoch": 0.5285433070866141, + "grad_norm": 12.874736785888672, + "learning_rate": 5.288270377733599e-07, + "loss": 1.2439, + "step": 537 + }, + { + "epoch": 0.5295275590551181, + "grad_norm": 13.560784339904785, + "learning_rate": 5.298210735586483e-07, + "loss": 1.3851, + "step": 538 + }, + { + "epoch": 0.530511811023622, + "grad_norm": 48.91597366333008, + "learning_rate": 5.308151093439365e-07, + "loss": 5.4258, + "step": 539 + }, + { + "epoch": 0.531496062992126, + "grad_norm": 63.82156753540039, + "learning_rate": 5.318091451292247e-07, + "loss": 6.0733, + "step": 540 + }, + { + "epoch": 0.53248031496063, + "grad_norm": 11.866876602172852, + "learning_rate": 5.32803180914513e-07, + "loss": 1.1416, + "step": 541 + }, + { + "epoch": 0.5334645669291339, + "grad_norm": 13.9098482131958, + "learning_rate": 5.337972166998013e-07, + "loss": 1.6498, + "step": 542 + }, + { + "epoch": 0.5344488188976378, + "grad_norm": 15.363849639892578, + "learning_rate": 5.347912524850895e-07, + "loss": 1.9752, + "step": 543 + }, + { + "epoch": 0.5354330708661418, + "grad_norm": 53.16739273071289, + "learning_rate": 5.357852882703778e-07, + "loss": 5.0479, + "step": 544 + }, + { + "epoch": 0.5364173228346457, + "grad_norm": 11.984163284301758, + "learning_rate": 5.367793240556661e-07, + "loss": 1.4322, + "step": 545 + }, + { + "epoch": 0.5374015748031497, + "grad_norm": 9.779897689819336, + "learning_rate": 5.377733598409544e-07, + "loss": 1.1549, + "step": 546 + }, + { + "epoch": 0.5383858267716536, + "grad_norm": 12.604474067687988, + "learning_rate": 5.387673956262426e-07, + "loss": 1.623, + "step": 547 + }, + { + "epoch": 0.5393700787401575, + "grad_norm": 10.861937522888184, + "learning_rate": 5.397614314115308e-07, + "loss": 1.2541, + "step": 548 + }, + { + "epoch": 0.5403543307086615, + "grad_norm": 11.303938865661621, + "learning_rate": 5.407554671968192e-07, + "loss": 1.3253, + "step": 549 + }, + { + "epoch": 0.5413385826771654, + "grad_norm": 11.52147388458252, + "learning_rate": 5.417495029821074e-07, + "loss": 1.4802, + "step": 550 + }, + { + "epoch": 0.5423228346456693, + "grad_norm": 9.711042404174805, + "learning_rate": 5.427435387673957e-07, + "loss": 0.993, + "step": 551 + }, + { + "epoch": 0.5433070866141733, + "grad_norm": 10.261675834655762, + "learning_rate": 5.43737574552684e-07, + "loss": 1.2354, + "step": 552 + }, + { + "epoch": 0.5442913385826772, + "grad_norm": 9.668618202209473, + "learning_rate": 5.447316103379722e-07, + "loss": 0.7297, + "step": 553 + }, + { + "epoch": 0.5452755905511811, + "grad_norm": 59.58056640625, + "learning_rate": 5.457256461232605e-07, + "loss": 5.1613, + "step": 554 + }, + { + "epoch": 0.5462598425196851, + "grad_norm": 10.98855209350586, + "learning_rate": 5.467196819085487e-07, + "loss": 1.0977, + "step": 555 + }, + { + "epoch": 0.547244094488189, + "grad_norm": 10.048385620117188, + "learning_rate": 5.477137176938371e-07, + "loss": 1.0681, + "step": 556 + }, + { + "epoch": 0.5482283464566929, + "grad_norm": 59.76496124267578, + "learning_rate": 5.487077534791253e-07, + "loss": 5.1421, + "step": 557 + }, + { + "epoch": 0.5492125984251969, + "grad_norm": 12.414549827575684, + "learning_rate": 5.497017892644135e-07, + "loss": 1.1879, + "step": 558 + }, + { + "epoch": 0.5501968503937008, + "grad_norm": 10.891206741333008, + "learning_rate": 5.50695825049702e-07, + "loss": 1.106, + "step": 559 + }, + { + "epoch": 0.5511811023622047, + "grad_norm": 46.950321197509766, + "learning_rate": 5.516898608349902e-07, + "loss": 4.4247, + "step": 560 + }, + { + "epoch": 0.5521653543307087, + "grad_norm": 10.903446197509766, + "learning_rate": 5.526838966202784e-07, + "loss": 1.2965, + "step": 561 + }, + { + "epoch": 0.5531496062992126, + "grad_norm": 10.382317543029785, + "learning_rate": 5.536779324055667e-07, + "loss": 0.878, + "step": 562 + }, + { + "epoch": 0.5541338582677166, + "grad_norm": 11.724767684936523, + "learning_rate": 5.54671968190855e-07, + "loss": 1.3601, + "step": 563 + }, + { + "epoch": 0.5551181102362205, + "grad_norm": 12.926590919494629, + "learning_rate": 5.556660039761433e-07, + "loss": 1.4492, + "step": 564 + }, + { + "epoch": 0.5561023622047244, + "grad_norm": 10.146951675415039, + "learning_rate": 5.566600397614315e-07, + "loss": 0.9265, + "step": 565 + }, + { + "epoch": 0.5570866141732284, + "grad_norm": 11.620803833007812, + "learning_rate": 5.576540755467198e-07, + "loss": 1.2781, + "step": 566 + }, + { + "epoch": 0.5580708661417323, + "grad_norm": 9.574142456054688, + "learning_rate": 5.586481113320081e-07, + "loss": 0.8514, + "step": 567 + }, + { + "epoch": 0.5590551181102362, + "grad_norm": 13.647089004516602, + "learning_rate": 5.596421471172963e-07, + "loss": 1.3312, + "step": 568 + }, + { + "epoch": 0.5600393700787402, + "grad_norm": 28.685733795166016, + "learning_rate": 5.606361829025846e-07, + "loss": 1.9533, + "step": 569 + }, + { + "epoch": 0.5610236220472441, + "grad_norm": 11.902270317077637, + "learning_rate": 5.616302186878729e-07, + "loss": 0.9773, + "step": 570 + }, + { + "epoch": 0.562007874015748, + "grad_norm": 12.151090621948242, + "learning_rate": 5.626242544731611e-07, + "loss": 1.5187, + "step": 571 + }, + { + "epoch": 0.562992125984252, + "grad_norm": 11.929316520690918, + "learning_rate": 5.636182902584494e-07, + "loss": 1.1786, + "step": 572 + }, + { + "epoch": 0.5639763779527559, + "grad_norm": 11.160627365112305, + "learning_rate": 5.646123260437377e-07, + "loss": 1.1154, + "step": 573 + }, + { + "epoch": 0.5649606299212598, + "grad_norm": 10.790230751037598, + "learning_rate": 5.656063618290259e-07, + "loss": 0.9937, + "step": 574 + }, + { + "epoch": 0.5659448818897638, + "grad_norm": 46.701942443847656, + "learning_rate": 5.666003976143142e-07, + "loss": 3.9975, + "step": 575 + }, + { + "epoch": 0.5669291338582677, + "grad_norm": 10.748030662536621, + "learning_rate": 5.675944333996024e-07, + "loss": 0.8966, + "step": 576 + }, + { + "epoch": 0.5679133858267716, + "grad_norm": 43.31929397583008, + "learning_rate": 5.685884691848908e-07, + "loss": 3.9168, + "step": 577 + }, + { + "epoch": 0.5688976377952756, + "grad_norm": 11.500582695007324, + "learning_rate": 5.69582504970179e-07, + "loss": 1.0795, + "step": 578 + }, + { + "epoch": 0.5698818897637795, + "grad_norm": 11.253996849060059, + "learning_rate": 5.705765407554672e-07, + "loss": 0.8258, + "step": 579 + }, + { + "epoch": 0.5708661417322834, + "grad_norm": 34.77997970581055, + "learning_rate": 5.715705765407556e-07, + "loss": 2.9027, + "step": 580 + }, + { + "epoch": 0.5718503937007874, + "grad_norm": 43.59735107421875, + "learning_rate": 5.725646123260438e-07, + "loss": 3.9928, + "step": 581 + }, + { + "epoch": 0.5728346456692913, + "grad_norm": 10.751973152160645, + "learning_rate": 5.735586481113321e-07, + "loss": 1.0328, + "step": 582 + }, + { + "epoch": 0.5738188976377953, + "grad_norm": 37.12349319458008, + "learning_rate": 5.745526838966203e-07, + "loss": 3.5106, + "step": 583 + }, + { + "epoch": 0.5748031496062992, + "grad_norm": 10.426263809204102, + "learning_rate": 5.755467196819086e-07, + "loss": 1.0963, + "step": 584 + }, + { + "epoch": 0.5757874015748031, + "grad_norm": 35.70732116699219, + "learning_rate": 5.765407554671969e-07, + "loss": 3.2798, + "step": 585 + }, + { + "epoch": 0.5767716535433071, + "grad_norm": 11.900673866271973, + "learning_rate": 5.775347912524851e-07, + "loss": 1.3846, + "step": 586 + }, + { + "epoch": 0.577755905511811, + "grad_norm": 12.225541114807129, + "learning_rate": 5.785288270377734e-07, + "loss": 1.4765, + "step": 587 + }, + { + "epoch": 0.5787401574803149, + "grad_norm": 12.415202140808105, + "learning_rate": 5.795228628230618e-07, + "loss": 1.5319, + "step": 588 + }, + { + "epoch": 0.5797244094488189, + "grad_norm": 8.8843412399292, + "learning_rate": 5.8051689860835e-07, + "loss": 0.8532, + "step": 589 + }, + { + "epoch": 0.5807086614173228, + "grad_norm": 10.474663734436035, + "learning_rate": 5.815109343936383e-07, + "loss": 0.9573, + "step": 590 + }, + { + "epoch": 0.5816929133858267, + "grad_norm": 11.181775093078613, + "learning_rate": 5.825049701789266e-07, + "loss": 1.2687, + "step": 591 + }, + { + "epoch": 0.5826771653543307, + "grad_norm": 23.9904842376709, + "learning_rate": 5.834990059642148e-07, + "loss": 2.5192, + "step": 592 + }, + { + "epoch": 0.5836614173228346, + "grad_norm": 13.039846420288086, + "learning_rate": 5.844930417495031e-07, + "loss": 1.3108, + "step": 593 + }, + { + "epoch": 0.5846456692913385, + "grad_norm": 13.52120590209961, + "learning_rate": 5.854870775347914e-07, + "loss": 1.7258, + "step": 594 + }, + { + "epoch": 0.5856299212598425, + "grad_norm": 10.26899528503418, + "learning_rate": 5.864811133200797e-07, + "loss": 0.9112, + "step": 595 + }, + { + "epoch": 0.5866141732283464, + "grad_norm": 11.116101264953613, + "learning_rate": 5.874751491053679e-07, + "loss": 1.2594, + "step": 596 + }, + { + "epoch": 0.5875984251968503, + "grad_norm": 32.65583038330078, + "learning_rate": 5.884691848906561e-07, + "loss": 2.8427, + "step": 597 + }, + { + "epoch": 0.5885826771653543, + "grad_norm": 8.176895141601562, + "learning_rate": 5.894632206759445e-07, + "loss": 0.7377, + "step": 598 + }, + { + "epoch": 0.5895669291338582, + "grad_norm": 12.208337783813477, + "learning_rate": 5.904572564612327e-07, + "loss": 1.0449, + "step": 599 + }, + { + "epoch": 0.5905511811023622, + "grad_norm": 35.78986740112305, + "learning_rate": 5.914512922465209e-07, + "loss": 3.2569, + "step": 600 + }, + { + "epoch": 0.5915354330708661, + "grad_norm": 10.76382827758789, + "learning_rate": 5.924453280318093e-07, + "loss": 1.0936, + "step": 601 + }, + { + "epoch": 0.59251968503937, + "grad_norm": 11.242823600769043, + "learning_rate": 5.934393638170975e-07, + "loss": 1.0935, + "step": 602 + }, + { + "epoch": 0.593503937007874, + "grad_norm": 15.131819725036621, + "learning_rate": 5.944333996023858e-07, + "loss": 1.5585, + "step": 603 + }, + { + "epoch": 0.594488188976378, + "grad_norm": 10.615683555603027, + "learning_rate": 5.95427435387674e-07, + "loss": 0.8383, + "step": 604 + }, + { + "epoch": 0.5954724409448819, + "grad_norm": 11.556757926940918, + "learning_rate": 5.964214711729623e-07, + "loss": 0.9299, + "step": 605 + }, + { + "epoch": 0.5964566929133859, + "grad_norm": 12.370502471923828, + "learning_rate": 5.974155069582506e-07, + "loss": 1.2652, + "step": 606 + }, + { + "epoch": 0.5974409448818898, + "grad_norm": 10.400749206542969, + "learning_rate": 5.984095427435388e-07, + "loss": 0.9225, + "step": 607 + }, + { + "epoch": 0.5984251968503937, + "grad_norm": 11.763941764831543, + "learning_rate": 5.994035785288272e-07, + "loss": 1.2258, + "step": 608 + }, + { + "epoch": 0.5994094488188977, + "grad_norm": 11.520954132080078, + "learning_rate": 6.003976143141154e-07, + "loss": 1.0937, + "step": 609 + }, + { + "epoch": 0.6003937007874016, + "grad_norm": 12.111637115478516, + "learning_rate": 6.013916500994036e-07, + "loss": 1.0317, + "step": 610 + }, + { + "epoch": 0.6013779527559056, + "grad_norm": 12.023083686828613, + "learning_rate": 6.02385685884692e-07, + "loss": 0.9925, + "step": 611 + }, + { + "epoch": 0.6023622047244095, + "grad_norm": 10.958356857299805, + "learning_rate": 6.033797216699802e-07, + "loss": 1.0347, + "step": 612 + }, + { + "epoch": 0.6023622047244095, + "eval_Qnli-dev_cosine_accuracy": 0.6328125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8777655363082886, + "eval_Qnli-dev_cosine_ap": 0.636082184555122, + "eval_Qnli-dev_cosine_f1": 0.6467817896389324, + "eval_Qnli-dev_cosine_f1_threshold": 0.78821861743927, + "eval_Qnli-dev_cosine_precision": 0.513715710723192, + "eval_Qnli-dev_cosine_recall": 0.8728813559322034, + "eval_Qnli-dev_dot_accuracy": 0.6328125, + "eval_Qnli-dev_dot_accuracy_threshold": 671.5428466796875, + "eval_Qnli-dev_dot_ap": 0.6345552658179916, + "eval_Qnli-dev_dot_f1": 0.6466876971608833, + "eval_Qnli-dev_dot_f1_threshold": 603.1448974609375, + "eval_Qnli-dev_dot_precision": 0.5150753768844221, + "eval_Qnli-dev_dot_recall": 0.8686440677966102, + "eval_Qnli-dev_euclidean_accuracy": 0.6328125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 13.673242568969727, + "eval_Qnli-dev_euclidean_ap": 0.6364530484256814, + "eval_Qnli-dev_euclidean_f1": 0.6467817896389324, + "eval_Qnli-dev_euclidean_f1_threshold": 17.990753173828125, + "eval_Qnli-dev_euclidean_precision": 0.513715710723192, + "eval_Qnli-dev_euclidean_recall": 0.8728813559322034, + "eval_Qnli-dev_manhattan_accuracy": 0.646484375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 294.42578125, + "eval_Qnli-dev_manhattan_ap": 0.6537409929922839, + "eval_Qnli-dev_manhattan_f1": 0.6527570789865872, + "eval_Qnli-dev_manhattan_f1_threshold": 385.14154052734375, + "eval_Qnli-dev_manhattan_precision": 0.503448275862069, + "eval_Qnli-dev_manhattan_recall": 0.9279661016949152, + "eval_Qnli-dev_max_accuracy": 0.646484375, + "eval_Qnli-dev_max_accuracy_threshold": 671.5428466796875, + "eval_Qnli-dev_max_ap": 0.6537409929922839, + "eval_Qnli-dev_max_f1": 0.6527570789865872, + "eval_Qnli-dev_max_f1_threshold": 603.1448974609375, + "eval_Qnli-dev_max_precision": 0.5150753768844221, + "eval_Qnli-dev_max_recall": 0.9279661016949152, + "eval_allNLI-dev_cosine_accuracy": 0.6796875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9677231311798096, + "eval_allNLI-dev_cosine_ap": 0.46405508143112484, + "eval_allNLI-dev_cosine_f1": 0.5277246653919695, + "eval_allNLI-dev_cosine_f1_threshold": 0.8723927140235901, + "eval_allNLI-dev_cosine_precision": 0.3942857142857143, + "eval_allNLI-dev_cosine_recall": 0.7976878612716763, + "eval_allNLI-dev_dot_accuracy": 0.681640625, + "eval_allNLI-dev_dot_accuracy_threshold": 728.7240600585938, + "eval_allNLI-dev_dot_ap": 0.46535056618997517, + "eval_allNLI-dev_dot_f1": 0.5288640595903166, + "eval_allNLI-dev_dot_f1_threshold": 664.1909790039062, + "eval_allNLI-dev_dot_precision": 0.3901098901098901, + "eval_allNLI-dev_dot_recall": 0.8208092485549133, + "eval_allNLI-dev_euclidean_accuracy": 0.6796875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 7.031080722808838, + "eval_allNLI-dev_euclidean_ap": 0.46418150524434637, + "eval_allNLI-dev_euclidean_f1": 0.5287356321839081, + "eval_allNLI-dev_euclidean_f1_threshold": 13.958231925964355, + "eval_allNLI-dev_euclidean_precision": 0.3954154727793696, + "eval_allNLI-dev_euclidean_recall": 0.7976878612716763, + "eval_allNLI-dev_manhattan_accuracy": 0.67578125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 178.7028045654297, + "eval_allNLI-dev_manhattan_ap": 0.45843625593780585, + "eval_allNLI-dev_manhattan_f1": 0.5275142314990512, + "eval_allNLI-dev_manhattan_f1_threshold": 285.9658203125, + "eval_allNLI-dev_manhattan_precision": 0.3926553672316384, + "eval_allNLI-dev_manhattan_recall": 0.8034682080924855, + "eval_allNLI-dev_max_accuracy": 0.681640625, + "eval_allNLI-dev_max_accuracy_threshold": 728.7240600585938, + "eval_allNLI-dev_max_ap": 0.46535056618997517, + "eval_allNLI-dev_max_f1": 0.5288640595903166, + "eval_allNLI-dev_max_f1_threshold": 664.1909790039062, + "eval_allNLI-dev_max_precision": 0.3954154727793696, + "eval_allNLI-dev_max_recall": 0.8208092485549133, + "eval_loss": 1.6341288089752197, + "eval_runtime": 54.4154, + "eval_samples_per_second": 30.58, + "eval_sequential_score": 0.6537409929922839, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.37733602550373435, + "eval_sts-test_pearson_dot": 0.37306741740811944, + "eval_sts-test_pearson_euclidean": 0.39932799761186, + "eval_sts-test_pearson_manhattan": 0.4113008973824271, + "eval_sts-test_pearson_max": 0.4113008973824271, + "eval_sts-test_spearman_cosine": 0.4216275445487205, + "eval_sts-test_spearman_dot": 0.41610367259887304, + "eval_sts-test_spearman_euclidean": 0.4220386743871232, + "eval_sts-test_spearman_manhattan": 0.4336883675037937, + "eval_sts-test_spearman_max": 0.4336883675037937, + "step": 612 + }, + { + "epoch": 0.6033464566929134, + "grad_norm": 10.701803207397461, + "learning_rate": 6.043737574552684e-07, + "loss": 0.7597, + "step": 613 + }, + { + "epoch": 0.6043307086614174, + "grad_norm": 44.23878860473633, + "learning_rate": 6.053677932405567e-07, + "loss": 4.1533, + "step": 614 + }, + { + "epoch": 0.6053149606299213, + "grad_norm": 11.628232955932617, + "learning_rate": 6.06361829025845e-07, + "loss": 0.9556, + "step": 615 + }, + { + "epoch": 0.6062992125984252, + "grad_norm": 15.238367080688477, + "learning_rate": 6.073558648111333e-07, + "loss": 1.3079, + "step": 616 + }, + { + "epoch": 0.6072834645669292, + "grad_norm": 14.660409927368164, + "learning_rate": 6.083499005964215e-07, + "loss": 1.5096, + "step": 617 + }, + { + "epoch": 0.6082677165354331, + "grad_norm": 40.2839469909668, + "learning_rate": 6.093439363817097e-07, + "loss": 3.8311, + "step": 618 + }, + { + "epoch": 0.609251968503937, + "grad_norm": 13.805837631225586, + "learning_rate": 6.103379721669982e-07, + "loss": 1.379, + "step": 619 + }, + { + "epoch": 0.610236220472441, + "grad_norm": 46.00523376464844, + "learning_rate": 6.113320079522864e-07, + "loss": 3.8273, + "step": 620 + }, + { + "epoch": 0.6112204724409449, + "grad_norm": 13.986534118652344, + "learning_rate": 6.123260437375747e-07, + "loss": 1.296, + "step": 621 + }, + { + "epoch": 0.6122047244094488, + "grad_norm": 15.609278678894043, + "learning_rate": 6.13320079522863e-07, + "loss": 1.6233, + "step": 622 + }, + { + "epoch": 0.6131889763779528, + "grad_norm": 11.434459686279297, + "learning_rate": 6.143141153081512e-07, + "loss": 0.7542, + "step": 623 + }, + { + "epoch": 0.6141732283464567, + "grad_norm": 13.063506126403809, + "learning_rate": 6.153081510934395e-07, + "loss": 1.0578, + "step": 624 + }, + { + "epoch": 0.6151574803149606, + "grad_norm": 30.77376365661621, + "learning_rate": 6.163021868787277e-07, + "loss": 2.4721, + "step": 625 + }, + { + "epoch": 0.6161417322834646, + "grad_norm": 30.752870559692383, + "learning_rate": 6.172962226640161e-07, + "loss": 2.6996, + "step": 626 + }, + { + "epoch": 0.6171259842519685, + "grad_norm": 13.579484939575195, + "learning_rate": 6.182902584493043e-07, + "loss": 1.6649, + "step": 627 + }, + { + "epoch": 0.6181102362204725, + "grad_norm": 30.221582412719727, + "learning_rate": 6.192842942345925e-07, + "loss": 2.5207, + "step": 628 + }, + { + "epoch": 0.6190944881889764, + "grad_norm": 12.64453411102295, + "learning_rate": 6.202783300198809e-07, + "loss": 1.2837, + "step": 629 + }, + { + "epoch": 0.6200787401574803, + "grad_norm": 11.140278816223145, + "learning_rate": 6.212723658051691e-07, + "loss": 1.1731, + "step": 630 + }, + { + "epoch": 0.6210629921259843, + "grad_norm": 11.152597427368164, + "learning_rate": 6.222664015904573e-07, + "loss": 0.9778, + "step": 631 + }, + { + "epoch": 0.6220472440944882, + "grad_norm": 12.832036972045898, + "learning_rate": 6.232604373757457e-07, + "loss": 1.6909, + "step": 632 + }, + { + "epoch": 0.6230314960629921, + "grad_norm": 12.344291687011719, + "learning_rate": 6.242544731610339e-07, + "loss": 1.5441, + "step": 633 + }, + { + "epoch": 0.6240157480314961, + "grad_norm": 11.180489540100098, + "learning_rate": 6.252485089463222e-07, + "loss": 0.9197, + "step": 634 + }, + { + "epoch": 0.625, + "grad_norm": 23.448375701904297, + "learning_rate": 6.262425447316104e-07, + "loss": 1.781, + "step": 635 + }, + { + "epoch": 0.6259842519685039, + "grad_norm": 10.159034729003906, + "learning_rate": 6.272365805168987e-07, + "loss": 1.0285, + "step": 636 + }, + { + "epoch": 0.6269685039370079, + "grad_norm": 13.20181941986084, + "learning_rate": 6.28230616302187e-07, + "loss": 1.213, + "step": 637 + }, + { + "epoch": 0.6279527559055118, + "grad_norm": 11.607129096984863, + "learning_rate": 6.292246520874752e-07, + "loss": 1.0478, + "step": 638 + }, + { + "epoch": 0.6289370078740157, + "grad_norm": 16.6951961517334, + "learning_rate": 6.302186878727636e-07, + "loss": 1.6243, + "step": 639 + }, + { + "epoch": 0.6299212598425197, + "grad_norm": 13.98421573638916, + "learning_rate": 6.312127236580518e-07, + "loss": 1.4093, + "step": 640 + }, + { + "epoch": 0.6309055118110236, + "grad_norm": 26.42371368408203, + "learning_rate": 6.3220675944334e-07, + "loss": 1.8847, + "step": 641 + }, + { + "epoch": 0.6318897637795275, + "grad_norm": 10.854850769042969, + "learning_rate": 6.332007952286283e-07, + "loss": 1.0713, + "step": 642 + }, + { + "epoch": 0.6328740157480315, + "grad_norm": 11.060506820678711, + "learning_rate": 6.341948310139166e-07, + "loss": 0.9998, + "step": 643 + }, + { + "epoch": 0.6338582677165354, + "grad_norm": 10.658958435058594, + "learning_rate": 6.351888667992048e-07, + "loss": 0.9684, + "step": 644 + }, + { + "epoch": 0.6348425196850394, + "grad_norm": 14.677898406982422, + "learning_rate": 6.361829025844931e-07, + "loss": 2.1272, + "step": 645 + }, + { + "epoch": 0.6358267716535433, + "grad_norm": 12.596633911132812, + "learning_rate": 6.371769383697813e-07, + "loss": 1.416, + "step": 646 + }, + { + "epoch": 0.6368110236220472, + "grad_norm": 10.448723793029785, + "learning_rate": 6.381709741550698e-07, + "loss": 1.0463, + "step": 647 + }, + { + "epoch": 0.6377952755905512, + "grad_norm": 11.114981651306152, + "learning_rate": 6.39165009940358e-07, + "loss": 0.9779, + "step": 648 + }, + { + "epoch": 0.6387795275590551, + "grad_norm": 8.395352363586426, + "learning_rate": 6.401590457256462e-07, + "loss": 0.5093, + "step": 649 + }, + { + "epoch": 0.639763779527559, + "grad_norm": 12.399072647094727, + "learning_rate": 6.411530815109346e-07, + "loss": 1.1007, + "step": 650 + }, + { + "epoch": 0.640748031496063, + "grad_norm": 10.01904582977295, + "learning_rate": 6.421471172962228e-07, + "loss": 0.9436, + "step": 651 + }, + { + "epoch": 0.6417322834645669, + "grad_norm": 14.65745735168457, + "learning_rate": 6.431411530815111e-07, + "loss": 1.4446, + "step": 652 + }, + { + "epoch": 0.6427165354330708, + "grad_norm": 15.491090774536133, + "learning_rate": 6.441351888667994e-07, + "loss": 1.7327, + "step": 653 + }, + { + "epoch": 0.6437007874015748, + "grad_norm": 10.455428123474121, + "learning_rate": 6.451292246520876e-07, + "loss": 0.8326, + "step": 654 + }, + { + "epoch": 0.6446850393700787, + "grad_norm": 9.583117485046387, + "learning_rate": 6.461232604373759e-07, + "loss": 0.6514, + "step": 655 + }, + { + "epoch": 0.6456692913385826, + "grad_norm": 9.990079879760742, + "learning_rate": 6.471172962226641e-07, + "loss": 0.6155, + "step": 656 + }, + { + "epoch": 0.6466535433070866, + "grad_norm": 10.108931541442871, + "learning_rate": 6.481113320079524e-07, + "loss": 0.7906, + "step": 657 + }, + { + "epoch": 0.6476377952755905, + "grad_norm": 13.665596008300781, + "learning_rate": 6.491053677932407e-07, + "loss": 1.4234, + "step": 658 + }, + { + "epoch": 0.6486220472440944, + "grad_norm": 11.515149116516113, + "learning_rate": 6.500994035785289e-07, + "loss": 0.9298, + "step": 659 + }, + { + "epoch": 0.6496062992125984, + "grad_norm": 17.00029945373535, + "learning_rate": 6.510934393638173e-07, + "loss": 1.7384, + "step": 660 + }, + { + "epoch": 0.6505905511811023, + "grad_norm": 15.062308311462402, + "learning_rate": 6.520874751491055e-07, + "loss": 1.7171, + "step": 661 + }, + { + "epoch": 0.6515748031496063, + "grad_norm": 15.243254661560059, + "learning_rate": 6.530815109343937e-07, + "loss": 1.5247, + "step": 662 + }, + { + "epoch": 0.6525590551181102, + "grad_norm": 12.88428783416748, + "learning_rate": 6.54075546719682e-07, + "loss": 1.2509, + "step": 663 + }, + { + "epoch": 0.6535433070866141, + "grad_norm": 34.313804626464844, + "learning_rate": 6.550695825049703e-07, + "loss": 3.0989, + "step": 664 + }, + { + "epoch": 0.6545275590551181, + "grad_norm": 10.890783309936523, + "learning_rate": 6.560636182902586e-07, + "loss": 0.8984, + "step": 665 + }, + { + "epoch": 0.655511811023622, + "grad_norm": 12.740053176879883, + "learning_rate": 6.570576540755468e-07, + "loss": 1.0749, + "step": 666 + }, + { + "epoch": 0.656496062992126, + "grad_norm": 65.30494689941406, + "learning_rate": 6.580516898608351e-07, + "loss": 5.5445, + "step": 667 + }, + { + "epoch": 0.65748031496063, + "grad_norm": 12.40191650390625, + "learning_rate": 6.590457256461234e-07, + "loss": 1.15, + "step": 668 + }, + { + "epoch": 0.6584645669291339, + "grad_norm": 13.165613174438477, + "learning_rate": 6.600397614314116e-07, + "loss": 1.0755, + "step": 669 + }, + { + "epoch": 0.6594488188976378, + "grad_norm": 11.294352531433105, + "learning_rate": 6.610337972166998e-07, + "loss": 0.93, + "step": 670 + }, + { + "epoch": 0.6604330708661418, + "grad_norm": 14.429823875427246, + "learning_rate": 6.620278330019882e-07, + "loss": 1.6097, + "step": 671 + }, + { + "epoch": 0.6614173228346457, + "grad_norm": 13.069869995117188, + "learning_rate": 6.630218687872764e-07, + "loss": 1.0799, + "step": 672 + }, + { + "epoch": 0.6624015748031497, + "grad_norm": 12.076016426086426, + "learning_rate": 6.640159045725647e-07, + "loss": 1.1492, + "step": 673 + }, + { + "epoch": 0.6633858267716536, + "grad_norm": 13.563544273376465, + "learning_rate": 6.65009940357853e-07, + "loss": 1.0605, + "step": 674 + }, + { + "epoch": 0.6643700787401575, + "grad_norm": 12.923056602478027, + "learning_rate": 6.660039761431412e-07, + "loss": 0.9923, + "step": 675 + }, + { + "epoch": 0.6653543307086615, + "grad_norm": 8.722559928894043, + "learning_rate": 6.669980119284295e-07, + "loss": 0.5501, + "step": 676 + }, + { + "epoch": 0.6663385826771654, + "grad_norm": 15.11403751373291, + "learning_rate": 6.679920477137177e-07, + "loss": 1.1595, + "step": 677 + }, + { + "epoch": 0.6673228346456693, + "grad_norm": 12.444266319274902, + "learning_rate": 6.689860834990062e-07, + "loss": 0.968, + "step": 678 + }, + { + "epoch": 0.6683070866141733, + "grad_norm": 39.157779693603516, + "learning_rate": 6.699801192842944e-07, + "loss": 3.3578, + "step": 679 + }, + { + "epoch": 0.6692913385826772, + "grad_norm": 47.95648193359375, + "learning_rate": 6.709741550695826e-07, + "loss": 4.1605, + "step": 680 + }, + { + "epoch": 0.6702755905511811, + "grad_norm": 11.361122131347656, + "learning_rate": 6.71968190854871e-07, + "loss": 0.7277, + "step": 681 + }, + { + "epoch": 0.6712598425196851, + "grad_norm": 8.56065559387207, + "learning_rate": 6.729622266401592e-07, + "loss": 0.7908, + "step": 682 + }, + { + "epoch": 0.672244094488189, + "grad_norm": 9.499028205871582, + "learning_rate": 6.739562624254474e-07, + "loss": 0.898, + "step": 683 + }, + { + "epoch": 0.6732283464566929, + "grad_norm": 11.447650909423828, + "learning_rate": 6.749502982107357e-07, + "loss": 1.0208, + "step": 684 + }, + { + "epoch": 0.6742125984251969, + "grad_norm": 14.240975379943848, + "learning_rate": 6.75944333996024e-07, + "loss": 1.652, + "step": 685 + }, + { + "epoch": 0.6751968503937008, + "grad_norm": 13.617683410644531, + "learning_rate": 6.769383697813123e-07, + "loss": 1.4247, + "step": 686 + }, + { + "epoch": 0.6761811023622047, + "grad_norm": 36.25943374633789, + "learning_rate": 6.779324055666005e-07, + "loss": 3.3876, + "step": 687 + }, + { + "epoch": 0.6771653543307087, + "grad_norm": 12.248785018920898, + "learning_rate": 6.789264413518888e-07, + "loss": 1.5638, + "step": 688 + }, + { + "epoch": 0.6781496062992126, + "grad_norm": 9.890121459960938, + "learning_rate": 6.799204771371771e-07, + "loss": 1.009, + "step": 689 + }, + { + "epoch": 0.6791338582677166, + "grad_norm": 11.641477584838867, + "learning_rate": 6.809145129224653e-07, + "loss": 1.1282, + "step": 690 + }, + { + "epoch": 0.6801181102362205, + "grad_norm": 11.62278938293457, + "learning_rate": 6.819085487077536e-07, + "loss": 1.2746, + "step": 691 + }, + { + "epoch": 0.6811023622047244, + "grad_norm": 11.104626655578613, + "learning_rate": 6.829025844930419e-07, + "loss": 1.1079, + "step": 692 + }, + { + "epoch": 0.6820866141732284, + "grad_norm": 10.597916603088379, + "learning_rate": 6.838966202783301e-07, + "loss": 1.1276, + "step": 693 + }, + { + "epoch": 0.6830708661417323, + "grad_norm": 28.668432235717773, + "learning_rate": 6.848906560636184e-07, + "loss": 2.5246, + "step": 694 + }, + { + "epoch": 0.6840551181102362, + "grad_norm": 9.761340141296387, + "learning_rate": 6.858846918489067e-07, + "loss": 0.807, + "step": 695 + }, + { + "epoch": 0.6850393700787402, + "grad_norm": 11.57912826538086, + "learning_rate": 6.86878727634195e-07, + "loss": 1.0713, + "step": 696 + }, + { + "epoch": 0.6860236220472441, + "grad_norm": 9.34850788116455, + "learning_rate": 6.878727634194832e-07, + "loss": 0.8295, + "step": 697 + }, + { + "epoch": 0.687007874015748, + "grad_norm": 9.771157264709473, + "learning_rate": 6.888667992047714e-07, + "loss": 1.1652, + "step": 698 + }, + { + "epoch": 0.687992125984252, + "grad_norm": 43.168426513671875, + "learning_rate": 6.898608349900598e-07, + "loss": 2.9347, + "step": 699 + }, + { + "epoch": 0.6889763779527559, + "grad_norm": 9.494790077209473, + "learning_rate": 6.90854870775348e-07, + "loss": 0.7807, + "step": 700 + }, + { + "epoch": 0.6899606299212598, + "grad_norm": 23.544239044189453, + "learning_rate": 6.918489065606362e-07, + "loss": 1.3768, + "step": 701 + }, + { + "epoch": 0.6909448818897638, + "grad_norm": 11.414657592773438, + "learning_rate": 6.928429423459246e-07, + "loss": 1.4596, + "step": 702 + }, + { + "epoch": 0.6919291338582677, + "grad_norm": 13.431307792663574, + "learning_rate": 6.938369781312128e-07, + "loss": 1.5273, + "step": 703 + }, + { + "epoch": 0.6929133858267716, + "grad_norm": 10.998498916625977, + "learning_rate": 6.948310139165011e-07, + "loss": 0.8857, + "step": 704 + }, + { + "epoch": 0.6938976377952756, + "grad_norm": 12.236613273620605, + "learning_rate": 6.958250497017893e-07, + "loss": 1.0254, + "step": 705 + }, + { + "epoch": 0.6948818897637795, + "grad_norm": 10.905499458312988, + "learning_rate": 6.968190854870776e-07, + "loss": 0.6571, + "step": 706 + }, + { + "epoch": 0.6958661417322834, + "grad_norm": 35.93000411987305, + "learning_rate": 6.97813121272366e-07, + "loss": 2.6553, + "step": 707 + }, + { + "epoch": 0.6968503937007874, + "grad_norm": 12.845637321472168, + "learning_rate": 6.988071570576542e-07, + "loss": 1.3157, + "step": 708 + }, + { + "epoch": 0.6978346456692913, + "grad_norm": 14.18862247467041, + "learning_rate": 6.998011928429426e-07, + "loss": 1.3675, + "step": 709 + }, + { + "epoch": 0.6988188976377953, + "grad_norm": 10.507378578186035, + "learning_rate": 7.007952286282308e-07, + "loss": 0.8492, + "step": 710 + }, + { + "epoch": 0.6998031496062992, + "grad_norm": 12.715144157409668, + "learning_rate": 7.01789264413519e-07, + "loss": 1.0831, + "step": 711 + }, + { + "epoch": 0.7007874015748031, + "grad_norm": 10.003392219543457, + "learning_rate": 7.027833001988073e-07, + "loss": 0.7888, + "step": 712 + }, + { + "epoch": 0.7017716535433071, + "grad_norm": 13.558688163757324, + "learning_rate": 7.037773359840956e-07, + "loss": 1.2725, + "step": 713 + }, + { + "epoch": 0.702755905511811, + "grad_norm": 12.053690910339355, + "learning_rate": 7.047713717693838e-07, + "loss": 1.1746, + "step": 714 + }, + { + "epoch": 0.7037401574803149, + "grad_norm": 8.193392753601074, + "learning_rate": 7.057654075546721e-07, + "loss": 0.6175, + "step": 715 + }, + { + "epoch": 0.7047244094488189, + "grad_norm": 8.654151916503906, + "learning_rate": 7.067594433399604e-07, + "loss": 0.5935, + "step": 716 + }, + { + "epoch": 0.7057086614173228, + "grad_norm": 13.256050109863281, + "learning_rate": 7.077534791252487e-07, + "loss": 1.1907, + "step": 717 + }, + { + "epoch": 0.7066929133858267, + "grad_norm": 13.537592887878418, + "learning_rate": 7.087475149105369e-07, + "loss": 1.4747, + "step": 718 + }, + { + "epoch": 0.7076771653543307, + "grad_norm": 11.818737030029297, + "learning_rate": 7.097415506958251e-07, + "loss": 0.7373, + "step": 719 + }, + { + "epoch": 0.7086614173228346, + "grad_norm": 38.487003326416016, + "learning_rate": 7.107355864811135e-07, + "loss": 3.7169, + "step": 720 + }, + { + "epoch": 0.7096456692913385, + "grad_norm": 13.312034606933594, + "learning_rate": 7.117296222664017e-07, + "loss": 1.0055, + "step": 721 + }, + { + "epoch": 0.7106299212598425, + "grad_norm": 12.317737579345703, + "learning_rate": 7.1272365805169e-07, + "loss": 1.0311, + "step": 722 + }, + { + "epoch": 0.7116141732283464, + "grad_norm": 8.979562759399414, + "learning_rate": 7.137176938369783e-07, + "loss": 0.6121, + "step": 723 + }, + { + "epoch": 0.7125984251968503, + "grad_norm": 11.30450439453125, + "learning_rate": 7.147117296222665e-07, + "loss": 0.8375, + "step": 724 + }, + { + "epoch": 0.7135826771653543, + "grad_norm": 10.606871604919434, + "learning_rate": 7.157057654075548e-07, + "loss": 0.6943, + "step": 725 + }, + { + "epoch": 0.7145669291338582, + "grad_norm": 12.569619178771973, + "learning_rate": 7.16699801192843e-07, + "loss": 1.0684, + "step": 726 + }, + { + "epoch": 0.7155511811023622, + "grad_norm": 9.287858009338379, + "learning_rate": 7.176938369781313e-07, + "loss": 0.7662, + "step": 727 + }, + { + "epoch": 0.7165354330708661, + "grad_norm": 11.010154724121094, + "learning_rate": 7.186878727634196e-07, + "loss": 0.8516, + "step": 728 + }, + { + "epoch": 0.71751968503937, + "grad_norm": 12.050772666931152, + "learning_rate": 7.196819085487078e-07, + "loss": 1.2911, + "step": 729 + }, + { + "epoch": 0.718503937007874, + "grad_norm": 12.157442092895508, + "learning_rate": 7.206759443339962e-07, + "loss": 1.0036, + "step": 730 + }, + { + "epoch": 0.719488188976378, + "grad_norm": 38.21113586425781, + "learning_rate": 7.216699801192844e-07, + "loss": 3.1986, + "step": 731 + }, + { + "epoch": 0.7204724409448819, + "grad_norm": 13.931038856506348, + "learning_rate": 7.226640159045726e-07, + "loss": 0.9704, + "step": 732 + }, + { + "epoch": 0.7214566929133859, + "grad_norm": 8.16901683807373, + "learning_rate": 7.236580516898609e-07, + "loss": 0.637, + "step": 733 + }, + { + "epoch": 0.7224409448818898, + "grad_norm": 9.032388687133789, + "learning_rate": 7.246520874751492e-07, + "loss": 0.5723, + "step": 734 + }, + { + "epoch": 0.7234251968503937, + "grad_norm": 10.507246971130371, + "learning_rate": 7.256461232604375e-07, + "loss": 1.0665, + "step": 735 + }, + { + "epoch": 0.7244094488188977, + "grad_norm": 12.778374671936035, + "learning_rate": 7.266401590457257e-07, + "loss": 1.0845, + "step": 736 + }, + { + "epoch": 0.7253937007874016, + "grad_norm": 12.195361137390137, + "learning_rate": 7.27634194831014e-07, + "loss": 1.0019, + "step": 737 + }, + { + "epoch": 0.7263779527559056, + "grad_norm": 8.443058013916016, + "learning_rate": 7.286282306163024e-07, + "loss": 0.616, + "step": 738 + }, + { + "epoch": 0.7273622047244095, + "grad_norm": 37.86750030517578, + "learning_rate": 7.296222664015906e-07, + "loss": 2.9126, + "step": 739 + }, + { + "epoch": 0.7283464566929134, + "grad_norm": 9.833470344543457, + "learning_rate": 7.306163021868788e-07, + "loss": 0.7203, + "step": 740 + }, + { + "epoch": 0.7293307086614174, + "grad_norm": 10.621623992919922, + "learning_rate": 7.316103379721672e-07, + "loss": 0.617, + "step": 741 + }, + { + "epoch": 0.7303149606299213, + "grad_norm": 54.747642517089844, + "learning_rate": 7.326043737574554e-07, + "loss": 4.7813, + "step": 742 + }, + { + "epoch": 0.7312992125984252, + "grad_norm": 11.494114875793457, + "learning_rate": 7.335984095427437e-07, + "loss": 0.8951, + "step": 743 + }, + { + "epoch": 0.7322834645669292, + "grad_norm": 9.549132347106934, + "learning_rate": 7.34592445328032e-07, + "loss": 0.6171, + "step": 744 + }, + { + "epoch": 0.7332677165354331, + "grad_norm": 11.160932540893555, + "learning_rate": 7.355864811133202e-07, + "loss": 0.8032, + "step": 745 + }, + { + "epoch": 0.734251968503937, + "grad_norm": 13.318976402282715, + "learning_rate": 7.365805168986085e-07, + "loss": 1.1307, + "step": 746 + }, + { + "epoch": 0.735236220472441, + "grad_norm": 38.743255615234375, + "learning_rate": 7.375745526838967e-07, + "loss": 3.4791, + "step": 747 + }, + { + "epoch": 0.7362204724409449, + "grad_norm": 28.96615982055664, + "learning_rate": 7.385685884691851e-07, + "loss": 1.6899, + "step": 748 + }, + { + "epoch": 0.7372047244094488, + "grad_norm": 9.659252166748047, + "learning_rate": 7.395626242544733e-07, + "loss": 0.7096, + "step": 749 + }, + { + "epoch": 0.7381889763779528, + "grad_norm": 11.343924522399902, + "learning_rate": 7.405566600397615e-07, + "loss": 0.7839, + "step": 750 + }, + { + "epoch": 0.7391732283464567, + "grad_norm": 13.733336448669434, + "learning_rate": 7.415506958250499e-07, + "loss": 1.3269, + "step": 751 + }, + { + "epoch": 0.7401574803149606, + "grad_norm": 12.469435691833496, + "learning_rate": 7.425447316103381e-07, + "loss": 1.4379, + "step": 752 + }, + { + "epoch": 0.7411417322834646, + "grad_norm": 13.951380729675293, + "learning_rate": 7.435387673956263e-07, + "loss": 1.055, + "step": 753 + }, + { + "epoch": 0.7421259842519685, + "grad_norm": 12.946548461914062, + "learning_rate": 7.445328031809146e-07, + "loss": 0.9262, + "step": 754 + }, + { + "epoch": 0.7431102362204725, + "grad_norm": 9.735945701599121, + "learning_rate": 7.455268389662029e-07, + "loss": 0.7023, + "step": 755 + }, + { + "epoch": 0.7440944881889764, + "grad_norm": 32.48512649536133, + "learning_rate": 7.465208747514912e-07, + "loss": 2.6348, + "step": 756 + }, + { + "epoch": 0.7450787401574803, + "grad_norm": 12.768858909606934, + "learning_rate": 7.475149105367794e-07, + "loss": 1.3171, + "step": 757 + }, + { + "epoch": 0.7460629921259843, + "grad_norm": 26.709489822387695, + "learning_rate": 7.485089463220677e-07, + "loss": 2.5155, + "step": 758 + }, + { + "epoch": 0.7470472440944882, + "grad_norm": 12.552984237670898, + "learning_rate": 7.49502982107356e-07, + "loss": 1.0247, + "step": 759 + }, + { + "epoch": 0.7480314960629921, + "grad_norm": 13.25454044342041, + "learning_rate": 7.504970178926442e-07, + "loss": 0.9433, + "step": 760 + }, + { + "epoch": 0.7490157480314961, + "grad_norm": 11.653966903686523, + "learning_rate": 7.514910536779325e-07, + "loss": 1.1192, + "step": 761 + }, + { + "epoch": 0.75, + "grad_norm": 11.929251670837402, + "learning_rate": 7.524850894632208e-07, + "loss": 1.0187, + "step": 762 + }, + { + "epoch": 0.7509842519685039, + "grad_norm": 11.60064697265625, + "learning_rate": 7.53479125248509e-07, + "loss": 0.8483, + "step": 763 + }, + { + "epoch": 0.7519685039370079, + "grad_norm": 10.26369571685791, + "learning_rate": 7.544731610337973e-07, + "loss": 0.7776, + "step": 764 + }, + { + "epoch": 0.7529527559055118, + "grad_norm": 32.42780303955078, + "learning_rate": 7.554671968190856e-07, + "loss": 2.2929, + "step": 765 + }, + { + "epoch": 0.7529527559055118, + "eval_Qnli-dev_cosine_accuracy": 0.654296875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8493565320968628, + "eval_Qnli-dev_cosine_ap": 0.6651485883101813, + "eval_Qnli-dev_cosine_f1": 0.6678023850085179, + "eval_Qnli-dev_cosine_f1_threshold": 0.7991494536399841, + "eval_Qnli-dev_cosine_precision": 0.5584045584045584, + "eval_Qnli-dev_cosine_recall": 0.8305084745762712, + "eval_Qnli-dev_dot_accuracy": 0.654296875, + "eval_Qnli-dev_dot_accuracy_threshold": 649.9716186523438, + "eval_Qnli-dev_dot_ap": 0.6631292557807136, + "eval_Qnli-dev_dot_f1": 0.6700507614213197, + "eval_Qnli-dev_dot_f1_threshold": 609.7298583984375, + "eval_Qnli-dev_dot_precision": 0.5577464788732395, + "eval_Qnli-dev_dot_recall": 0.8389830508474576, + "eval_Qnli-dev_euclidean_accuracy": 0.654296875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 14.864194869995117, + "eval_Qnli-dev_euclidean_ap": 0.6652217784881351, + "eval_Qnli-dev_euclidean_f1": 0.6678023850085179, + "eval_Qnli-dev_euclidean_f1_threshold": 17.510093688964844, + "eval_Qnli-dev_euclidean_precision": 0.5584045584045584, + "eval_Qnli-dev_euclidean_recall": 0.8305084745762712, + "eval_Qnli-dev_manhattan_accuracy": 0.662109375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 301.39422607421875, + "eval_Qnli-dev_manhattan_ap": 0.6763570733667617, + "eval_Qnli-dev_manhattan_f1": 0.6762820512820512, + "eval_Qnli-dev_manhattan_f1_threshold": 376.59600830078125, + "eval_Qnli-dev_manhattan_precision": 0.5438144329896907, + "eval_Qnli-dev_manhattan_recall": 0.8940677966101694, + "eval_Qnli-dev_max_accuracy": 0.662109375, + "eval_Qnli-dev_max_accuracy_threshold": 649.9716186523438, + "eval_Qnli-dev_max_ap": 0.6763570733667617, + "eval_Qnli-dev_max_f1": 0.6762820512820512, + "eval_Qnli-dev_max_f1_threshold": 609.7298583984375, + "eval_Qnli-dev_max_precision": 0.5584045584045584, + "eval_Qnli-dev_max_recall": 0.8940677966101694, + "eval_allNLI-dev_cosine_accuracy": 0.6875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9424331188201904, + "eval_allNLI-dev_cosine_ap": 0.4692955852848747, + "eval_allNLI-dev_cosine_f1": 0.5341365461847389, + "eval_allNLI-dev_cosine_f1_threshold": 0.8611772060394287, + "eval_allNLI-dev_cosine_precision": 0.40923076923076923, + "eval_allNLI-dev_cosine_recall": 0.7687861271676301, + "eval_allNLI-dev_dot_accuracy": 0.6875, + "eval_allNLI-dev_dot_accuracy_threshold": 722.18798828125, + "eval_allNLI-dev_dot_ap": 0.4694961600900589, + "eval_allNLI-dev_dot_f1": 0.5357873210633947, + "eval_allNLI-dev_dot_f1_threshold": 661.792724609375, + "eval_allNLI-dev_dot_precision": 0.41455696202531644, + "eval_allNLI-dev_dot_recall": 0.7572254335260116, + "eval_allNLI-dev_euclidean_accuracy": 0.6875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.392923355102539, + "eval_allNLI-dev_euclidean_ap": 0.4693675979912956, + "eval_allNLI-dev_euclidean_f1": 0.5341365461847389, + "eval_allNLI-dev_euclidean_f1_threshold": 14.569525718688965, + "eval_allNLI-dev_euclidean_precision": 0.40923076923076923, + "eval_allNLI-dev_euclidean_recall": 0.7687861271676301, + "eval_allNLI-dev_manhattan_accuracy": 0.677734375, + "eval_allNLI-dev_manhattan_accuracy_threshold": 195.32534790039062, + "eval_allNLI-dev_manhattan_ap": 0.4693162755377951, + "eval_allNLI-dev_manhattan_f1": 0.5338645418326693, + "eval_allNLI-dev_manhattan_f1_threshold": 301.98321533203125, + "eval_allNLI-dev_manhattan_precision": 0.4072948328267477, + "eval_allNLI-dev_manhattan_recall": 0.7745664739884393, + "eval_allNLI-dev_max_accuracy": 0.6875, + "eval_allNLI-dev_max_accuracy_threshold": 722.18798828125, + "eval_allNLI-dev_max_ap": 0.4694961600900589, + "eval_allNLI-dev_max_f1": 0.5357873210633947, + "eval_allNLI-dev_max_f1_threshold": 661.792724609375, + "eval_allNLI-dev_max_precision": 0.41455696202531644, + "eval_allNLI-dev_max_recall": 0.7745664739884393, + "eval_loss": 1.460240364074707, + "eval_runtime": 54.1723, + "eval_samples_per_second": 30.717, + "eval_sequential_score": 0.6763570733667617, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.39163897925771674, + "eval_sts-test_pearson_dot": 0.3859127227394329, + "eval_sts-test_pearson_euclidean": 0.416510609899249, + "eval_sts-test_pearson_manhattan": 0.42444677513396134, + "eval_sts-test_pearson_max": 0.42444677513396134, + "eval_sts-test_spearman_cosine": 0.43941218808953797, + "eval_sts-test_spearman_dot": 0.4330934889861178, + "eval_sts-test_spearman_euclidean": 0.4398173660101757, + "eval_sts-test_spearman_manhattan": 0.4483853082623888, + "eval_sts-test_spearman_max": 0.4483853082623888, + "step": 765 + }, + { + "epoch": 0.7539370078740157, + "grad_norm": 14.009566307067871, + "learning_rate": 7.56461232604374e-07, + "loss": 1.1385, + "step": 766 + }, + { + "epoch": 0.7549212598425197, + "grad_norm": 9.024477005004883, + "learning_rate": 7.574552683896621e-07, + "loss": 0.5783, + "step": 767 + }, + { + "epoch": 0.7559055118110236, + "grad_norm": 14.691167831420898, + "learning_rate": 7.584493041749503e-07, + "loss": 1.2594, + "step": 768 + }, + { + "epoch": 0.7568897637795275, + "grad_norm": 30.12328338623047, + "learning_rate": 7.594433399602388e-07, + "loss": 2.1814, + "step": 769 + }, + { + "epoch": 0.7578740157480315, + "grad_norm": 11.479731559753418, + "learning_rate": 7.60437375745527e-07, + "loss": 0.9557, + "step": 770 + }, + { + "epoch": 0.7588582677165354, + "grad_norm": 9.11027717590332, + "learning_rate": 7.614314115308152e-07, + "loss": 0.6649, + "step": 771 + }, + { + "epoch": 0.7598425196850394, + "grad_norm": 11.371647834777832, + "learning_rate": 7.624254473161036e-07, + "loss": 0.8908, + "step": 772 + }, + { + "epoch": 0.7608267716535433, + "grad_norm": 10.068663597106934, + "learning_rate": 7.634194831013918e-07, + "loss": 0.8234, + "step": 773 + }, + { + "epoch": 0.7618110236220472, + "grad_norm": 12.031545639038086, + "learning_rate": 7.644135188866801e-07, + "loss": 1.0889, + "step": 774 + }, + { + "epoch": 0.7627952755905512, + "grad_norm": 11.676977157592773, + "learning_rate": 7.654075546719683e-07, + "loss": 1.0294, + "step": 775 + }, + { + "epoch": 0.7637795275590551, + "grad_norm": 12.664962768554688, + "learning_rate": 7.664015904572566e-07, + "loss": 1.207, + "step": 776 + }, + { + "epoch": 0.764763779527559, + "grad_norm": 11.427294731140137, + "learning_rate": 7.673956262425449e-07, + "loss": 0.9177, + "step": 777 + }, + { + "epoch": 0.765748031496063, + "grad_norm": 11.521842002868652, + "learning_rate": 7.683896620278331e-07, + "loss": 1.118, + "step": 778 + }, + { + "epoch": 0.7667322834645669, + "grad_norm": 12.860966682434082, + "learning_rate": 7.693836978131215e-07, + "loss": 1.2775, + "step": 779 + }, + { + "epoch": 0.7677165354330708, + "grad_norm": 34.428462982177734, + "learning_rate": 7.703777335984097e-07, + "loss": 3.6265, + "step": 780 + }, + { + "epoch": 0.7687007874015748, + "grad_norm": 11.847333908081055, + "learning_rate": 7.713717693836979e-07, + "loss": 1.0477, + "step": 781 + }, + { + "epoch": 0.7696850393700787, + "grad_norm": 11.537668228149414, + "learning_rate": 7.723658051689862e-07, + "loss": 1.0475, + "step": 782 + }, + { + "epoch": 0.7706692913385826, + "grad_norm": 10.223224639892578, + "learning_rate": 7.733598409542745e-07, + "loss": 0.8965, + "step": 783 + }, + { + "epoch": 0.7716535433070866, + "grad_norm": 9.149170875549316, + "learning_rate": 7.743538767395627e-07, + "loss": 0.6398, + "step": 784 + }, + { + "epoch": 0.7726377952755905, + "grad_norm": 9.898460388183594, + "learning_rate": 7.75347912524851e-07, + "loss": 0.8387, + "step": 785 + }, + { + "epoch": 0.7736220472440944, + "grad_norm": 14.476304054260254, + "learning_rate": 7.763419483101393e-07, + "loss": 1.2206, + "step": 786 + }, + { + "epoch": 0.7746062992125984, + "grad_norm": 10.1232328414917, + "learning_rate": 7.773359840954276e-07, + "loss": 0.7254, + "step": 787 + }, + { + "epoch": 0.7755905511811023, + "grad_norm": 29.228551864624023, + "learning_rate": 7.783300198807158e-07, + "loss": 2.674, + "step": 788 + }, + { + "epoch": 0.7765748031496063, + "grad_norm": 11.327478408813477, + "learning_rate": 7.79324055666004e-07, + "loss": 0.9569, + "step": 789 + }, + { + "epoch": 0.7775590551181102, + "grad_norm": 12.248971939086914, + "learning_rate": 7.803180914512924e-07, + "loss": 1.1421, + "step": 790 + }, + { + "epoch": 0.7785433070866141, + "grad_norm": 11.644203186035156, + "learning_rate": 7.813121272365806e-07, + "loss": 1.0097, + "step": 791 + }, + { + "epoch": 0.7795275590551181, + "grad_norm": 9.664244651794434, + "learning_rate": 7.823061630218689e-07, + "loss": 0.7003, + "step": 792 + }, + { + "epoch": 0.780511811023622, + "grad_norm": 10.34313678741455, + "learning_rate": 7.833001988071572e-07, + "loss": 0.9617, + "step": 793 + }, + { + "epoch": 0.781496062992126, + "grad_norm": 10.9442720413208, + "learning_rate": 7.842942345924454e-07, + "loss": 0.7657, + "step": 794 + }, + { + "epoch": 0.78248031496063, + "grad_norm": 13.110860824584961, + "learning_rate": 7.852882703777337e-07, + "loss": 1.0451, + "step": 795 + }, + { + "epoch": 0.7834645669291339, + "grad_norm": 32.66413879394531, + "learning_rate": 7.862823061630219e-07, + "loss": 2.6532, + "step": 796 + }, + { + "epoch": 0.7844488188976378, + "grad_norm": 31.113216400146484, + "learning_rate": 7.872763419483102e-07, + "loss": 2.2881, + "step": 797 + }, + { + "epoch": 0.7854330708661418, + "grad_norm": 8.918703079223633, + "learning_rate": 7.882703777335986e-07, + "loss": 0.6881, + "step": 798 + }, + { + "epoch": 0.7864173228346457, + "grad_norm": 13.535049438476562, + "learning_rate": 7.892644135188868e-07, + "loss": 0.8976, + "step": 799 + }, + { + "epoch": 0.7874015748031497, + "grad_norm": 26.764535903930664, + "learning_rate": 7.902584493041752e-07, + "loss": 2.1921, + "step": 800 + }, + { + "epoch": 0.7883858267716536, + "grad_norm": 9.47220516204834, + "learning_rate": 7.912524850894634e-07, + "loss": 0.5112, + "step": 801 + }, + { + "epoch": 0.7893700787401575, + "grad_norm": 10.911540985107422, + "learning_rate": 7.922465208747516e-07, + "loss": 0.775, + "step": 802 + }, + { + "epoch": 0.7903543307086615, + "grad_norm": 28.311187744140625, + "learning_rate": 7.932405566600399e-07, + "loss": 1.9774, + "step": 803 + }, + { + "epoch": 0.7913385826771654, + "grad_norm": 28.16878890991211, + "learning_rate": 7.942345924453282e-07, + "loss": 1.8192, + "step": 804 + }, + { + "epoch": 0.7923228346456693, + "grad_norm": 9.268046379089355, + "learning_rate": 7.952286282306165e-07, + "loss": 0.6617, + "step": 805 + }, + { + "epoch": 0.7933070866141733, + "grad_norm": 12.68924331665039, + "learning_rate": 7.962226640159047e-07, + "loss": 1.1582, + "step": 806 + }, + { + "epoch": 0.7942913385826772, + "grad_norm": 21.48848533630371, + "learning_rate": 7.97216699801193e-07, + "loss": 1.7469, + "step": 807 + }, + { + "epoch": 0.7952755905511811, + "grad_norm": 10.589978218078613, + "learning_rate": 7.982107355864813e-07, + "loss": 1.0597, + "step": 808 + }, + { + "epoch": 0.7962598425196851, + "grad_norm": 9.825285911560059, + "learning_rate": 7.992047713717695e-07, + "loss": 0.4811, + "step": 809 + }, + { + "epoch": 0.797244094488189, + "grad_norm": 14.238350868225098, + "learning_rate": 8.001988071570577e-07, + "loss": 1.1675, + "step": 810 + }, + { + "epoch": 0.7982283464566929, + "grad_norm": 11.2032470703125, + "learning_rate": 8.011928429423461e-07, + "loss": 0.9254, + "step": 811 + }, + { + "epoch": 0.7992125984251969, + "grad_norm": 12.099225997924805, + "learning_rate": 8.021868787276343e-07, + "loss": 0.9448, + "step": 812 + }, + { + "epoch": 0.8001968503937008, + "grad_norm": 14.134735107421875, + "learning_rate": 8.031809145129226e-07, + "loss": 1.2215, + "step": 813 + }, + { + "epoch": 0.8011811023622047, + "grad_norm": 13.462716102600098, + "learning_rate": 8.041749502982109e-07, + "loss": 1.1021, + "step": 814 + }, + { + "epoch": 0.8021653543307087, + "grad_norm": 15.155725479125977, + "learning_rate": 8.051689860834991e-07, + "loss": 1.6263, + "step": 815 + }, + { + "epoch": 0.8031496062992126, + "grad_norm": 9.317359924316406, + "learning_rate": 8.061630218687874e-07, + "loss": 0.6675, + "step": 816 + }, + { + "epoch": 0.8041338582677166, + "grad_norm": 11.237923622131348, + "learning_rate": 8.071570576540756e-07, + "loss": 0.8904, + "step": 817 + }, + { + "epoch": 0.8051181102362205, + "grad_norm": 11.468343734741211, + "learning_rate": 8.08151093439364e-07, + "loss": 0.8919, + "step": 818 + }, + { + "epoch": 0.8061023622047244, + "grad_norm": 10.934569358825684, + "learning_rate": 8.091451292246522e-07, + "loss": 0.6128, + "step": 819 + }, + { + "epoch": 0.8070866141732284, + "grad_norm": 10.651111602783203, + "learning_rate": 8.101391650099404e-07, + "loss": 0.7364, + "step": 820 + }, + { + "epoch": 0.8080708661417323, + "grad_norm": 28.13587760925293, + "learning_rate": 8.111332007952288e-07, + "loss": 1.8094, + "step": 821 + }, + { + "epoch": 0.8090551181102362, + "grad_norm": 11.294393539428711, + "learning_rate": 8.12127236580517e-07, + "loss": 0.7423, + "step": 822 + }, + { + "epoch": 0.8100393700787402, + "grad_norm": 8.074543952941895, + "learning_rate": 8.131212723658052e-07, + "loss": 0.4273, + "step": 823 + }, + { + "epoch": 0.8110236220472441, + "grad_norm": 10.790502548217773, + "learning_rate": 8.141153081510935e-07, + "loss": 0.7915, + "step": 824 + }, + { + "epoch": 0.812007874015748, + "grad_norm": 11.241899490356445, + "learning_rate": 8.151093439363818e-07, + "loss": 0.6667, + "step": 825 + }, + { + "epoch": 0.812992125984252, + "grad_norm": 14.739508628845215, + "learning_rate": 8.161033797216701e-07, + "loss": 1.0842, + "step": 826 + }, + { + "epoch": 0.8139763779527559, + "grad_norm": 10.02929973602295, + "learning_rate": 8.170974155069583e-07, + "loss": 0.6472, + "step": 827 + }, + { + "epoch": 0.8149606299212598, + "grad_norm": 11.892731666564941, + "learning_rate": 8.180914512922467e-07, + "loss": 0.6185, + "step": 828 + }, + { + "epoch": 0.8159448818897638, + "grad_norm": 15.663006782531738, + "learning_rate": 8.19085487077535e-07, + "loss": 1.1912, + "step": 829 + }, + { + "epoch": 0.8169291338582677, + "grad_norm": 15.349569320678711, + "learning_rate": 8.200795228628232e-07, + "loss": 1.3546, + "step": 830 + }, + { + "epoch": 0.8179133858267716, + "grad_norm": 11.670661926269531, + "learning_rate": 8.210735586481115e-07, + "loss": 0.8114, + "step": 831 + }, + { + "epoch": 0.8188976377952756, + "grad_norm": 27.554731369018555, + "learning_rate": 8.220675944333998e-07, + "loss": 1.5822, + "step": 832 + }, + { + "epoch": 0.8198818897637795, + "grad_norm": 12.875972747802734, + "learning_rate": 8.23061630218688e-07, + "loss": 0.7137, + "step": 833 + }, + { + "epoch": 0.8208661417322834, + "grad_norm": 6.835382461547852, + "learning_rate": 8.240556660039763e-07, + "loss": 0.2886, + "step": 834 + }, + { + "epoch": 0.8218503937007874, + "grad_norm": 15.828883171081543, + "learning_rate": 8.250497017892646e-07, + "loss": 1.2391, + "step": 835 + }, + { + "epoch": 0.8228346456692913, + "grad_norm": 15.309415817260742, + "learning_rate": 8.260437375745529e-07, + "loss": 1.1536, + "step": 836 + }, + { + "epoch": 0.8238188976377953, + "grad_norm": 16.423490524291992, + "learning_rate": 8.270377733598411e-07, + "loss": 1.1422, + "step": 837 + }, + { + "epoch": 0.8248031496062992, + "grad_norm": 35.645301818847656, + "learning_rate": 8.280318091451293e-07, + "loss": 2.7215, + "step": 838 + }, + { + "epoch": 0.8257874015748031, + "grad_norm": 30.38851547241211, + "learning_rate": 8.290258449304177e-07, + "loss": 1.9721, + "step": 839 + }, + { + "epoch": 0.8267716535433071, + "grad_norm": 9.624545097351074, + "learning_rate": 8.300198807157059e-07, + "loss": 0.5596, + "step": 840 + }, + { + "epoch": 0.827755905511811, + "grad_norm": 24.00203514099121, + "learning_rate": 8.310139165009941e-07, + "loss": 1.1538, + "step": 841 + }, + { + "epoch": 0.8287401574803149, + "grad_norm": 15.759881973266602, + "learning_rate": 8.320079522862825e-07, + "loss": 1.1463, + "step": 842 + }, + { + "epoch": 0.8297244094488189, + "grad_norm": 15.030658721923828, + "learning_rate": 8.330019880715707e-07, + "loss": 1.2118, + "step": 843 + }, + { + "epoch": 0.8307086614173228, + "grad_norm": 17.78954315185547, + "learning_rate": 8.33996023856859e-07, + "loss": 1.0761, + "step": 844 + }, + { + "epoch": 0.8316929133858267, + "grad_norm": 28.6446533203125, + "learning_rate": 8.349900596421472e-07, + "loss": 2.2337, + "step": 845 + }, + { + "epoch": 0.8326771653543307, + "grad_norm": 15.061113357543945, + "learning_rate": 8.359840954274355e-07, + "loss": 1.4387, + "step": 846 + }, + { + "epoch": 0.8336614173228346, + "grad_norm": 16.369281768798828, + "learning_rate": 8.369781312127238e-07, + "loss": 1.3689, + "step": 847 + }, + { + "epoch": 0.8346456692913385, + "grad_norm": 11.550053596496582, + "learning_rate": 8.37972166998012e-07, + "loss": 0.6953, + "step": 848 + }, + { + "epoch": 0.8356299212598425, + "grad_norm": 11.115499496459961, + "learning_rate": 8.389662027833004e-07, + "loss": 0.7809, + "step": 849 + }, + { + "epoch": 0.8366141732283464, + "grad_norm": 11.123430252075195, + "learning_rate": 8.399602385685886e-07, + "loss": 0.8823, + "step": 850 + }, + { + "epoch": 0.8375984251968503, + "grad_norm": 14.044934272766113, + "learning_rate": 8.409542743538768e-07, + "loss": 1.2892, + "step": 851 + }, + { + "epoch": 0.8385826771653543, + "grad_norm": 11.716177940368652, + "learning_rate": 8.419483101391651e-07, + "loss": 1.0262, + "step": 852 + }, + { + "epoch": 0.8395669291338582, + "grad_norm": 14.854663848876953, + "learning_rate": 8.429423459244534e-07, + "loss": 1.053, + "step": 853 + }, + { + "epoch": 0.8405511811023622, + "grad_norm": 12.035706520080566, + "learning_rate": 8.439363817097416e-07, + "loss": 0.8695, + "step": 854 + }, + { + "epoch": 0.8415354330708661, + "grad_norm": 11.932565689086914, + "learning_rate": 8.449304174950299e-07, + "loss": 0.9463, + "step": 855 + }, + { + "epoch": 0.84251968503937, + "grad_norm": 30.260635375976562, + "learning_rate": 8.459244532803182e-07, + "loss": 2.3975, + "step": 856 + }, + { + "epoch": 0.843503937007874, + "grad_norm": 13.055898666381836, + "learning_rate": 8.469184890656066e-07, + "loss": 1.032, + "step": 857 + }, + { + "epoch": 0.844488188976378, + "grad_norm": 29.741281509399414, + "learning_rate": 8.479125248508948e-07, + "loss": 2.1536, + "step": 858 + }, + { + "epoch": 0.8454724409448819, + "grad_norm": 24.180944442749023, + "learning_rate": 8.48906560636183e-07, + "loss": 1.5867, + "step": 859 + }, + { + "epoch": 0.8464566929133859, + "grad_norm": 10.7355318069458, + "learning_rate": 8.499005964214714e-07, + "loss": 0.8153, + "step": 860 + }, + { + "epoch": 0.8474409448818898, + "grad_norm": 10.552571296691895, + "learning_rate": 8.508946322067596e-07, + "loss": 0.7053, + "step": 861 + }, + { + "epoch": 0.8484251968503937, + "grad_norm": 11.440624237060547, + "learning_rate": 8.518886679920479e-07, + "loss": 1.0604, + "step": 862 + }, + { + "epoch": 0.8494094488188977, + "grad_norm": 24.281822204589844, + "learning_rate": 8.528827037773362e-07, + "loss": 2.0091, + "step": 863 + }, + { + "epoch": 0.8503937007874016, + "grad_norm": 13.694790840148926, + "learning_rate": 8.538767395626244e-07, + "loss": 0.9021, + "step": 864 + }, + { + "epoch": 0.8513779527559056, + "grad_norm": 23.159679412841797, + "learning_rate": 8.548707753479127e-07, + "loss": 1.6572, + "step": 865 + }, + { + "epoch": 0.8523622047244095, + "grad_norm": 9.769048690795898, + "learning_rate": 8.558648111332009e-07, + "loss": 0.7634, + "step": 866 + }, + { + "epoch": 0.8533464566929134, + "grad_norm": 15.126567840576172, + "learning_rate": 8.568588469184892e-07, + "loss": 1.3602, + "step": 867 + }, + { + "epoch": 0.8543307086614174, + "grad_norm": 18.267560958862305, + "learning_rate": 8.578528827037775e-07, + "loss": 1.8969, + "step": 868 + }, + { + "epoch": 0.8553149606299213, + "grad_norm": 19.578649520874023, + "learning_rate": 8.588469184890657e-07, + "loss": 1.8319, + "step": 869 + }, + { + "epoch": 0.8562992125984252, + "grad_norm": 19.867412567138672, + "learning_rate": 8.598409542743541e-07, + "loss": 1.8445, + "step": 870 + }, + { + "epoch": 0.8572834645669292, + "grad_norm": 15.06701374053955, + "learning_rate": 8.608349900596423e-07, + "loss": 1.3896, + "step": 871 + }, + { + "epoch": 0.8582677165354331, + "grad_norm": 9.85556697845459, + "learning_rate": 8.618290258449305e-07, + "loss": 0.5278, + "step": 872 + }, + { + "epoch": 0.859251968503937, + "grad_norm": 10.843356132507324, + "learning_rate": 8.628230616302189e-07, + "loss": 0.9049, + "step": 873 + }, + { + "epoch": 0.860236220472441, + "grad_norm": 11.80095386505127, + "learning_rate": 8.638170974155071e-07, + "loss": 0.8758, + "step": 874 + }, + { + "epoch": 0.8612204724409449, + "grad_norm": 13.865303993225098, + "learning_rate": 8.648111332007954e-07, + "loss": 1.2579, + "step": 875 + }, + { + "epoch": 0.8622047244094488, + "grad_norm": 10.507223129272461, + "learning_rate": 8.658051689860836e-07, + "loss": 0.7051, + "step": 876 + }, + { + "epoch": 0.8631889763779528, + "grad_norm": 12.494022369384766, + "learning_rate": 8.667992047713719e-07, + "loss": 0.9826, + "step": 877 + }, + { + "epoch": 0.8641732283464567, + "grad_norm": 13.588288307189941, + "learning_rate": 8.677932405566602e-07, + "loss": 0.7903, + "step": 878 + }, + { + "epoch": 0.8651574803149606, + "grad_norm": 13.005169868469238, + "learning_rate": 8.687872763419484e-07, + "loss": 1.0113, + "step": 879 + }, + { + "epoch": 0.8661417322834646, + "grad_norm": 14.407569885253906, + "learning_rate": 8.697813121272366e-07, + "loss": 1.0594, + "step": 880 + }, + { + "epoch": 0.8671259842519685, + "grad_norm": 12.332335472106934, + "learning_rate": 8.70775347912525e-07, + "loss": 0.9289, + "step": 881 + }, + { + "epoch": 0.8681102362204725, + "grad_norm": 30.00078010559082, + "learning_rate": 8.717693836978132e-07, + "loss": 2.2593, + "step": 882 + }, + { + "epoch": 0.8690944881889764, + "grad_norm": 12.043009757995605, + "learning_rate": 8.727634194831015e-07, + "loss": 0.8689, + "step": 883 + }, + { + "epoch": 0.8700787401574803, + "grad_norm": 11.589649200439453, + "learning_rate": 8.737574552683898e-07, + "loss": 0.7923, + "step": 884 + }, + { + "epoch": 0.8710629921259843, + "grad_norm": 15.759047508239746, + "learning_rate": 8.74751491053678e-07, + "loss": 0.9832, + "step": 885 + }, + { + "epoch": 0.8720472440944882, + "grad_norm": 40.35478591918945, + "learning_rate": 8.757455268389663e-07, + "loss": 2.681, + "step": 886 + }, + { + "epoch": 0.8730314960629921, + "grad_norm": 14.283058166503906, + "learning_rate": 8.767395626242545e-07, + "loss": 1.0878, + "step": 887 + }, + { + "epoch": 0.8740157480314961, + "grad_norm": 13.471712112426758, + "learning_rate": 8.77733598409543e-07, + "loss": 0.6764, + "step": 888 + }, + { + "epoch": 0.875, + "grad_norm": 24.483654022216797, + "learning_rate": 8.787276341948312e-07, + "loss": 1.7032, + "step": 889 + }, + { + "epoch": 0.8759842519685039, + "grad_norm": 16.756513595581055, + "learning_rate": 8.797216699801194e-07, + "loss": 1.4734, + "step": 890 + }, + { + "epoch": 0.8769685039370079, + "grad_norm": 12.61288833618164, + "learning_rate": 8.807157057654078e-07, + "loss": 0.8396, + "step": 891 + }, + { + "epoch": 0.8779527559055118, + "grad_norm": 13.15617561340332, + "learning_rate": 8.81709741550696e-07, + "loss": 1.0006, + "step": 892 + }, + { + "epoch": 0.8789370078740157, + "grad_norm": 14.844758987426758, + "learning_rate": 8.827037773359842e-07, + "loss": 1.2926, + "step": 893 + }, + { + "epoch": 0.8799212598425197, + "grad_norm": 11.292862892150879, + "learning_rate": 8.836978131212726e-07, + "loss": 0.7681, + "step": 894 + }, + { + "epoch": 0.8809055118110236, + "grad_norm": 11.68862247467041, + "learning_rate": 8.846918489065608e-07, + "loss": 0.8424, + "step": 895 + }, + { + "epoch": 0.8818897637795275, + "grad_norm": 8.260144233703613, + "learning_rate": 8.856858846918491e-07, + "loss": 0.4961, + "step": 896 + }, + { + "epoch": 0.8828740157480315, + "grad_norm": 21.9805965423584, + "learning_rate": 8.866799204771373e-07, + "loss": 1.8518, + "step": 897 + }, + { + "epoch": 0.8838582677165354, + "grad_norm": 12.657113075256348, + "learning_rate": 8.876739562624256e-07, + "loss": 0.7808, + "step": 898 + }, + { + "epoch": 0.8848425196850394, + "grad_norm": 10.799315452575684, + "learning_rate": 8.886679920477139e-07, + "loss": 0.7176, + "step": 899 + }, + { + "epoch": 0.8858267716535433, + "grad_norm": 12.964784622192383, + "learning_rate": 8.896620278330021e-07, + "loss": 0.8946, + "step": 900 + }, + { + "epoch": 0.8868110236220472, + "grad_norm": 23.136194229125977, + "learning_rate": 8.906560636182905e-07, + "loss": 1.5311, + "step": 901 + }, + { + "epoch": 0.8877952755905512, + "grad_norm": 12.264001846313477, + "learning_rate": 8.916500994035787e-07, + "loss": 1.0758, + "step": 902 + }, + { + "epoch": 0.8887795275590551, + "grad_norm": 11.262816429138184, + "learning_rate": 8.926441351888669e-07, + "loss": 0.843, + "step": 903 + }, + { + "epoch": 0.889763779527559, + "grad_norm": 14.829652786254883, + "learning_rate": 8.936381709741552e-07, + "loss": 1.3121, + "step": 904 + }, + { + "epoch": 0.890748031496063, + "grad_norm": 12.111909866333008, + "learning_rate": 8.946322067594435e-07, + "loss": 0.9709, + "step": 905 + }, + { + "epoch": 0.8917322834645669, + "grad_norm": 13.451350212097168, + "learning_rate": 8.956262425447318e-07, + "loss": 0.9791, + "step": 906 + }, + { + "epoch": 0.8927165354330708, + "grad_norm": 12.7373628616333, + "learning_rate": 8.9662027833002e-07, + "loss": 1.0457, + "step": 907 + }, + { + "epoch": 0.8937007874015748, + "grad_norm": 12.571308135986328, + "learning_rate": 8.976143141153082e-07, + "loss": 0.8604, + "step": 908 + }, + { + "epoch": 0.8946850393700787, + "grad_norm": 15.486084938049316, + "learning_rate": 8.986083499005966e-07, + "loss": 1.1543, + "step": 909 + }, + { + "epoch": 0.8956692913385826, + "grad_norm": 12.579278945922852, + "learning_rate": 8.996023856858848e-07, + "loss": 0.642, + "step": 910 + }, + { + "epoch": 0.8966535433070866, + "grad_norm": 20.628713607788086, + "learning_rate": 9.00596421471173e-07, + "loss": 1.0866, + "step": 911 + }, + { + "epoch": 0.8976377952755905, + "grad_norm": 24.661611557006836, + "learning_rate": 9.015904572564614e-07, + "loss": 1.5848, + "step": 912 + }, + { + "epoch": 0.8986220472440944, + "grad_norm": 31.02180290222168, + "learning_rate": 9.025844930417496e-07, + "loss": 1.7155, + "step": 913 + }, + { + "epoch": 0.8996062992125984, + "grad_norm": 13.382023811340332, + "learning_rate": 9.035785288270379e-07, + "loss": 0.8018, + "step": 914 + }, + { + "epoch": 0.9005905511811023, + "grad_norm": 12.55356216430664, + "learning_rate": 9.045725646123262e-07, + "loss": 0.879, + "step": 915 + }, + { + "epoch": 0.9015748031496063, + "grad_norm": 12.50846004486084, + "learning_rate": 9.055666003976144e-07, + "loss": 0.7473, + "step": 916 + }, + { + "epoch": 0.9025590551181102, + "grad_norm": 8.537793159484863, + "learning_rate": 9.065606361829028e-07, + "loss": 0.4899, + "step": 917 + }, + { + "epoch": 0.9035433070866141, + "grad_norm": 9.605969429016113, + "learning_rate": 9.07554671968191e-07, + "loss": 0.6282, + "step": 918 + }, + { + "epoch": 0.9035433070866141, + "eval_Qnli-dev_cosine_accuracy": 0.662109375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8222019672393799, + "eval_Qnli-dev_cosine_ap": 0.6690525867530521, + "eval_Qnli-dev_cosine_f1": 0.6626506024096385, + "eval_Qnli-dev_cosine_f1_threshold": 0.736522376537323, + "eval_Qnli-dev_cosine_precision": 0.514018691588785, + "eval_Qnli-dev_cosine_recall": 0.9322033898305084, + "eval_Qnli-dev_dot_accuracy": 0.662109375, + "eval_Qnli-dev_dot_accuracy_threshold": 630.4925537109375, + "eval_Qnli-dev_dot_ap": 0.6666029674148094, + "eval_Qnli-dev_dot_f1": 0.6635802469135803, + "eval_Qnli-dev_dot_f1_threshold": 569.1029052734375, + "eval_Qnli-dev_dot_precision": 0.5218446601941747, + "eval_Qnli-dev_dot_recall": 0.9110169491525424, + "eval_Qnli-dev_euclidean_accuracy": 0.666015625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 16.523191452026367, + "eval_Qnli-dev_euclidean_ap": 0.6688456278122232, + "eval_Qnli-dev_euclidean_f1": 0.6626506024096385, + "eval_Qnli-dev_euclidean_f1_threshold": 20.06503677368164, + "eval_Qnli-dev_euclidean_precision": 0.514018691588785, + "eval_Qnli-dev_euclidean_recall": 0.9322033898305084, + "eval_Qnli-dev_manhattan_accuracy": 0.666015625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 349.4272766113281, + "eval_Qnli-dev_manhattan_ap": 0.6815804028415055, + "eval_Qnli-dev_manhattan_f1": 0.6720257234726689, + "eval_Qnli-dev_manhattan_f1_threshold": 396.77142333984375, + "eval_Qnli-dev_manhattan_precision": 0.5414507772020726, + "eval_Qnli-dev_manhattan_recall": 0.885593220338983, + "eval_Qnli-dev_max_accuracy": 0.666015625, + "eval_Qnli-dev_max_accuracy_threshold": 630.4925537109375, + "eval_Qnli-dev_max_ap": 0.6815804028415055, + "eval_Qnli-dev_max_f1": 0.6720257234726689, + "eval_Qnli-dev_max_f1_threshold": 569.1029052734375, + "eval_Qnli-dev_max_precision": 0.5414507772020726, + "eval_Qnli-dev_max_recall": 0.9322033898305084, + "eval_allNLI-dev_cosine_accuracy": 0.685546875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9381226897239685, + "eval_allNLI-dev_cosine_ap": 0.49098585826236835, + "eval_allNLI-dev_cosine_f1": 0.5296610169491525, + "eval_allNLI-dev_cosine_f1_threshold": 0.8554738163948059, + "eval_allNLI-dev_cosine_precision": 0.4180602006688963, + "eval_allNLI-dev_cosine_recall": 0.7225433526011561, + "eval_allNLI-dev_dot_accuracy": 0.685546875, + "eval_allNLI-dev_dot_accuracy_threshold": 715.43310546875, + "eval_allNLI-dev_dot_ap": 0.49162275304477426, + "eval_allNLI-dev_dot_f1": 0.5311871227364185, + "eval_allNLI-dev_dot_f1_threshold": 646.8074951171875, + "eval_allNLI-dev_dot_precision": 0.4074074074074074, + "eval_allNLI-dev_dot_recall": 0.7630057803468208, + "eval_allNLI-dev_euclidean_accuracy": 0.685546875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.73391342163086, + "eval_allNLI-dev_euclidean_ap": 0.4908290509025364, + "eval_allNLI-dev_euclidean_f1": 0.5296610169491525, + "eval_allNLI-dev_euclidean_f1_threshold": 14.872737884521484, + "eval_allNLI-dev_euclidean_precision": 0.4180602006688963, + "eval_allNLI-dev_euclidean_recall": 0.7225433526011561, + "eval_allNLI-dev_manhattan_accuracy": 0.6875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 210.50613403320312, + "eval_allNLI-dev_manhattan_ap": 0.49318808387723423, + "eval_allNLI-dev_manhattan_f1": 0.5353319057815846, + "eval_allNLI-dev_manhattan_f1_threshold": 307.93359375, + "eval_allNLI-dev_manhattan_precision": 0.42517006802721086, + "eval_allNLI-dev_manhattan_recall": 0.7225433526011561, + "eval_allNLI-dev_max_accuracy": 0.6875, + "eval_allNLI-dev_max_accuracy_threshold": 715.43310546875, + "eval_allNLI-dev_max_ap": 0.49318808387723423, + "eval_allNLI-dev_max_f1": 0.5353319057815846, + "eval_allNLI-dev_max_f1_threshold": 646.8074951171875, + "eval_allNLI-dev_max_precision": 0.42517006802721086, + "eval_allNLI-dev_max_recall": 0.7630057803468208, + "eval_loss": 1.3715771436691284, + "eval_runtime": 54.2633, + "eval_samples_per_second": 30.665, + "eval_sequential_score": 0.6815804028415055, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.4591231950836093, + "eval_sts-test_pearson_dot": 0.4551866277132004, + "eval_sts-test_pearson_euclidean": 0.48026487148147445, + "eval_sts-test_pearson_manhattan": 0.49118985842833635, + "eval_sts-test_pearson_max": 0.49118985842833635, + "eval_sts-test_spearman_cosine": 0.4969829405703791, + "eval_sts-test_spearman_dot": 0.491899063022532, + "eval_sts-test_spearman_euclidean": 0.49736032795769275, + "eval_sts-test_spearman_manhattan": 0.5076723611201129, + "eval_sts-test_spearman_max": 0.5076723611201129, + "step": 918 + }, + { + "epoch": 0.9045275590551181, + "grad_norm": 15.066818237304688, + "learning_rate": 9.085487077534794e-07, + "loss": 1.2808, + "step": 919 + }, + { + "epoch": 0.905511811023622, + "grad_norm": 15.124238014221191, + "learning_rate": 9.095427435387676e-07, + "loss": 1.2134, + "step": 920 + }, + { + "epoch": 0.906496062992126, + "grad_norm": 10.836934089660645, + "learning_rate": 9.105367793240558e-07, + "loss": 0.6815, + "step": 921 + }, + { + "epoch": 0.90748031496063, + "grad_norm": 12.821209907531738, + "learning_rate": 9.115308151093442e-07, + "loss": 1.0554, + "step": 922 + }, + { + "epoch": 0.9084645669291339, + "grad_norm": 23.893203735351562, + "learning_rate": 9.125248508946324e-07, + "loss": 1.8122, + "step": 923 + }, + { + "epoch": 0.9094488188976378, + "grad_norm": 25.48162841796875, + "learning_rate": 9.135188866799206e-07, + "loss": 1.041, + "step": 924 + }, + { + "epoch": 0.9104330708661418, + "grad_norm": 10.63532829284668, + "learning_rate": 9.145129224652089e-07, + "loss": 0.4742, + "step": 925 + }, + { + "epoch": 0.9114173228346457, + "grad_norm": 13.105849266052246, + "learning_rate": 9.155069582504972e-07, + "loss": 0.986, + "step": 926 + }, + { + "epoch": 0.9124015748031497, + "grad_norm": 16.00253677368164, + "learning_rate": 9.165009940357855e-07, + "loss": 1.1247, + "step": 927 + }, + { + "epoch": 0.9133858267716536, + "grad_norm": 13.528996467590332, + "learning_rate": 9.174950298210737e-07, + "loss": 0.8918, + "step": 928 + }, + { + "epoch": 0.9143700787401575, + "grad_norm": 12.672636985778809, + "learning_rate": 9.18489065606362e-07, + "loss": 0.9735, + "step": 929 + }, + { + "epoch": 0.9153543307086615, + "grad_norm": 15.981063842773438, + "learning_rate": 9.194831013916503e-07, + "loss": 1.4467, + "step": 930 + }, + { + "epoch": 0.9163385826771654, + "grad_norm": 10.165881156921387, + "learning_rate": 9.204771371769385e-07, + "loss": 0.6623, + "step": 931 + }, + { + "epoch": 0.9173228346456693, + "grad_norm": 35.951602935791016, + "learning_rate": 9.214711729622268e-07, + "loss": 2.3602, + "step": 932 + }, + { + "epoch": 0.9183070866141733, + "grad_norm": 37.820743560791016, + "learning_rate": 9.224652087475151e-07, + "loss": 2.5324, + "step": 933 + }, + { + "epoch": 0.9192913385826772, + "grad_norm": 11.071194648742676, + "learning_rate": 9.234592445328033e-07, + "loss": 0.6558, + "step": 934 + }, + { + "epoch": 0.9202755905511811, + "grad_norm": 14.736655235290527, + "learning_rate": 9.244532803180916e-07, + "loss": 0.856, + "step": 935 + }, + { + "epoch": 0.9212598425196851, + "grad_norm": 23.33149528503418, + "learning_rate": 9.254473161033799e-07, + "loss": 1.8413, + "step": 936 + }, + { + "epoch": 0.922244094488189, + "grad_norm": 29.5730037689209, + "learning_rate": 9.264413518886681e-07, + "loss": 2.0715, + "step": 937 + }, + { + "epoch": 0.9232283464566929, + "grad_norm": 17.301631927490234, + "learning_rate": 9.274353876739564e-07, + "loss": 0.9652, + "step": 938 + }, + { + "epoch": 0.9242125984251969, + "grad_norm": 14.46304702758789, + "learning_rate": 9.284294234592446e-07, + "loss": 0.7404, + "step": 939 + }, + { + "epoch": 0.9251968503937008, + "grad_norm": 15.705856323242188, + "learning_rate": 9.29423459244533e-07, + "loss": 1.6298, + "step": 940 + }, + { + "epoch": 0.9261811023622047, + "grad_norm": 10.920563697814941, + "learning_rate": 9.304174950298212e-07, + "loss": 0.7701, + "step": 941 + }, + { + "epoch": 0.9271653543307087, + "grad_norm": 11.190475463867188, + "learning_rate": 9.314115308151094e-07, + "loss": 0.967, + "step": 942 + }, + { + "epoch": 0.9281496062992126, + "grad_norm": 10.520164489746094, + "learning_rate": 9.324055666003978e-07, + "loss": 0.934, + "step": 943 + }, + { + "epoch": 0.9291338582677166, + "grad_norm": 26.160831451416016, + "learning_rate": 9.33399602385686e-07, + "loss": 1.7068, + "step": 944 + }, + { + "epoch": 0.9301181102362205, + "grad_norm": 13.173468589782715, + "learning_rate": 9.343936381709743e-07, + "loss": 1.1774, + "step": 945 + }, + { + "epoch": 0.9311023622047244, + "grad_norm": 8.563469886779785, + "learning_rate": 9.353876739562625e-07, + "loss": 0.6848, + "step": 946 + }, + { + "epoch": 0.9320866141732284, + "grad_norm": 9.22998046875, + "learning_rate": 9.363817097415509e-07, + "loss": 0.6364, + "step": 947 + }, + { + "epoch": 0.9330708661417323, + "grad_norm": 7.827816963195801, + "learning_rate": 9.373757455268392e-07, + "loss": 0.6477, + "step": 948 + }, + { + "epoch": 0.9340551181102362, + "grad_norm": 27.504056930541992, + "learning_rate": 9.383697813121274e-07, + "loss": 1.5631, + "step": 949 + }, + { + "epoch": 0.9350393700787402, + "grad_norm": 9.704866409301758, + "learning_rate": 9.393638170974157e-07, + "loss": 0.7204, + "step": 950 + }, + { + "epoch": 0.9360236220472441, + "grad_norm": 9.747822761535645, + "learning_rate": 9.40357852882704e-07, + "loss": 0.7169, + "step": 951 + }, + { + "epoch": 0.937007874015748, + "grad_norm": 10.555800437927246, + "learning_rate": 9.413518886679922e-07, + "loss": 0.7335, + "step": 952 + }, + { + "epoch": 0.937992125984252, + "grad_norm": 7.41068172454834, + "learning_rate": 9.423459244532805e-07, + "loss": 0.5237, + "step": 953 + }, + { + "epoch": 0.9389763779527559, + "grad_norm": 11.667238235473633, + "learning_rate": 9.433399602385688e-07, + "loss": 0.6509, + "step": 954 + }, + { + "epoch": 0.9399606299212598, + "grad_norm": 13.451628684997559, + "learning_rate": 9.44333996023857e-07, + "loss": 0.8209, + "step": 955 + }, + { + "epoch": 0.9409448818897638, + "grad_norm": 9.289586067199707, + "learning_rate": 9.453280318091453e-07, + "loss": 0.509, + "step": 956 + }, + { + "epoch": 0.9419291338582677, + "grad_norm": 12.067706108093262, + "learning_rate": 9.463220675944336e-07, + "loss": 0.8048, + "step": 957 + }, + { + "epoch": 0.9429133858267716, + "grad_norm": 10.956072807312012, + "learning_rate": 9.473161033797219e-07, + "loss": 0.663, + "step": 958 + }, + { + "epoch": 0.9438976377952756, + "grad_norm": 10.91523551940918, + "learning_rate": 9.483101391650101e-07, + "loss": 0.714, + "step": 959 + }, + { + "epoch": 0.9448818897637795, + "grad_norm": 29.80571174621582, + "learning_rate": 9.493041749502983e-07, + "loss": 2.1961, + "step": 960 + }, + { + "epoch": 0.9458661417322834, + "grad_norm": 12.550529479980469, + "learning_rate": 9.502982107355867e-07, + "loss": 0.8293, + "step": 961 + }, + { + "epoch": 0.9468503937007874, + "grad_norm": 22.91265296936035, + "learning_rate": 9.512922465208749e-07, + "loss": 1.4108, + "step": 962 + }, + { + "epoch": 0.9478346456692913, + "grad_norm": 15.225966453552246, + "learning_rate": 9.522862823061631e-07, + "loss": 1.2291, + "step": 963 + }, + { + "epoch": 0.9488188976377953, + "grad_norm": 26.223369598388672, + "learning_rate": 9.532803180914515e-07, + "loss": 1.7917, + "step": 964 + }, + { + "epoch": 0.9498031496062992, + "grad_norm": 25.121417999267578, + "learning_rate": 9.542743538767396e-07, + "loss": 1.5266, + "step": 965 + }, + { + "epoch": 0.9507874015748031, + "grad_norm": 17.08681869506836, + "learning_rate": 9.55268389662028e-07, + "loss": 1.4995, + "step": 966 + }, + { + "epoch": 0.9517716535433071, + "grad_norm": 14.251176834106445, + "learning_rate": 9.562624254473162e-07, + "loss": 1.1383, + "step": 967 + }, + { + "epoch": 0.952755905511811, + "grad_norm": 9.796082496643066, + "learning_rate": 9.572564612326044e-07, + "loss": 0.7347, + "step": 968 + }, + { + "epoch": 0.9537401574803149, + "grad_norm": 16.32306480407715, + "learning_rate": 9.582504970178928e-07, + "loss": 1.2644, + "step": 969 + }, + { + "epoch": 0.9547244094488189, + "grad_norm": 13.050923347473145, + "learning_rate": 9.59244532803181e-07, + "loss": 0.8973, + "step": 970 + }, + { + "epoch": 0.9557086614173228, + "grad_norm": 11.90676498413086, + "learning_rate": 9.602385685884694e-07, + "loss": 0.8478, + "step": 971 + }, + { + "epoch": 0.9566929133858267, + "grad_norm": 9.020946502685547, + "learning_rate": 9.612326043737576e-07, + "loss": 0.5924, + "step": 972 + }, + { + "epoch": 0.9576771653543307, + "grad_norm": 12.121260643005371, + "learning_rate": 9.622266401590458e-07, + "loss": 1.0661, + "step": 973 + }, + { + "epoch": 0.9586614173228346, + "grad_norm": 13.3746337890625, + "learning_rate": 9.632206759443342e-07, + "loss": 1.3449, + "step": 974 + }, + { + "epoch": 0.9596456692913385, + "grad_norm": 11.143105506896973, + "learning_rate": 9.642147117296224e-07, + "loss": 0.8305, + "step": 975 + }, + { + "epoch": 0.9606299212598425, + "grad_norm": 14.122886657714844, + "learning_rate": 9.652087475149109e-07, + "loss": 1.3399, + "step": 976 + }, + { + "epoch": 0.9616141732283464, + "grad_norm": 26.04822540283203, + "learning_rate": 9.66202783300199e-07, + "loss": 2.1596, + "step": 977 + }, + { + "epoch": 0.9625984251968503, + "grad_norm": 11.19150447845459, + "learning_rate": 9.671968190854873e-07, + "loss": 0.6784, + "step": 978 + }, + { + "epoch": 0.9635826771653543, + "grad_norm": 18.581256866455078, + "learning_rate": 9.681908548707755e-07, + "loss": 1.3365, + "step": 979 + }, + { + "epoch": 0.9645669291338582, + "grad_norm": 12.220732688903809, + "learning_rate": 9.691848906560639e-07, + "loss": 0.8564, + "step": 980 + }, + { + "epoch": 0.9655511811023622, + "grad_norm": 12.121989250183105, + "learning_rate": 9.70178926441352e-07, + "loss": 0.9305, + "step": 981 + }, + { + "epoch": 0.9665354330708661, + "grad_norm": 12.424619674682617, + "learning_rate": 9.711729622266403e-07, + "loss": 1.0597, + "step": 982 + }, + { + "epoch": 0.96751968503937, + "grad_norm": 25.869260787963867, + "learning_rate": 9.721669980119285e-07, + "loss": 2.0984, + "step": 983 + }, + { + "epoch": 0.968503937007874, + "grad_norm": 12.994894027709961, + "learning_rate": 9.731610337972169e-07, + "loss": 1.2005, + "step": 984 + }, + { + "epoch": 0.969488188976378, + "grad_norm": 11.917289733886719, + "learning_rate": 9.74155069582505e-07, + "loss": 0.9083, + "step": 985 + }, + { + "epoch": 0.9704724409448819, + "grad_norm": 24.574920654296875, + "learning_rate": 9.751491053677933e-07, + "loss": 1.8997, + "step": 986 + }, + { + "epoch": 0.9714566929133859, + "grad_norm": 9.57914924621582, + "learning_rate": 9.761431411530817e-07, + "loss": 0.7606, + "step": 987 + }, + { + "epoch": 0.9724409448818898, + "grad_norm": 11.350131034851074, + "learning_rate": 9.771371769383699e-07, + "loss": 0.6725, + "step": 988 + }, + { + "epoch": 0.9734251968503937, + "grad_norm": 11.479193687438965, + "learning_rate": 9.781312127236583e-07, + "loss": 0.8839, + "step": 989 + }, + { + "epoch": 0.9744094488188977, + "grad_norm": 8.321087837219238, + "learning_rate": 9.791252485089465e-07, + "loss": 0.4199, + "step": 990 + }, + { + "epoch": 0.9753937007874016, + "grad_norm": 10.547780990600586, + "learning_rate": 9.801192842942347e-07, + "loss": 0.6008, + "step": 991 + }, + { + "epoch": 0.9763779527559056, + "grad_norm": 6.902090549468994, + "learning_rate": 9.811133200795231e-07, + "loss": 0.3017, + "step": 992 + }, + { + "epoch": 0.9773622047244095, + "grad_norm": 9.781502723693848, + "learning_rate": 9.821073558648113e-07, + "loss": 0.5473, + "step": 993 + }, + { + "epoch": 0.9783464566929134, + "grad_norm": 12.86219310760498, + "learning_rate": 9.831013916500995e-07, + "loss": 0.854, + "step": 994 + }, + { + "epoch": 0.9793307086614174, + "grad_norm": 12.871740341186523, + "learning_rate": 9.84095427435388e-07, + "loss": 0.961, + "step": 995 + }, + { + "epoch": 0.9803149606299213, + "grad_norm": 12.065014839172363, + "learning_rate": 9.850894632206761e-07, + "loss": 0.8797, + "step": 996 + }, + { + "epoch": 0.9812992125984252, + "grad_norm": 15.222761154174805, + "learning_rate": 9.860834990059645e-07, + "loss": 0.9166, + "step": 997 + }, + { + "epoch": 0.9822834645669292, + "grad_norm": 13.594207763671875, + "learning_rate": 9.870775347912527e-07, + "loss": 0.8646, + "step": 998 + }, + { + "epoch": 0.9832677165354331, + "grad_norm": 9.33657455444336, + "learning_rate": 9.88071570576541e-07, + "loss": 0.4182, + "step": 999 + }, + { + "epoch": 0.984251968503937, + "grad_norm": 13.540481567382812, + "learning_rate": 9.890656063618291e-07, + "loss": 1.1083, + "step": 1000 + }, + { + "epoch": 0.985236220472441, + "grad_norm": 27.116676330566406, + "learning_rate": 9.900596421471175e-07, + "loss": 1.6306, + "step": 1001 + }, + { + "epoch": 0.9862204724409449, + "grad_norm": 40.88335037231445, + "learning_rate": 9.910536779324057e-07, + "loss": 3.1844, + "step": 1002 + }, + { + "epoch": 0.9872047244094488, + "grad_norm": 11.378805160522461, + "learning_rate": 9.92047713717694e-07, + "loss": 0.7701, + "step": 1003 + }, + { + "epoch": 0.9881889763779528, + "grad_norm": 13.258501052856445, + "learning_rate": 9.930417495029821e-07, + "loss": 0.8628, + "step": 1004 + }, + { + "epoch": 0.9891732283464567, + "grad_norm": 11.201334953308105, + "learning_rate": 9.940357852882705e-07, + "loss": 0.6919, + "step": 1005 + }, + { + "epoch": 0.9901574803149606, + "grad_norm": 10.30268669128418, + "learning_rate": 9.950298210735587e-07, + "loss": 0.5191, + "step": 1006 + }, + { + "epoch": 0.9911417322834646, + "grad_norm": 16.8773136138916, + "learning_rate": 9.96023856858847e-07, + "loss": 0.8675, + "step": 1007 + }, + { + "epoch": 0.9921259842519685, + "grad_norm": 21.878103256225586, + "learning_rate": 9.970178926441354e-07, + "loss": 1.4837, + "step": 1008 + }, + { + "epoch": 0.9931102362204725, + "grad_norm": 25.748367309570312, + "learning_rate": 9.980119284294236e-07, + "loss": 1.5177, + "step": 1009 + }, + { + "epoch": 0.9940944881889764, + "grad_norm": 16.585498809814453, + "learning_rate": 9.99005964214712e-07, + "loss": 1.0422, + "step": 1010 + }, + { + "epoch": 0.9950787401574803, + "grad_norm": 20.055673599243164, + "learning_rate": 1.0000000000000002e-06, + "loss": 1.4038, + "step": 1011 + }, + { + "epoch": 0.9960629921259843, + "grad_norm": 10.144420623779297, + "learning_rate": 9.999994102361802e-07, + "loss": 0.6402, + "step": 1012 + }, + { + "epoch": 0.9970472440944882, + "grad_norm": 17.399948120117188, + "learning_rate": 9.999976409461164e-07, + "loss": 1.0376, + "step": 1013 + }, + { + "epoch": 0.9980314960629921, + "grad_norm": 16.318910598754883, + "learning_rate": 9.999946921339963e-07, + "loss": 1.5319, + "step": 1014 + }, + { + "epoch": 0.9990157480314961, + "grad_norm": 12.954554557800293, + "learning_rate": 9.999905638068e-07, + "loss": 1.0956, + "step": 1015 + }, + { + "epoch": 1.0, + "grad_norm": 14.94405460357666, + "learning_rate": 9.999852559742985e-07, + "loss": 0.6806, + "step": 1016 + }, + { + "epoch": 1.000984251968504, + "grad_norm": 0.0, + "learning_rate": 9.999787686490555e-07, + "loss": 0.0, + "step": 1017 + }, + { + "epoch": 1.000984251968504, + "grad_norm": 13.080833435058594, + "learning_rate": 9.99971101846426e-07, + "loss": 1.0824, + "step": 1018 + }, + { + "epoch": 1.0019685039370079, + "grad_norm": 15.298651695251465, + "learning_rate": 9.999622555845573e-07, + "loss": 1.0969, + "step": 1019 + }, + { + "epoch": 1.0029527559055118, + "grad_norm": 11.248276710510254, + "learning_rate": 9.999522298843875e-07, + "loss": 0.8334, + "step": 1020 + }, + { + "epoch": 1.0039370078740157, + "grad_norm": 10.074872016906738, + "learning_rate": 9.99941024769647e-07, + "loss": 0.7022, + "step": 1021 + }, + { + "epoch": 1.0049212598425197, + "grad_norm": 8.650634765625, + "learning_rate": 9.999286402668578e-07, + "loss": 0.502, + "step": 1022 + }, + { + "epoch": 1.0059055118110236, + "grad_norm": 8.788339614868164, + "learning_rate": 9.999150764053334e-07, + "loss": 0.5146, + "step": 1023 + }, + { + "epoch": 1.0068897637795275, + "grad_norm": 12.089606285095215, + "learning_rate": 9.999003332171784e-07, + "loss": 0.9768, + "step": 1024 + }, + { + "epoch": 1.0078740157480315, + "grad_norm": 11.496919631958008, + "learning_rate": 9.998844107372894e-07, + "loss": 0.7295, + "step": 1025 + }, + { + "epoch": 1.0088582677165354, + "grad_norm": 51.444732666015625, + "learning_rate": 9.99867309003354e-07, + "loss": 2.0448, + "step": 1026 + }, + { + "epoch": 1.0098425196850394, + "grad_norm": 11.470032691955566, + "learning_rate": 9.998490280558511e-07, + "loss": 0.7609, + "step": 1027 + }, + { + "epoch": 1.0108267716535433, + "grad_norm": 14.040739059448242, + "learning_rate": 9.998295679380506e-07, + "loss": 1.3264, + "step": 1028 + }, + { + "epoch": 1.0118110236220472, + "grad_norm": 16.276897430419922, + "learning_rate": 9.998089286960135e-07, + "loss": 1.1084, + "step": 1029 + }, + { + "epoch": 1.0127952755905512, + "grad_norm": 13.747846603393555, + "learning_rate": 9.99787110378592e-07, + "loss": 1.0857, + "step": 1030 + }, + { + "epoch": 1.013779527559055, + "grad_norm": 34.06081008911133, + "learning_rate": 9.997641130374284e-07, + "loss": 1.7838, + "step": 1031 + }, + { + "epoch": 1.014763779527559, + "grad_norm": 23.071144104003906, + "learning_rate": 9.997399367269565e-07, + "loss": 1.2398, + "step": 1032 + }, + { + "epoch": 1.015748031496063, + "grad_norm": 11.253214836120605, + "learning_rate": 9.997145815044003e-07, + "loss": 0.7096, + "step": 1033 + }, + { + "epoch": 1.016732283464567, + "grad_norm": 10.326289176940918, + "learning_rate": 9.99688047429774e-07, + "loss": 0.7082, + "step": 1034 + }, + { + "epoch": 1.0177165354330708, + "grad_norm": 10.67347526550293, + "learning_rate": 9.996603345658824e-07, + "loss": 0.6649, + "step": 1035 + }, + { + "epoch": 1.0187007874015748, + "grad_norm": 20.553695678710938, + "learning_rate": 9.996314429783204e-07, + "loss": 1.6809, + "step": 1036 + }, + { + "epoch": 1.0196850393700787, + "grad_norm": 14.0806303024292, + "learning_rate": 9.99601372735473e-07, + "loss": 1.0725, + "step": 1037 + }, + { + "epoch": 1.0206692913385826, + "grad_norm": 12.312373161315918, + "learning_rate": 9.995701239085143e-07, + "loss": 0.7607, + "step": 1038 + }, + { + "epoch": 1.0216535433070866, + "grad_norm": 13.746301651000977, + "learning_rate": 9.99537696571409e-07, + "loss": 0.8318, + "step": 1039 + }, + { + "epoch": 1.0226377952755905, + "grad_norm": 27.883398056030273, + "learning_rate": 9.995040908009107e-07, + "loss": 1.2795, + "step": 1040 + }, + { + "epoch": 1.0236220472440944, + "grad_norm": 12.371521949768066, + "learning_rate": 9.994693066765622e-07, + "loss": 0.5544, + "step": 1041 + }, + { + "epoch": 1.0246062992125984, + "grad_norm": 10.142796516418457, + "learning_rate": 9.994333442806959e-07, + "loss": 0.6573, + "step": 1042 + }, + { + "epoch": 1.0255905511811023, + "grad_norm": 13.032113075256348, + "learning_rate": 9.993962036984325e-07, + "loss": 0.868, + "step": 1043 + }, + { + "epoch": 1.0265748031496063, + "grad_norm": 15.365907669067383, + "learning_rate": 9.99357885017682e-07, + "loss": 1.1831, + "step": 1044 + }, + { + "epoch": 1.0275590551181102, + "grad_norm": 14.052511215209961, + "learning_rate": 9.993183883291425e-07, + "loss": 0.9724, + "step": 1045 + }, + { + "epoch": 1.0285433070866141, + "grad_norm": 16.52604866027832, + "learning_rate": 9.992777137263007e-07, + "loss": 1.2018, + "step": 1046 + }, + { + "epoch": 1.029527559055118, + "grad_norm": 14.135658264160156, + "learning_rate": 9.99235861305431e-07, + "loss": 0.9539, + "step": 1047 + }, + { + "epoch": 1.030511811023622, + "grad_norm": 9.573715209960938, + "learning_rate": 9.991928311655956e-07, + "loss": 0.5348, + "step": 1048 + }, + { + "epoch": 1.031496062992126, + "grad_norm": 11.995317459106445, + "learning_rate": 9.991486234086447e-07, + "loss": 0.7531, + "step": 1049 + }, + { + "epoch": 1.0324803149606299, + "grad_norm": 14.20138168334961, + "learning_rate": 9.991032381392158e-07, + "loss": 1.1062, + "step": 1050 + }, + { + "epoch": 1.0334645669291338, + "grad_norm": 13.821763038635254, + "learning_rate": 9.990566754647328e-07, + "loss": 1.067, + "step": 1051 + }, + { + "epoch": 1.0344488188976377, + "grad_norm": 13.316102027893066, + "learning_rate": 9.990089354954073e-07, + "loss": 0.866, + "step": 1052 + }, + { + "epoch": 1.0354330708661417, + "grad_norm": 15.538060188293457, + "learning_rate": 9.989600183442373e-07, + "loss": 1.2976, + "step": 1053 + }, + { + "epoch": 1.0364173228346456, + "grad_norm": 25.953405380249023, + "learning_rate": 9.98909924127007e-07, + "loss": 2.035, + "step": 1054 + }, + { + "epoch": 1.0374015748031495, + "grad_norm": 14.323229789733887, + "learning_rate": 9.988586529622864e-07, + "loss": 1.1908, + "step": 1055 + }, + { + "epoch": 1.0383858267716535, + "grad_norm": 13.404814720153809, + "learning_rate": 9.98806204971432e-07, + "loss": 0.5581, + "step": 1056 + }, + { + "epoch": 1.0393700787401574, + "grad_norm": 14.532567977905273, + "learning_rate": 9.987525802785847e-07, + "loss": 1.0315, + "step": 1057 + }, + { + "epoch": 1.0403543307086613, + "grad_norm": 13.722807884216309, + "learning_rate": 9.986977790106715e-07, + "loss": 1.1023, + "step": 1058 + }, + { + "epoch": 1.0413385826771653, + "grad_norm": 10.097068786621094, + "learning_rate": 9.986418012974042e-07, + "loss": 0.4518, + "step": 1059 + }, + { + "epoch": 1.0423228346456692, + "grad_norm": 13.188447952270508, + "learning_rate": 9.985846472712786e-07, + "loss": 0.8027, + "step": 1060 + }, + { + "epoch": 1.0433070866141732, + "grad_norm": 27.605440139770508, + "learning_rate": 9.985263170675756e-07, + "loss": 1.9934, + "step": 1061 + }, + { + "epoch": 1.044291338582677, + "grad_norm": 12.822807312011719, + "learning_rate": 9.98466810824359e-07, + "loss": 0.9329, + "step": 1062 + }, + { + "epoch": 1.045275590551181, + "grad_norm": 14.287766456604004, + "learning_rate": 9.984061286824774e-07, + "loss": 0.9931, + "step": 1063 + }, + { + "epoch": 1.046259842519685, + "grad_norm": 14.493118286132812, + "learning_rate": 9.983442707855616e-07, + "loss": 0.7169, + "step": 1064 + }, + { + "epoch": 1.047244094488189, + "grad_norm": 20.291234970092773, + "learning_rate": 9.982812372800262e-07, + "loss": 1.0229, + "step": 1065 + }, + { + "epoch": 1.0482283464566928, + "grad_norm": 10.355449676513672, + "learning_rate": 9.982170283150678e-07, + "loss": 0.6757, + "step": 1066 + }, + { + "epoch": 1.0492125984251968, + "grad_norm": 13.256071090698242, + "learning_rate": 9.981516440426659e-07, + "loss": 0.7128, + "step": 1067 + }, + { + "epoch": 1.0501968503937007, + "grad_norm": 8.21869945526123, + "learning_rate": 9.98085084617581e-07, + "loss": 0.3775, + "step": 1068 + }, + { + "epoch": 1.0511811023622046, + "grad_norm": 14.035961151123047, + "learning_rate": 9.980173501973558e-07, + "loss": 1.1158, + "step": 1069 + }, + { + "epoch": 1.0521653543307086, + "grad_norm": 14.340359687805176, + "learning_rate": 9.97948440942314e-07, + "loss": 1.084, + "step": 1070 + }, + { + "epoch": 1.0531496062992125, + "grad_norm": 16.62276840209961, + "learning_rate": 9.978783570155597e-07, + "loss": 1.0829, + "step": 1071 + }, + { + "epoch": 1.0531496062992125, + "eval_Qnli-dev_cosine_accuracy": 0.654296875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.842217206954956, + "eval_Qnli-dev_cosine_ap": 0.6736642973454704, + "eval_Qnli-dev_cosine_f1": 0.6656534954407294, + "eval_Qnli-dev_cosine_f1_threshold": 0.7517247200012207, + "eval_Qnli-dev_cosine_precision": 0.518957345971564, + "eval_Qnli-dev_cosine_recall": 0.9279661016949152, + "eval_Qnli-dev_dot_accuracy": 0.658203125, + "eval_Qnli-dev_dot_accuracy_threshold": 660.471923828125, + "eval_Qnli-dev_dot_ap": 0.6716435318899119, + "eval_Qnli-dev_dot_f1": 0.6666666666666667, + "eval_Qnli-dev_dot_f1_threshold": 575.9989013671875, + "eval_Qnli-dev_dot_precision": 0.5228915662650603, + "eval_Qnli-dev_dot_recall": 0.9194915254237288, + "eval_Qnli-dev_euclidean_accuracy": 0.65625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.113401412963867, + "eval_Qnli-dev_euclidean_ap": 0.673907896308458, + "eval_Qnli-dev_euclidean_f1": 0.6666666666666667, + "eval_Qnli-dev_euclidean_f1_threshold": 19.382671356201172, + "eval_Qnli-dev_euclidean_precision": 0.5228915662650603, + "eval_Qnli-dev_euclidean_recall": 0.9194915254237288, + "eval_Qnli-dev_manhattan_accuracy": 0.66015625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 342.40484619140625, + "eval_Qnli-dev_manhattan_ap": 0.6866744707607304, + "eval_Qnli-dev_manhattan_f1": 0.6666666666666667, + "eval_Qnli-dev_manhattan_f1_threshold": 357.13262939453125, + "eval_Qnli-dev_manhattan_precision": 0.5846645367412141, + "eval_Qnli-dev_manhattan_recall": 0.7754237288135594, + "eval_Qnli-dev_max_accuracy": 0.66015625, + "eval_Qnli-dev_max_accuracy_threshold": 660.471923828125, + "eval_Qnli-dev_max_ap": 0.6866744707607304, + "eval_Qnli-dev_max_f1": 0.6666666666666667, + "eval_Qnli-dev_max_f1_threshold": 575.9989013671875, + "eval_Qnli-dev_max_precision": 0.5846645367412141, + "eval_Qnli-dev_max_recall": 0.9279661016949152, + "eval_allNLI-dev_cosine_accuracy": 0.689453125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9341496229171753, + "eval_allNLI-dev_cosine_ap": 0.489327543271253, + "eval_allNLI-dev_cosine_f1": 0.5300668151447662, + "eval_allNLI-dev_cosine_f1_threshold": 0.865491509437561, + "eval_allNLI-dev_cosine_precision": 0.4311594202898551, + "eval_allNLI-dev_cosine_recall": 0.6878612716763006, + "eval_allNLI-dev_dot_accuracy": 0.689453125, + "eval_allNLI-dev_dot_accuracy_threshold": 715.212890625, + "eval_allNLI-dev_dot_ap": 0.4881798241956048, + "eval_allNLI-dev_dot_f1": 0.5309734513274337, + "eval_allNLI-dev_dot_f1_threshold": 661.7647094726562, + "eval_allNLI-dev_dot_precision": 0.43010752688172044, + "eval_allNLI-dev_dot_recall": 0.6936416184971098, + "eval_allNLI-dev_euclidean_accuracy": 0.689453125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.04066276550293, + "eval_allNLI-dev_euclidean_ap": 0.4892035422795702, + "eval_allNLI-dev_euclidean_f1": 0.5330296127562643, + "eval_allNLI-dev_euclidean_f1_threshold": 14.194154739379883, + "eval_allNLI-dev_euclidean_precision": 0.4398496240601504, + "eval_allNLI-dev_euclidean_recall": 0.6763005780346821, + "eval_allNLI-dev_manhattan_accuracy": 0.6875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 212.68006896972656, + "eval_allNLI-dev_manhattan_ap": 0.4893894820749276, + "eval_allNLI-dev_manhattan_f1": 0.5400457665903889, + "eval_allNLI-dev_manhattan_f1_threshold": 294.4931640625, + "eval_allNLI-dev_manhattan_precision": 0.44696969696969696, + "eval_allNLI-dev_manhattan_recall": 0.6820809248554913, + "eval_allNLI-dev_max_accuracy": 0.689453125, + "eval_allNLI-dev_max_accuracy_threshold": 715.212890625, + "eval_allNLI-dev_max_ap": 0.4893894820749276, + "eval_allNLI-dev_max_f1": 0.5400457665903889, + "eval_allNLI-dev_max_f1_threshold": 661.7647094726562, + "eval_allNLI-dev_max_precision": 0.44696969696969696, + "eval_allNLI-dev_max_recall": 0.6936416184971098, + "eval_loss": 1.2497835159301758, + "eval_runtime": 54.1817, + "eval_samples_per_second": 30.711, + "eval_sequential_score": 0.6866744707607304, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.5218244250314455, + "eval_sts-test_pearson_dot": 0.5186838879324001, + "eval_sts-test_pearson_euclidean": 0.5454337101963522, + "eval_sts-test_pearson_manhattan": 0.554711626366034, + "eval_sts-test_pearson_max": 0.554711626366034, + "eval_sts-test_spearman_cosine": 0.5611271163344802, + "eval_sts-test_spearman_dot": 0.5566891784935565, + "eval_sts-test_spearman_euclidean": 0.5611494472139146, + "eval_sts-test_spearman_manhattan": 0.5698060417149441, + "eval_sts-test_spearman_max": 0.5698060417149441, + "step": 1071 + }, + { + "epoch": 1.0541338582677164, + "grad_norm": 28.540851593017578, + "learning_rate": 9.978070985829782e-07, + "loss": 1.4476, + "step": 1072 + }, + { + "epoch": 1.0551181102362204, + "grad_norm": 17.484149932861328, + "learning_rate": 9.97734665813234e-07, + "loss": 1.008, + "step": 1073 + }, + { + "epoch": 1.0561023622047243, + "grad_norm": 13.71534538269043, + "learning_rate": 9.976610588777717e-07, + "loss": 0.8832, + "step": 1074 + }, + { + "epoch": 1.0570866141732282, + "grad_norm": 9.226608276367188, + "learning_rate": 9.975862779508149e-07, + "loss": 0.5766, + "step": 1075 + }, + { + "epoch": 1.0580708661417322, + "grad_norm": 10.914620399475098, + "learning_rate": 9.975103232093656e-07, + "loss": 0.8722, + "step": 1076 + }, + { + "epoch": 1.0590551181102361, + "grad_norm": 10.53770923614502, + "learning_rate": 9.974331948332048e-07, + "loss": 0.6285, + "step": 1077 + }, + { + "epoch": 1.06003937007874, + "grad_norm": 11.654394149780273, + "learning_rate": 9.97354893004891e-07, + "loss": 0.8074, + "step": 1078 + }, + { + "epoch": 1.061023622047244, + "grad_norm": 11.405344009399414, + "learning_rate": 9.972754179097602e-07, + "loss": 0.7646, + "step": 1079 + }, + { + "epoch": 1.0620078740157481, + "grad_norm": 8.713008880615234, + "learning_rate": 9.971947697359257e-07, + "loss": 0.3947, + "step": 1080 + }, + { + "epoch": 1.0629921259842519, + "grad_norm": 14.822614669799805, + "learning_rate": 9.971129486742777e-07, + "loss": 1.1946, + "step": 1081 + }, + { + "epoch": 1.063976377952756, + "grad_norm": 13.049259185791016, + "learning_rate": 9.970299549184814e-07, + "loss": 1.3833, + "step": 1082 + }, + { + "epoch": 1.0649606299212597, + "grad_norm": 24.78580665588379, + "learning_rate": 9.969457886649791e-07, + "loss": 1.5897, + "step": 1083 + }, + { + "epoch": 1.0659448818897639, + "grad_norm": 24.568845748901367, + "learning_rate": 9.968604501129877e-07, + "loss": 1.7621, + "step": 1084 + }, + { + "epoch": 1.0669291338582678, + "grad_norm": 15.05627155303955, + "learning_rate": 9.967739394644982e-07, + "loss": 1.4584, + "step": 1085 + }, + { + "epoch": 1.0679133858267718, + "grad_norm": 11.705534934997559, + "learning_rate": 9.966862569242773e-07, + "loss": 0.715, + "step": 1086 + }, + { + "epoch": 1.0688976377952757, + "grad_norm": 11.141972541809082, + "learning_rate": 9.965974026998642e-07, + "loss": 0.7493, + "step": 1087 + }, + { + "epoch": 1.0698818897637796, + "grad_norm": 11.882009506225586, + "learning_rate": 9.965073770015722e-07, + "loss": 0.7076, + "step": 1088 + }, + { + "epoch": 1.0708661417322836, + "grad_norm": 16.43975067138672, + "learning_rate": 9.964161800424873e-07, + "loss": 0.7919, + "step": 1089 + }, + { + "epoch": 1.0718503937007875, + "grad_norm": 11.75568675994873, + "learning_rate": 9.963238120384677e-07, + "loss": 0.9057, + "step": 1090 + }, + { + "epoch": 1.0728346456692914, + "grad_norm": 9.294759750366211, + "learning_rate": 9.962302732081432e-07, + "loss": 0.5515, + "step": 1091 + }, + { + "epoch": 1.0738188976377954, + "grad_norm": 9.674452781677246, + "learning_rate": 9.961355637729153e-07, + "loss": 0.5275, + "step": 1092 + }, + { + "epoch": 1.0748031496062993, + "grad_norm": 10.010574340820312, + "learning_rate": 9.960396839569557e-07, + "loss": 0.4794, + "step": 1093 + }, + { + "epoch": 1.0757874015748032, + "grad_norm": 23.5357608795166, + "learning_rate": 9.95942633987207e-07, + "loss": 1.4877, + "step": 1094 + }, + { + "epoch": 1.0767716535433072, + "grad_norm": 18.584951400756836, + "learning_rate": 9.95844414093381e-07, + "loss": 1.1599, + "step": 1095 + }, + { + "epoch": 1.077755905511811, + "grad_norm": 10.287535667419434, + "learning_rate": 9.957450245079588e-07, + "loss": 0.5241, + "step": 1096 + }, + { + "epoch": 1.078740157480315, + "grad_norm": 21.15087890625, + "learning_rate": 9.9564446546619e-07, + "loss": 1.7957, + "step": 1097 + }, + { + "epoch": 1.079724409448819, + "grad_norm": 13.045212745666504, + "learning_rate": 9.955427372060928e-07, + "loss": 1.1295, + "step": 1098 + }, + { + "epoch": 1.080708661417323, + "grad_norm": 19.12482452392578, + "learning_rate": 9.954398399684515e-07, + "loss": 1.5688, + "step": 1099 + }, + { + "epoch": 1.0816929133858268, + "grad_norm": 9.251370429992676, + "learning_rate": 9.953357739968192e-07, + "loss": 0.7658, + "step": 1100 + }, + { + "epoch": 1.0826771653543308, + "grad_norm": 9.279908180236816, + "learning_rate": 9.952305395375137e-07, + "loss": 0.5828, + "step": 1101 + }, + { + "epoch": 1.0836614173228347, + "grad_norm": 10.163260459899902, + "learning_rate": 9.951241368396194e-07, + "loss": 0.8962, + "step": 1102 + }, + { + "epoch": 1.0846456692913387, + "grad_norm": 10.750070571899414, + "learning_rate": 9.950165661549856e-07, + "loss": 0.9649, + "step": 1103 + }, + { + "epoch": 1.0856299212598426, + "grad_norm": 10.144333839416504, + "learning_rate": 9.949078277382262e-07, + "loss": 0.8328, + "step": 1104 + }, + { + "epoch": 1.0866141732283465, + "grad_norm": 19.090774536132812, + "learning_rate": 9.947979218467193e-07, + "loss": 1.2495, + "step": 1105 + }, + { + "epoch": 1.0875984251968505, + "grad_norm": 12.030488967895508, + "learning_rate": 9.946868487406057e-07, + "loss": 1.1155, + "step": 1106 + }, + { + "epoch": 1.0885826771653544, + "grad_norm": 10.546430587768555, + "learning_rate": 9.945746086827896e-07, + "loss": 0.6959, + "step": 1107 + }, + { + "epoch": 1.0895669291338583, + "grad_norm": 7.1781415939331055, + "learning_rate": 9.94461201938937e-07, + "loss": 0.3202, + "step": 1108 + }, + { + "epoch": 1.0905511811023623, + "grad_norm": 12.0957670211792, + "learning_rate": 9.943466287774756e-07, + "loss": 0.7917, + "step": 1109 + }, + { + "epoch": 1.0915354330708662, + "grad_norm": 21.6634578704834, + "learning_rate": 9.942308894695932e-07, + "loss": 1.6036, + "step": 1110 + }, + { + "epoch": 1.0925196850393701, + "grad_norm": 19.509628295898438, + "learning_rate": 9.941139842892391e-07, + "loss": 1.2433, + "step": 1111 + }, + { + "epoch": 1.093503937007874, + "grad_norm": 10.067360877990723, + "learning_rate": 9.939959135131214e-07, + "loss": 0.5881, + "step": 1112 + }, + { + "epoch": 1.094488188976378, + "grad_norm": 15.549373626708984, + "learning_rate": 9.938766774207064e-07, + "loss": 1.2943, + "step": 1113 + }, + { + "epoch": 1.095472440944882, + "grad_norm": 8.6616792678833, + "learning_rate": 9.937562762942202e-07, + "loss": 0.4795, + "step": 1114 + }, + { + "epoch": 1.0964566929133859, + "grad_norm": 17.8710994720459, + "learning_rate": 9.936347104186452e-07, + "loss": 1.2879, + "step": 1115 + }, + { + "epoch": 1.0974409448818898, + "grad_norm": 11.639845848083496, + "learning_rate": 9.935119800817215e-07, + "loss": 0.7128, + "step": 1116 + }, + { + "epoch": 1.0984251968503937, + "grad_norm": 11.503822326660156, + "learning_rate": 9.933880855739446e-07, + "loss": 0.579, + "step": 1117 + }, + { + "epoch": 1.0994094488188977, + "grad_norm": 8.639110565185547, + "learning_rate": 9.932630271885662e-07, + "loss": 0.4919, + "step": 1118 + }, + { + "epoch": 1.1003937007874016, + "grad_norm": 11.358433723449707, + "learning_rate": 9.93136805221593e-07, + "loss": 0.8719, + "step": 1119 + }, + { + "epoch": 1.1013779527559056, + "grad_norm": 12.555047988891602, + "learning_rate": 9.930094199717848e-07, + "loss": 1.1696, + "step": 1120 + }, + { + "epoch": 1.1023622047244095, + "grad_norm": 16.483301162719727, + "learning_rate": 9.92880871740656e-07, + "loss": 1.0572, + "step": 1121 + }, + { + "epoch": 1.1033464566929134, + "grad_norm": 10.812005043029785, + "learning_rate": 9.927511608324732e-07, + "loss": 0.746, + "step": 1122 + }, + { + "epoch": 1.1043307086614174, + "grad_norm": 7.33451509475708, + "learning_rate": 9.926202875542546e-07, + "loss": 0.4024, + "step": 1123 + }, + { + "epoch": 1.1053149606299213, + "grad_norm": 22.47677230834961, + "learning_rate": 9.924882522157706e-07, + "loss": 1.506, + "step": 1124 + }, + { + "epoch": 1.1062992125984252, + "grad_norm": 12.610832214355469, + "learning_rate": 9.923550551295411e-07, + "loss": 0.8658, + "step": 1125 + }, + { + "epoch": 1.1072834645669292, + "grad_norm": 11.658797264099121, + "learning_rate": 9.922206966108367e-07, + "loss": 0.8811, + "step": 1126 + }, + { + "epoch": 1.108267716535433, + "grad_norm": 10.466711044311523, + "learning_rate": 9.920851769776761e-07, + "loss": 0.5166, + "step": 1127 + }, + { + "epoch": 1.109251968503937, + "grad_norm": 12.000123023986816, + "learning_rate": 9.919484965508276e-07, + "loss": 0.6676, + "step": 1128 + }, + { + "epoch": 1.110236220472441, + "grad_norm": 10.173340797424316, + "learning_rate": 9.918106556538057e-07, + "loss": 0.6104, + "step": 1129 + }, + { + "epoch": 1.111220472440945, + "grad_norm": 12.212109565734863, + "learning_rate": 9.91671654612872e-07, + "loss": 0.9263, + "step": 1130 + }, + { + "epoch": 1.1122047244094488, + "grad_norm": 15.39807415008545, + "learning_rate": 9.91531493757035e-07, + "loss": 1.27, + "step": 1131 + }, + { + "epoch": 1.1131889763779528, + "grad_norm": 11.779406547546387, + "learning_rate": 9.913901734180476e-07, + "loss": 0.7009, + "step": 1132 + }, + { + "epoch": 1.1141732283464567, + "grad_norm": 22.418588638305664, + "learning_rate": 9.912476939304069e-07, + "loss": 1.7367, + "step": 1133 + }, + { + "epoch": 1.1151574803149606, + "grad_norm": 9.897189140319824, + "learning_rate": 9.911040556313543e-07, + "loss": 0.5924, + "step": 1134 + }, + { + "epoch": 1.1161417322834646, + "grad_norm": 25.989843368530273, + "learning_rate": 9.909592588608738e-07, + "loss": 1.4601, + "step": 1135 + }, + { + "epoch": 1.1171259842519685, + "grad_norm": 24.011293411254883, + "learning_rate": 9.908133039616911e-07, + "loss": 1.4386, + "step": 1136 + }, + { + "epoch": 1.1181102362204725, + "grad_norm": 11.448659896850586, + "learning_rate": 9.90666191279274e-07, + "loss": 0.7663, + "step": 1137 + }, + { + "epoch": 1.1190944881889764, + "grad_norm": 13.132770538330078, + "learning_rate": 9.905179211618296e-07, + "loss": 1.4043, + "step": 1138 + }, + { + "epoch": 1.1200787401574803, + "grad_norm": 12.378631591796875, + "learning_rate": 9.903684939603054e-07, + "loss": 1.0486, + "step": 1139 + }, + { + "epoch": 1.1210629921259843, + "grad_norm": 10.946438789367676, + "learning_rate": 9.902179100283875e-07, + "loss": 0.4541, + "step": 1140 + }, + { + "epoch": 1.1220472440944882, + "grad_norm": 9.304766654968262, + "learning_rate": 9.900661697224995e-07, + "loss": 0.6372, + "step": 1141 + }, + { + "epoch": 1.1230314960629921, + "grad_norm": 9.965361595153809, + "learning_rate": 9.899132734018024e-07, + "loss": 0.6563, + "step": 1142 + }, + { + "epoch": 1.124015748031496, + "grad_norm": 9.754569053649902, + "learning_rate": 9.897592214281936e-07, + "loss": 0.4799, + "step": 1143 + }, + { + "epoch": 1.125, + "grad_norm": 17.194753646850586, + "learning_rate": 9.896040141663054e-07, + "loss": 1.1878, + "step": 1144 + }, + { + "epoch": 1.125984251968504, + "grad_norm": 10.091248512268066, + "learning_rate": 9.894476519835052e-07, + "loss": 0.6507, + "step": 1145 + }, + { + "epoch": 1.1269685039370079, + "grad_norm": 16.77588653564453, + "learning_rate": 9.892901352498932e-07, + "loss": 0.8475, + "step": 1146 + }, + { + "epoch": 1.1279527559055118, + "grad_norm": 12.568476676940918, + "learning_rate": 9.891314643383037e-07, + "loss": 1.107, + "step": 1147 + }, + { + "epoch": 1.1289370078740157, + "grad_norm": 9.016369819641113, + "learning_rate": 9.88971639624301e-07, + "loss": 0.3035, + "step": 1148 + }, + { + "epoch": 1.1299212598425197, + "grad_norm": 9.463494300842285, + "learning_rate": 9.888106614861823e-07, + "loss": 0.6317, + "step": 1149 + }, + { + "epoch": 1.1309055118110236, + "grad_norm": 16.098876953125, + "learning_rate": 9.886485303049734e-07, + "loss": 1.0571, + "step": 1150 + }, + { + "epoch": 1.1318897637795275, + "grad_norm": 11.583295822143555, + "learning_rate": 9.884852464644305e-07, + "loss": 0.6284, + "step": 1151 + }, + { + "epoch": 1.1328740157480315, + "grad_norm": 11.977601051330566, + "learning_rate": 9.883208103510367e-07, + "loss": 0.9719, + "step": 1152 + }, + { + "epoch": 1.1338582677165354, + "grad_norm": 12.4282808303833, + "learning_rate": 9.88155222354004e-07, + "loss": 0.9043, + "step": 1153 + }, + { + "epoch": 1.1348425196850394, + "grad_norm": 11.01272201538086, + "learning_rate": 9.879884828652698e-07, + "loss": 0.6841, + "step": 1154 + }, + { + "epoch": 1.1358267716535433, + "grad_norm": 12.006009101867676, + "learning_rate": 9.878205922794972e-07, + "loss": 0.608, + "step": 1155 + }, + { + "epoch": 1.1368110236220472, + "grad_norm": 9.32274055480957, + "learning_rate": 9.876515509940742e-07, + "loss": 0.5592, + "step": 1156 + }, + { + "epoch": 1.1377952755905512, + "grad_norm": 22.739818572998047, + "learning_rate": 9.874813594091122e-07, + "loss": 1.4329, + "step": 1157 + }, + { + "epoch": 1.138779527559055, + "grad_norm": 26.053585052490234, + "learning_rate": 9.873100179274453e-07, + "loss": 1.6512, + "step": 1158 + }, + { + "epoch": 1.139763779527559, + "grad_norm": 23.698720932006836, + "learning_rate": 9.871375269546292e-07, + "loss": 1.9971, + "step": 1159 + }, + { + "epoch": 1.140748031496063, + "grad_norm": 16.412256240844727, + "learning_rate": 9.869638868989407e-07, + "loss": 0.9536, + "step": 1160 + }, + { + "epoch": 1.141732283464567, + "grad_norm": 14.230514526367188, + "learning_rate": 9.867890981713764e-07, + "loss": 0.7213, + "step": 1161 + }, + { + "epoch": 1.1427165354330708, + "grad_norm": 10.371112823486328, + "learning_rate": 9.866131611856514e-07, + "loss": 0.4714, + "step": 1162 + }, + { + "epoch": 1.1437007874015748, + "grad_norm": 9.045125961303711, + "learning_rate": 9.864360763581994e-07, + "loss": 0.5469, + "step": 1163 + }, + { + "epoch": 1.1446850393700787, + "grad_norm": 13.141439437866211, + "learning_rate": 9.862578441081699e-07, + "loss": 0.7808, + "step": 1164 + }, + { + "epoch": 1.1456692913385826, + "grad_norm": 20.23772430419922, + "learning_rate": 9.860784648574288e-07, + "loss": 2.1225, + "step": 1165 + }, + { + "epoch": 1.1466535433070866, + "grad_norm": 12.796348571777344, + "learning_rate": 9.858979390305573e-07, + "loss": 1.0035, + "step": 1166 + }, + { + "epoch": 1.1476377952755905, + "grad_norm": 11.537859916687012, + "learning_rate": 9.857162670548502e-07, + "loss": 0.8881, + "step": 1167 + }, + { + "epoch": 1.1486220472440944, + "grad_norm": 11.287174224853516, + "learning_rate": 9.855334493603147e-07, + "loss": 1.0899, + "step": 1168 + }, + { + "epoch": 1.1496062992125984, + "grad_norm": 10.592592239379883, + "learning_rate": 9.853494863796705e-07, + "loss": 0.814, + "step": 1169 + }, + { + "epoch": 1.1505905511811023, + "grad_norm": 13.20379638671875, + "learning_rate": 9.851643785483478e-07, + "loss": 1.1564, + "step": 1170 + }, + { + "epoch": 1.1515748031496063, + "grad_norm": 12.289298057556152, + "learning_rate": 9.849781263044867e-07, + "loss": 1.1262, + "step": 1171 + }, + { + "epoch": 1.1525590551181102, + "grad_norm": 10.34423828125, + "learning_rate": 9.847907300889358e-07, + "loss": 0.4826, + "step": 1172 + }, + { + "epoch": 1.1535433070866141, + "grad_norm": 13.520644187927246, + "learning_rate": 9.846021903452519e-07, + "loss": 1.072, + "step": 1173 + }, + { + "epoch": 1.154527559055118, + "grad_norm": 10.882362365722656, + "learning_rate": 9.844125075196983e-07, + "loss": 0.7588, + "step": 1174 + }, + { + "epoch": 1.155511811023622, + "grad_norm": 12.282682418823242, + "learning_rate": 9.842216820612438e-07, + "loss": 0.9597, + "step": 1175 + }, + { + "epoch": 1.156496062992126, + "grad_norm": 14.361637115478516, + "learning_rate": 9.840297144215615e-07, + "loss": 0.9954, + "step": 1176 + }, + { + "epoch": 1.1574803149606299, + "grad_norm": 10.931534767150879, + "learning_rate": 9.838366050550287e-07, + "loss": 0.7169, + "step": 1177 + }, + { + "epoch": 1.1584645669291338, + "grad_norm": 8.973247528076172, + "learning_rate": 9.836423544187245e-07, + "loss": 0.6022, + "step": 1178 + }, + { + "epoch": 1.1594488188976377, + "grad_norm": 13.785079002380371, + "learning_rate": 9.834469629724293e-07, + "loss": 0.9108, + "step": 1179 + }, + { + "epoch": 1.1604330708661417, + "grad_norm": 13.577901840209961, + "learning_rate": 9.832504311786238e-07, + "loss": 0.7757, + "step": 1180 + }, + { + "epoch": 1.1614173228346456, + "grad_norm": 15.905202865600586, + "learning_rate": 9.830527595024885e-07, + "loss": 1.3626, + "step": 1181 + }, + { + "epoch": 1.1624015748031495, + "grad_norm": 10.73210334777832, + "learning_rate": 9.82853948411901e-07, + "loss": 0.7546, + "step": 1182 + }, + { + "epoch": 1.1633858267716535, + "grad_norm": 7.652894020080566, + "learning_rate": 9.826539983774365e-07, + "loss": 0.4991, + "step": 1183 + }, + { + "epoch": 1.1643700787401574, + "grad_norm": 7.244395732879639, + "learning_rate": 9.824529098723657e-07, + "loss": 0.359, + "step": 1184 + }, + { + "epoch": 1.1653543307086613, + "grad_norm": 12.004920959472656, + "learning_rate": 9.822506833726537e-07, + "loss": 0.8804, + "step": 1185 + }, + { + "epoch": 1.1663385826771653, + "grad_norm": 12.034069061279297, + "learning_rate": 9.8204731935696e-07, + "loss": 0.7941, + "step": 1186 + }, + { + "epoch": 1.1673228346456692, + "grad_norm": 22.406587600708008, + "learning_rate": 9.818428183066356e-07, + "loss": 1.6374, + "step": 1187 + }, + { + "epoch": 1.1683070866141732, + "grad_norm": 13.935935020446777, + "learning_rate": 9.816371807057237e-07, + "loss": 0.8655, + "step": 1188 + }, + { + "epoch": 1.169291338582677, + "grad_norm": 12.156387329101562, + "learning_rate": 9.81430407040957e-07, + "loss": 0.7996, + "step": 1189 + }, + { + "epoch": 1.170275590551181, + "grad_norm": 10.461490631103516, + "learning_rate": 9.812224978017573e-07, + "loss": 0.77, + "step": 1190 + }, + { + "epoch": 1.171259842519685, + "grad_norm": 9.46623706817627, + "learning_rate": 9.810134534802346e-07, + "loss": 0.7023, + "step": 1191 + }, + { + "epoch": 1.172244094488189, + "grad_norm": 10.952614784240723, + "learning_rate": 9.80803274571185e-07, + "loss": 0.8983, + "step": 1192 + }, + { + "epoch": 1.1732283464566928, + "grad_norm": 10.143698692321777, + "learning_rate": 9.80591961572091e-07, + "loss": 0.7334, + "step": 1193 + }, + { + "epoch": 1.1742125984251968, + "grad_norm": 10.362936019897461, + "learning_rate": 9.803795149831183e-07, + "loss": 0.588, + "step": 1194 + }, + { + "epoch": 1.1751968503937007, + "grad_norm": 7.492079257965088, + "learning_rate": 9.801659353071165e-07, + "loss": 0.3758, + "step": 1195 + }, + { + "epoch": 1.1761811023622046, + "grad_norm": 11.673321723937988, + "learning_rate": 9.799512230496172e-07, + "loss": 0.8057, + "step": 1196 + }, + { + "epoch": 1.1771653543307086, + "grad_norm": 28.830455780029297, + "learning_rate": 9.79735378718832e-07, + "loss": 1.6137, + "step": 1197 + }, + { + "epoch": 1.1781496062992125, + "grad_norm": 12.126893043518066, + "learning_rate": 9.795184028256534e-07, + "loss": 0.6031, + "step": 1198 + }, + { + "epoch": 1.1791338582677164, + "grad_norm": 13.487756729125977, + "learning_rate": 9.793002958836506e-07, + "loss": 0.9049, + "step": 1199 + }, + { + "epoch": 1.1801181102362204, + "grad_norm": 21.260787963867188, + "learning_rate": 9.790810584090712e-07, + "loss": 0.9688, + "step": 1200 + }, + { + "epoch": 1.1811023622047245, + "grad_norm": 10.346322059631348, + "learning_rate": 9.78860690920838e-07, + "loss": 0.4923, + "step": 1201 + }, + { + "epoch": 1.1820866141732282, + "grad_norm": 12.38548755645752, + "learning_rate": 9.78639193940549e-07, + "loss": 0.6468, + "step": 1202 + }, + { + "epoch": 1.1830708661417324, + "grad_norm": 26.022510528564453, + "learning_rate": 9.784165679924752e-07, + "loss": 1.907, + "step": 1203 + }, + { + "epoch": 1.1840551181102361, + "grad_norm": 12.170500755310059, + "learning_rate": 9.781928136035604e-07, + "loss": 0.81, + "step": 1204 + }, + { + "epoch": 1.1850393700787403, + "grad_norm": 19.44597625732422, + "learning_rate": 9.779679313034182e-07, + "loss": 1.1321, + "step": 1205 + }, + { + "epoch": 1.186023622047244, + "grad_norm": 24.822078704833984, + "learning_rate": 9.777419216243333e-07, + "loss": 1.901, + "step": 1206 + }, + { + "epoch": 1.1870078740157481, + "grad_norm": 12.606575965881348, + "learning_rate": 9.77514785101258e-07, + "loss": 0.9074, + "step": 1207 + }, + { + "epoch": 1.1879921259842519, + "grad_norm": 12.32088851928711, + "learning_rate": 9.77286522271812e-07, + "loss": 0.7982, + "step": 1208 + }, + { + "epoch": 1.188976377952756, + "grad_norm": 11.44482707977295, + "learning_rate": 9.77057133676281e-07, + "loss": 0.6944, + "step": 1209 + }, + { + "epoch": 1.1899606299212597, + "grad_norm": 8.395225524902344, + "learning_rate": 9.768266198576147e-07, + "loss": 0.3568, + "step": 1210 + }, + { + "epoch": 1.1909448818897639, + "grad_norm": 10.126123428344727, + "learning_rate": 9.765949813614273e-07, + "loss": 0.605, + "step": 1211 + }, + { + "epoch": 1.1919291338582676, + "grad_norm": 9.255550384521484, + "learning_rate": 9.763622187359942e-07, + "loss": 0.515, + "step": 1212 + }, + { + "epoch": 1.1929133858267718, + "grad_norm": 18.876310348510742, + "learning_rate": 9.761283325322515e-07, + "loss": 1.2768, + "step": 1213 + }, + { + "epoch": 1.1938976377952755, + "grad_norm": 10.210433959960938, + "learning_rate": 9.758933233037952e-07, + "loss": 0.5625, + "step": 1214 + }, + { + "epoch": 1.1948818897637796, + "grad_norm": 8.423827171325684, + "learning_rate": 9.756571916068793e-07, + "loss": 0.3923, + "step": 1215 + }, + { + "epoch": 1.1958661417322836, + "grad_norm": 11.02546215057373, + "learning_rate": 9.754199380004145e-07, + "loss": 0.5358, + "step": 1216 + }, + { + "epoch": 1.1968503937007875, + "grad_norm": 20.572431564331055, + "learning_rate": 9.75181563045967e-07, + "loss": 0.9994, + "step": 1217 + }, + { + "epoch": 1.1978346456692914, + "grad_norm": 13.444281578063965, + "learning_rate": 9.749420673077574e-07, + "loss": 0.5955, + "step": 1218 + }, + { + "epoch": 1.1988188976377954, + "grad_norm": 11.807435035705566, + "learning_rate": 9.747014513526588e-07, + "loss": 0.6657, + "step": 1219 + }, + { + "epoch": 1.1998031496062993, + "grad_norm": 11.162291526794434, + "learning_rate": 9.744597157501957e-07, + "loss": 0.7751, + "step": 1220 + }, + { + "epoch": 1.2007874015748032, + "grad_norm": 12.134571075439453, + "learning_rate": 9.742168610725437e-07, + "loss": 0.8443, + "step": 1221 + }, + { + "epoch": 1.2017716535433072, + "grad_norm": 12.475569725036621, + "learning_rate": 9.73972887894526e-07, + "loss": 0.6662, + "step": 1222 + }, + { + "epoch": 1.202755905511811, + "grad_norm": 10.001572608947754, + "learning_rate": 9.737277967936137e-07, + "loss": 0.7022, + "step": 1223 + }, + { + "epoch": 1.203740157480315, + "grad_norm": 13.214725494384766, + "learning_rate": 9.734815883499241e-07, + "loss": 0.8183, + "step": 1224 + }, + { + "epoch": 1.203740157480315, + "eval_Qnli-dev_cosine_accuracy": 0.681640625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8366125226020813, + "eval_Qnli-dev_cosine_ap": 0.6914476487679233, + "eval_Qnli-dev_cosine_f1": 0.6784, + "eval_Qnli-dev_cosine_f1_threshold": 0.7529718279838562, + "eval_Qnli-dev_cosine_precision": 0.5449871465295629, + "eval_Qnli-dev_cosine_recall": 0.8983050847457628, + "eval_Qnli-dev_dot_accuracy": 0.67578125, + "eval_Qnli-dev_dot_accuracy_threshold": 639.7266845703125, + "eval_Qnli-dev_dot_ap": 0.6893503306461788, + "eval_Qnli-dev_dot_f1": 0.6772655007949125, + "eval_Qnli-dev_dot_f1_threshold": 571.5009155273438, + "eval_Qnli-dev_dot_precision": 0.5419847328244275, + "eval_Qnli-dev_dot_recall": 0.902542372881356, + "eval_Qnli-dev_euclidean_accuracy": 0.681640625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.806737899780273, + "eval_Qnli-dev_euclidean_ap": 0.6915811548453872, + "eval_Qnli-dev_euclidean_f1": 0.6784, + "eval_Qnli-dev_euclidean_f1_threshold": 19.451509475708008, + "eval_Qnli-dev_euclidean_precision": 0.5449871465295629, + "eval_Qnli-dev_euclidean_recall": 0.8983050847457628, + "eval_Qnli-dev_manhattan_accuracy": 0.68359375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 344.76611328125, + "eval_Qnli-dev_manhattan_ap": 0.6984341364216773, + "eval_Qnli-dev_manhattan_f1": 0.6807817589576547, + "eval_Qnli-dev_manhattan_f1_threshold": 400.9693603515625, + "eval_Qnli-dev_manhattan_precision": 0.5529100529100529, + "eval_Qnli-dev_manhattan_recall": 0.885593220338983, + "eval_Qnli-dev_max_accuracy": 0.68359375, + "eval_Qnli-dev_max_accuracy_threshold": 639.7266845703125, + "eval_Qnli-dev_max_ap": 0.6984341364216773, + "eval_Qnli-dev_max_f1": 0.6807817589576547, + "eval_Qnli-dev_max_f1_threshold": 571.5009155273438, + "eval_Qnli-dev_max_precision": 0.5529100529100529, + "eval_Qnli-dev_max_recall": 0.902542372881356, + "eval_allNLI-dev_cosine_accuracy": 0.68359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9374961853027344, + "eval_allNLI-dev_cosine_ap": 0.4873775690193095, + "eval_allNLI-dev_cosine_f1": 0.533066132264529, + "eval_allNLI-dev_cosine_f1_threshold": 0.8453455567359924, + "eval_allNLI-dev_cosine_precision": 0.40797546012269936, + "eval_allNLI-dev_cosine_recall": 0.7687861271676301, + "eval_allNLI-dev_dot_accuracy": 0.68359375, + "eval_allNLI-dev_dot_accuracy_threshold": 717.646484375, + "eval_allNLI-dev_dot_ap": 0.48976942081457003, + "eval_allNLI-dev_dot_f1": 0.5338645418326693, + "eval_allNLI-dev_dot_f1_threshold": 646.2739868164062, + "eval_allNLI-dev_dot_precision": 0.4072948328267477, + "eval_allNLI-dev_dot_recall": 0.7745664739884393, + "eval_allNLI-dev_euclidean_accuracy": 0.68359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.785663604736328, + "eval_allNLI-dev_euclidean_ap": 0.4873970729071831, + "eval_allNLI-dev_euclidean_f1": 0.5319999999999999, + "eval_allNLI-dev_euclidean_f1_threshold": 15.387359619140625, + "eval_allNLI-dev_euclidean_precision": 0.40672782874617736, + "eval_allNLI-dev_euclidean_recall": 0.7687861271676301, + "eval_allNLI-dev_manhattan_accuracy": 0.681640625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 205.621337890625, + "eval_allNLI-dev_manhattan_ap": 0.48655776096738934, + "eval_allNLI-dev_manhattan_f1": 0.5379876796714579, + "eval_allNLI-dev_manhattan_f1_threshold": 313.74407958984375, + "eval_allNLI-dev_manhattan_precision": 0.4171974522292994, + "eval_allNLI-dev_manhattan_recall": 0.7572254335260116, + "eval_allNLI-dev_max_accuracy": 0.68359375, + "eval_allNLI-dev_max_accuracy_threshold": 717.646484375, + "eval_allNLI-dev_max_ap": 0.48976942081457003, + "eval_allNLI-dev_max_f1": 0.5379876796714579, + "eval_allNLI-dev_max_f1_threshold": 646.2739868164062, + "eval_allNLI-dev_max_precision": 0.4171974522292994, + "eval_allNLI-dev_max_recall": 0.7745664739884393, + "eval_loss": 1.1757967472076416, + "eval_runtime": 54.4087, + "eval_samples_per_second": 30.583, + "eval_sequential_score": 0.6984341364216773, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.5452067506531562, + "eval_sts-test_pearson_dot": 0.5408381648828177, + "eval_sts-test_pearson_euclidean": 0.5684167455418432, + "eval_sts-test_pearson_manhattan": 0.5766689893476588, + "eval_sts-test_pearson_max": 0.5766689893476588, + "eval_sts-test_spearman_cosine": 0.5782599130018705, + "eval_sts-test_spearman_dot": 0.5738021503521504, + "eval_sts-test_spearman_euclidean": 0.5785303001231974, + "eval_sts-test_spearman_manhattan": 0.5863883978672202, + "eval_sts-test_spearman_max": 0.5863883978672202, + "step": 1224 + }, + { + "epoch": 1.204724409448819, + "grad_norm": 27.127208709716797, + "learning_rate": 9.73234263146219e-07, + "loss": 1.7184, + "step": 1225 + }, + { + "epoch": 1.205708661417323, + "grad_norm": 27.30764389038086, + "learning_rate": 9.72985821767904e-07, + "loss": 2.0612, + "step": 1226 + }, + { + "epoch": 1.2066929133858268, + "grad_norm": 12.28317928314209, + "learning_rate": 9.727362648030254e-07, + "loss": 0.5645, + "step": 1227 + }, + { + "epoch": 1.2076771653543308, + "grad_norm": 11.45736312866211, + "learning_rate": 9.724855928422714e-07, + "loss": 1.0122, + "step": 1228 + }, + { + "epoch": 1.2086614173228347, + "grad_norm": 7.623450756072998, + "learning_rate": 9.722338064789688e-07, + "loss": 0.4022, + "step": 1229 + }, + { + "epoch": 1.2096456692913387, + "grad_norm": 18.705665588378906, + "learning_rate": 9.719809063090818e-07, + "loss": 1.2996, + "step": 1230 + }, + { + "epoch": 1.2106299212598426, + "grad_norm": 10.509492874145508, + "learning_rate": 9.717268929312112e-07, + "loss": 0.5619, + "step": 1231 + }, + { + "epoch": 1.2116141732283465, + "grad_norm": 7.465246200561523, + "learning_rate": 9.714717669465932e-07, + "loss": 0.2787, + "step": 1232 + }, + { + "epoch": 1.2125984251968505, + "grad_norm": 10.551802635192871, + "learning_rate": 9.712155289590964e-07, + "loss": 0.423, + "step": 1233 + }, + { + "epoch": 1.2135826771653544, + "grad_norm": 14.872739791870117, + "learning_rate": 9.709581795752225e-07, + "loss": 0.7389, + "step": 1234 + }, + { + "epoch": 1.2145669291338583, + "grad_norm": 14.757356643676758, + "learning_rate": 9.70699719404103e-07, + "loss": 1.0838, + "step": 1235 + }, + { + "epoch": 1.2155511811023623, + "grad_norm": 10.091594696044922, + "learning_rate": 9.704401490574993e-07, + "loss": 0.5171, + "step": 1236 + }, + { + "epoch": 1.2165354330708662, + "grad_norm": 11.12486743927002, + "learning_rate": 9.701794691497997e-07, + "loss": 0.5735, + "step": 1237 + }, + { + "epoch": 1.2175196850393701, + "grad_norm": 10.067895889282227, + "learning_rate": 9.699176802980199e-07, + "loss": 0.716, + "step": 1238 + }, + { + "epoch": 1.218503937007874, + "grad_norm": 14.799269676208496, + "learning_rate": 9.69654783121799e-07, + "loss": 1.0944, + "step": 1239 + }, + { + "epoch": 1.219488188976378, + "grad_norm": 8.924025535583496, + "learning_rate": 9.693907782434008e-07, + "loss": 0.409, + "step": 1240 + }, + { + "epoch": 1.220472440944882, + "grad_norm": 23.94953155517578, + "learning_rate": 9.691256662877099e-07, + "loss": 1.5226, + "step": 1241 + }, + { + "epoch": 1.2214566929133859, + "grad_norm": 11.415987014770508, + "learning_rate": 9.688594478822317e-07, + "loss": 0.6814, + "step": 1242 + }, + { + "epoch": 1.2224409448818898, + "grad_norm": 13.597917556762695, + "learning_rate": 9.685921236570914e-07, + "loss": 0.9776, + "step": 1243 + }, + { + "epoch": 1.2234251968503937, + "grad_norm": 12.646177291870117, + "learning_rate": 9.683236942450294e-07, + "loss": 0.6928, + "step": 1244 + }, + { + "epoch": 1.2244094488188977, + "grad_norm": 24.075674057006836, + "learning_rate": 9.680541602814047e-07, + "loss": 1.1014, + "step": 1245 + }, + { + "epoch": 1.2253937007874016, + "grad_norm": 13.862969398498535, + "learning_rate": 9.677835224041885e-07, + "loss": 1.0794, + "step": 1246 + }, + { + "epoch": 1.2263779527559056, + "grad_norm": 10.0800199508667, + "learning_rate": 9.67511781253966e-07, + "loss": 0.4316, + "step": 1247 + }, + { + "epoch": 1.2273622047244095, + "grad_norm": 16.01119613647461, + "learning_rate": 9.67238937473934e-07, + "loss": 1.0016, + "step": 1248 + }, + { + "epoch": 1.2283464566929134, + "grad_norm": 9.573379516601562, + "learning_rate": 9.669649917098983e-07, + "loss": 0.4825, + "step": 1249 + }, + { + "epoch": 1.2293307086614174, + "grad_norm": 19.94251823425293, + "learning_rate": 9.666899446102736e-07, + "loss": 1.2514, + "step": 1250 + }, + { + "epoch": 1.2303149606299213, + "grad_norm": 9.705304145812988, + "learning_rate": 9.664137968260814e-07, + "loss": 0.462, + "step": 1251 + }, + { + "epoch": 1.2312992125984252, + "grad_norm": 13.630583763122559, + "learning_rate": 9.661365490109482e-07, + "loss": 0.9723, + "step": 1252 + }, + { + "epoch": 1.2322834645669292, + "grad_norm": 10.716950416564941, + "learning_rate": 9.658582018211045e-07, + "loss": 0.6441, + "step": 1253 + }, + { + "epoch": 1.233267716535433, + "grad_norm": 6.912999153137207, + "learning_rate": 9.655787559153824e-07, + "loss": 0.3009, + "step": 1254 + }, + { + "epoch": 1.234251968503937, + "grad_norm": 15.139958381652832, + "learning_rate": 9.652982119552154e-07, + "loss": 1.0618, + "step": 1255 + }, + { + "epoch": 1.235236220472441, + "grad_norm": 29.90904998779297, + "learning_rate": 9.650165706046359e-07, + "loss": 1.9636, + "step": 1256 + }, + { + "epoch": 1.236220472440945, + "grad_norm": 22.840852737426758, + "learning_rate": 9.64733832530273e-07, + "loss": 1.7589, + "step": 1257 + }, + { + "epoch": 1.2372047244094488, + "grad_norm": 12.522883415222168, + "learning_rate": 9.644499984013525e-07, + "loss": 0.6347, + "step": 1258 + }, + { + "epoch": 1.2381889763779528, + "grad_norm": 14.57872200012207, + "learning_rate": 9.64165068889694e-07, + "loss": 0.6183, + "step": 1259 + }, + { + "epoch": 1.2391732283464567, + "grad_norm": 11.782417297363281, + "learning_rate": 9.6387904466971e-07, + "loss": 0.6868, + "step": 1260 + }, + { + "epoch": 1.2401574803149606, + "grad_norm": 10.705400466918945, + "learning_rate": 9.635919264184043e-07, + "loss": 0.6504, + "step": 1261 + }, + { + "epoch": 1.2411417322834646, + "grad_norm": 9.323182106018066, + "learning_rate": 9.6330371481537e-07, + "loss": 0.4346, + "step": 1262 + }, + { + "epoch": 1.2421259842519685, + "grad_norm": 11.712157249450684, + "learning_rate": 9.630144105427883e-07, + "loss": 0.6985, + "step": 1263 + }, + { + "epoch": 1.2431102362204725, + "grad_norm": 11.031139373779297, + "learning_rate": 9.627240142854262e-07, + "loss": 0.5558, + "step": 1264 + }, + { + "epoch": 1.2440944881889764, + "grad_norm": 15.171723365783691, + "learning_rate": 9.624325267306357e-07, + "loss": 0.9465, + "step": 1265 + }, + { + "epoch": 1.2450787401574803, + "grad_norm": 16.123682022094727, + "learning_rate": 9.621399485683523e-07, + "loss": 1.0173, + "step": 1266 + }, + { + "epoch": 1.2460629921259843, + "grad_norm": 11.502947807312012, + "learning_rate": 9.618462804910917e-07, + "loss": 0.6847, + "step": 1267 + }, + { + "epoch": 1.2470472440944882, + "grad_norm": 13.352543830871582, + "learning_rate": 9.61551523193951e-07, + "loss": 1.0888, + "step": 1268 + }, + { + "epoch": 1.2480314960629921, + "grad_norm": 26.212068557739258, + "learning_rate": 9.612556773746038e-07, + "loss": 1.5393, + "step": 1269 + }, + { + "epoch": 1.249015748031496, + "grad_norm": 7.848837852478027, + "learning_rate": 9.609587437333011e-07, + "loss": 0.466, + "step": 1270 + }, + { + "epoch": 1.25, + "grad_norm": 9.938773155212402, + "learning_rate": 9.606607229728688e-07, + "loss": 0.4438, + "step": 1271 + }, + { + "epoch": 1.250984251968504, + "grad_norm": 11.60131549835205, + "learning_rate": 9.603616157987054e-07, + "loss": 0.6365, + "step": 1272 + }, + { + "epoch": 1.2519685039370079, + "grad_norm": 20.468914031982422, + "learning_rate": 9.600614229187812e-07, + "loss": 1.244, + "step": 1273 + }, + { + "epoch": 1.2529527559055118, + "grad_norm": 14.907753944396973, + "learning_rate": 9.597601450436364e-07, + "loss": 1.0352, + "step": 1274 + }, + { + "epoch": 1.2539370078740157, + "grad_norm": 13.460160255432129, + "learning_rate": 9.594577828863794e-07, + "loss": 0.8419, + "step": 1275 + }, + { + "epoch": 1.2549212598425197, + "grad_norm": 22.66904067993164, + "learning_rate": 9.591543371626842e-07, + "loss": 1.6384, + "step": 1276 + }, + { + "epoch": 1.2559055118110236, + "grad_norm": 19.56178855895996, + "learning_rate": 9.588498085907905e-07, + "loss": 1.3742, + "step": 1277 + }, + { + "epoch": 1.2568897637795275, + "grad_norm": 7.275899410247803, + "learning_rate": 9.585441978915008e-07, + "loss": 0.2408, + "step": 1278 + }, + { + "epoch": 1.2578740157480315, + "grad_norm": 12.173895835876465, + "learning_rate": 9.582375057881785e-07, + "loss": 0.4633, + "step": 1279 + }, + { + "epoch": 1.2588582677165354, + "grad_norm": 12.677711486816406, + "learning_rate": 9.579297330067476e-07, + "loss": 0.9162, + "step": 1280 + }, + { + "epoch": 1.2598425196850394, + "grad_norm": 17.700607299804688, + "learning_rate": 9.576208802756888e-07, + "loss": 1.0861, + "step": 1281 + }, + { + "epoch": 1.2608267716535433, + "grad_norm": 10.266142845153809, + "learning_rate": 9.5731094832604e-07, + "loss": 0.5915, + "step": 1282 + }, + { + "epoch": 1.2618110236220472, + "grad_norm": 20.697246551513672, + "learning_rate": 9.569999378913925e-07, + "loss": 0.9804, + "step": 1283 + }, + { + "epoch": 1.2627952755905512, + "grad_norm": 12.583255767822266, + "learning_rate": 9.566878497078914e-07, + "loss": 0.7697, + "step": 1284 + }, + { + "epoch": 1.263779527559055, + "grad_norm": 12.262706756591797, + "learning_rate": 9.563746845142326e-07, + "loss": 0.6647, + "step": 1285 + }, + { + "epoch": 1.264763779527559, + "grad_norm": 10.353370666503906, + "learning_rate": 9.5606044305166e-07, + "loss": 0.5544, + "step": 1286 + }, + { + "epoch": 1.265748031496063, + "grad_norm": 9.361234664916992, + "learning_rate": 9.557451260639665e-07, + "loss": 0.6577, + "step": 1287 + }, + { + "epoch": 1.266732283464567, + "grad_norm": 9.571710586547852, + "learning_rate": 9.5542873429749e-07, + "loss": 0.4959, + "step": 1288 + }, + { + "epoch": 1.2677165354330708, + "grad_norm": 18.31670570373535, + "learning_rate": 9.551112685011127e-07, + "loss": 0.8166, + "step": 1289 + }, + { + "epoch": 1.2687007874015748, + "grad_norm": 10.438338279724121, + "learning_rate": 9.547927294262583e-07, + "loss": 0.7087, + "step": 1290 + }, + { + "epoch": 1.2696850393700787, + "grad_norm": 9.65834903717041, + "learning_rate": 9.544731178268913e-07, + "loss": 0.5905, + "step": 1291 + }, + { + "epoch": 1.2706692913385826, + "grad_norm": 23.963003158569336, + "learning_rate": 9.54152434459515e-07, + "loss": 1.406, + "step": 1292 + }, + { + "epoch": 1.2716535433070866, + "grad_norm": 23.614809036254883, + "learning_rate": 9.538306800831695e-07, + "loss": 1.1811, + "step": 1293 + }, + { + "epoch": 1.2726377952755905, + "grad_norm": 10.113353729248047, + "learning_rate": 9.535078554594293e-07, + "loss": 0.6425, + "step": 1294 + }, + { + "epoch": 1.2736220472440944, + "grad_norm": 16.463342666625977, + "learning_rate": 9.531839613524029e-07, + "loss": 0.9818, + "step": 1295 + }, + { + "epoch": 1.2746062992125984, + "grad_norm": 13.193273544311523, + "learning_rate": 9.528589985287297e-07, + "loss": 0.8521, + "step": 1296 + }, + { + "epoch": 1.2755905511811023, + "grad_norm": 11.638998031616211, + "learning_rate": 9.525329677575787e-07, + "loss": 0.6998, + "step": 1297 + }, + { + "epoch": 1.2765748031496063, + "grad_norm": 15.33539867401123, + "learning_rate": 9.522058698106475e-07, + "loss": 0.8087, + "step": 1298 + }, + { + "epoch": 1.2775590551181102, + "grad_norm": 11.569411277770996, + "learning_rate": 9.518777054621582e-07, + "loss": 0.6473, + "step": 1299 + }, + { + "epoch": 1.2785433070866141, + "grad_norm": 13.36752700805664, + "learning_rate": 9.515484754888582e-07, + "loss": 1.2283, + "step": 1300 + }, + { + "epoch": 1.279527559055118, + "grad_norm": 11.46446418762207, + "learning_rate": 9.512181806700167e-07, + "loss": 0.7829, + "step": 1301 + }, + { + "epoch": 1.280511811023622, + "grad_norm": 12.435479164123535, + "learning_rate": 9.508868217874233e-07, + "loss": 0.7921, + "step": 1302 + }, + { + "epoch": 1.281496062992126, + "grad_norm": 8.405648231506348, + "learning_rate": 9.505543996253865e-07, + "loss": 0.3911, + "step": 1303 + }, + { + "epoch": 1.2824803149606299, + "grad_norm": 9.165940284729004, + "learning_rate": 9.502209149707312e-07, + "loss": 0.4446, + "step": 1304 + }, + { + "epoch": 1.2834645669291338, + "grad_norm": 9.730753898620605, + "learning_rate": 9.498863686127971e-07, + "loss": 0.53, + "step": 1305 + }, + { + "epoch": 1.2844488188976377, + "grad_norm": 13.173008918762207, + "learning_rate": 9.495507613434374e-07, + "loss": 0.9651, + "step": 1306 + }, + { + "epoch": 1.2854330708661417, + "grad_norm": 11.323593139648438, + "learning_rate": 9.492140939570159e-07, + "loss": 0.8219, + "step": 1307 + }, + { + "epoch": 1.2864173228346456, + "grad_norm": 12.519793510437012, + "learning_rate": 9.488763672504058e-07, + "loss": 0.7061, + "step": 1308 + }, + { + "epoch": 1.2874015748031495, + "grad_norm": 8.543204307556152, + "learning_rate": 9.485375820229876e-07, + "loss": 0.493, + "step": 1309 + }, + { + "epoch": 1.2883858267716535, + "grad_norm": 17.205244064331055, + "learning_rate": 9.481977390766475e-07, + "loss": 1.315, + "step": 1310 + }, + { + "epoch": 1.2893700787401574, + "grad_norm": 18.341398239135742, + "learning_rate": 9.47856839215775e-07, + "loss": 1.4424, + "step": 1311 + }, + { + "epoch": 1.2903543307086613, + "grad_norm": 13.18166446685791, + "learning_rate": 9.475148832472613e-07, + "loss": 0.6492, + "step": 1312 + }, + { + "epoch": 1.2913385826771653, + "grad_norm": 20.222919464111328, + "learning_rate": 9.471718719804975e-07, + "loss": 1.3142, + "step": 1313 + }, + { + "epoch": 1.2923228346456692, + "grad_norm": 13.445477485656738, + "learning_rate": 9.468278062273726e-07, + "loss": 0.8536, + "step": 1314 + }, + { + "epoch": 1.2933070866141732, + "grad_norm": 11.16597843170166, + "learning_rate": 9.464826868022709e-07, + "loss": 0.6758, + "step": 1315 + }, + { + "epoch": 1.294291338582677, + "grad_norm": 16.122098922729492, + "learning_rate": 9.461365145220715e-07, + "loss": 1.0877, + "step": 1316 + }, + { + "epoch": 1.295275590551181, + "grad_norm": 19.477783203125, + "learning_rate": 9.457892902061452e-07, + "loss": 1.2856, + "step": 1317 + }, + { + "epoch": 1.296259842519685, + "grad_norm": 16.64370346069336, + "learning_rate": 9.454410146763525e-07, + "loss": 1.1316, + "step": 1318 + }, + { + "epoch": 1.297244094488189, + "grad_norm": 16.68716049194336, + "learning_rate": 9.450916887570429e-07, + "loss": 1.137, + "step": 1319 + }, + { + "epoch": 1.298228346456693, + "grad_norm": 10.482457160949707, + "learning_rate": 9.447413132750515e-07, + "loss": 0.6438, + "step": 1320 + }, + { + "epoch": 1.2992125984251968, + "grad_norm": 10.563558578491211, + "learning_rate": 9.443898890596977e-07, + "loss": 0.5434, + "step": 1321 + }, + { + "epoch": 1.300196850393701, + "grad_norm": 20.146427154541016, + "learning_rate": 9.440374169427834e-07, + "loss": 1.117, + "step": 1322 + }, + { + "epoch": 1.3011811023622046, + "grad_norm": 17.821557998657227, + "learning_rate": 9.43683897758591e-07, + "loss": 1.0887, + "step": 1323 + }, + { + "epoch": 1.3021653543307088, + "grad_norm": 17.06334686279297, + "learning_rate": 9.433293323438807e-07, + "loss": 0.9753, + "step": 1324 + }, + { + "epoch": 1.3031496062992125, + "grad_norm": 14.24149227142334, + "learning_rate": 9.429737215378894e-07, + "loss": 0.7109, + "step": 1325 + }, + { + "epoch": 1.3041338582677167, + "grad_norm": 16.65839958190918, + "learning_rate": 9.426170661823285e-07, + "loss": 1.1552, + "step": 1326 + }, + { + "epoch": 1.3051181102362204, + "grad_norm": 13.389101028442383, + "learning_rate": 9.422593671213816e-07, + "loss": 0.8534, + "step": 1327 + }, + { + "epoch": 1.3061023622047245, + "grad_norm": 10.902589797973633, + "learning_rate": 9.419006252017027e-07, + "loss": 0.7198, + "step": 1328 + }, + { + "epoch": 1.3070866141732282, + "grad_norm": 13.474040985107422, + "learning_rate": 9.415408412724141e-07, + "loss": 1.2167, + "step": 1329 + }, + { + "epoch": 1.3080708661417324, + "grad_norm": 11.535694122314453, + "learning_rate": 9.411800161851048e-07, + "loss": 0.979, + "step": 1330 + }, + { + "epoch": 1.3090551181102361, + "grad_norm": 15.960291862487793, + "learning_rate": 9.408181507938279e-07, + "loss": 1.1667, + "step": 1331 + }, + { + "epoch": 1.3100393700787403, + "grad_norm": 10.989014625549316, + "learning_rate": 9.404552459550988e-07, + "loss": 0.5707, + "step": 1332 + }, + { + "epoch": 1.311023622047244, + "grad_norm": 12.71398639678955, + "learning_rate": 9.400913025278937e-07, + "loss": 0.8196, + "step": 1333 + }, + { + "epoch": 1.3120078740157481, + "grad_norm": 15.065571784973145, + "learning_rate": 9.397263213736464e-07, + "loss": 1.1255, + "step": 1334 + }, + { + "epoch": 1.3129921259842519, + "grad_norm": 9.93442440032959, + "learning_rate": 9.393603033562471e-07, + "loss": 0.7614, + "step": 1335 + }, + { + "epoch": 1.313976377952756, + "grad_norm": 10.066417694091797, + "learning_rate": 9.389932493420407e-07, + "loss": 0.616, + "step": 1336 + }, + { + "epoch": 1.3149606299212597, + "grad_norm": 11.152246475219727, + "learning_rate": 9.386251601998239e-07, + "loss": 0.8282, + "step": 1337 + }, + { + "epoch": 1.3159448818897639, + "grad_norm": 13.376919746398926, + "learning_rate": 9.38256036800843e-07, + "loss": 0.9118, + "step": 1338 + }, + { + "epoch": 1.3169291338582676, + "grad_norm": 8.617725372314453, + "learning_rate": 9.378858800187935e-07, + "loss": 0.578, + "step": 1339 + }, + { + "epoch": 1.3179133858267718, + "grad_norm": 9.212502479553223, + "learning_rate": 9.375146907298157e-07, + "loss": 0.5914, + "step": 1340 + }, + { + "epoch": 1.3188976377952755, + "grad_norm": 9.490756034851074, + "learning_rate": 9.371424698124945e-07, + "loss": 0.612, + "step": 1341 + }, + { + "epoch": 1.3198818897637796, + "grad_norm": 10.681628227233887, + "learning_rate": 9.367692181478564e-07, + "loss": 0.5543, + "step": 1342 + }, + { + "epoch": 1.3208661417322833, + "grad_norm": 9.452579498291016, + "learning_rate": 9.363949366193673e-07, + "loss": 0.4898, + "step": 1343 + }, + { + "epoch": 1.3218503937007875, + "grad_norm": 11.30627727508545, + "learning_rate": 9.360196261129313e-07, + "loss": 0.7363, + "step": 1344 + }, + { + "epoch": 1.3228346456692912, + "grad_norm": 8.573918342590332, + "learning_rate": 9.356432875168876e-07, + "loss": 0.3767, + "step": 1345 + }, + { + "epoch": 1.3238188976377954, + "grad_norm": 13.486555099487305, + "learning_rate": 9.352659217220091e-07, + "loss": 0.4314, + "step": 1346 + }, + { + "epoch": 1.324803149606299, + "grad_norm": 10.113042831420898, + "learning_rate": 9.348875296215001e-07, + "loss": 0.4874, + "step": 1347 + }, + { + "epoch": 1.3257874015748032, + "grad_norm": 9.970690727233887, + "learning_rate": 9.345081121109935e-07, + "loss": 0.467, + "step": 1348 + }, + { + "epoch": 1.326771653543307, + "grad_norm": 10.185608863830566, + "learning_rate": 9.341276700885501e-07, + "loss": 0.4353, + "step": 1349 + }, + { + "epoch": 1.327755905511811, + "grad_norm": 9.48005199432373, + "learning_rate": 9.337462044546548e-07, + "loss": 0.6328, + "step": 1350 + }, + { + "epoch": 1.328740157480315, + "grad_norm": 19.29398536682129, + "learning_rate": 9.333637161122166e-07, + "loss": 0.9637, + "step": 1351 + }, + { + "epoch": 1.329724409448819, + "grad_norm": 12.811509132385254, + "learning_rate": 9.329802059665637e-07, + "loss": 0.9188, + "step": 1352 + }, + { + "epoch": 1.330708661417323, + "grad_norm": 11.238883018493652, + "learning_rate": 9.325956749254437e-07, + "loss": 0.5365, + "step": 1353 + }, + { + "epoch": 1.3316929133858268, + "grad_norm": 9.164770126342773, + "learning_rate": 9.322101238990206e-07, + "loss": 0.2543, + "step": 1354 + }, + { + "epoch": 1.3326771653543308, + "grad_norm": 13.3213472366333, + "learning_rate": 9.318235537998726e-07, + "loss": 0.8462, + "step": 1355 + }, + { + "epoch": 1.3336614173228347, + "grad_norm": 12.852092742919922, + "learning_rate": 9.314359655429897e-07, + "loss": 0.824, + "step": 1356 + }, + { + "epoch": 1.3346456692913387, + "grad_norm": 10.420133590698242, + "learning_rate": 9.310473600457721e-07, + "loss": 0.6009, + "step": 1357 + }, + { + "epoch": 1.3356299212598426, + "grad_norm": 13.898282051086426, + "learning_rate": 9.306577382280277e-07, + "loss": 0.8251, + "step": 1358 + }, + { + "epoch": 1.3366141732283465, + "grad_norm": 8.93106746673584, + "learning_rate": 9.3026710101197e-07, + "loss": 0.5137, + "step": 1359 + }, + { + "epoch": 1.3375984251968505, + "grad_norm": 20.12264633178711, + "learning_rate": 9.298754493222155e-07, + "loss": 1.1384, + "step": 1360 + }, + { + "epoch": 1.3385826771653544, + "grad_norm": 6.826462268829346, + "learning_rate": 9.294827840857827e-07, + "loss": 0.3927, + "step": 1361 + }, + { + "epoch": 1.3395669291338583, + "grad_norm": 21.808618545532227, + "learning_rate": 9.290891062320884e-07, + "loss": 1.2509, + "step": 1362 + }, + { + "epoch": 1.3405511811023623, + "grad_norm": 17.233945846557617, + "learning_rate": 9.286944166929466e-07, + "loss": 1.0992, + "step": 1363 + }, + { + "epoch": 1.3415354330708662, + "grad_norm": 16.560678482055664, + "learning_rate": 9.282987164025658e-07, + "loss": 1.0847, + "step": 1364 + }, + { + "epoch": 1.3425196850393701, + "grad_norm": 13.578652381896973, + "learning_rate": 9.279020062975465e-07, + "loss": 1.1022, + "step": 1365 + }, + { + "epoch": 1.343503937007874, + "grad_norm": 16.567989349365234, + "learning_rate": 9.275042873168802e-07, + "loss": 0.8804, + "step": 1366 + }, + { + "epoch": 1.344488188976378, + "grad_norm": 13.353636741638184, + "learning_rate": 9.271055604019458e-07, + "loss": 0.8076, + "step": 1367 + }, + { + "epoch": 1.345472440944882, + "grad_norm": 11.190862655639648, + "learning_rate": 9.267058264965076e-07, + "loss": 0.7171, + "step": 1368 + }, + { + "epoch": 1.3464566929133859, + "grad_norm": 11.54306411743164, + "learning_rate": 9.263050865467143e-07, + "loss": 0.7283, + "step": 1369 + }, + { + "epoch": 1.3474409448818898, + "grad_norm": 12.018163681030273, + "learning_rate": 9.259033415010952e-07, + "loss": 0.7313, + "step": 1370 + }, + { + "epoch": 1.3484251968503937, + "grad_norm": 14.016282081604004, + "learning_rate": 9.255005923105587e-07, + "loss": 0.9396, + "step": 1371 + }, + { + "epoch": 1.3494094488188977, + "grad_norm": 12.047124862670898, + "learning_rate": 9.250968399283902e-07, + "loss": 0.661, + "step": 1372 + }, + { + "epoch": 1.3503937007874016, + "grad_norm": 21.686264038085938, + "learning_rate": 9.24692085310249e-07, + "loss": 1.8785, + "step": 1373 + }, + { + "epoch": 1.3513779527559056, + "grad_norm": 13.831476211547852, + "learning_rate": 9.242863294141675e-07, + "loss": 0.4109, + "step": 1374 + }, + { + "epoch": 1.3523622047244095, + "grad_norm": 10.79392147064209, + "learning_rate": 9.238795732005475e-07, + "loss": 0.7084, + "step": 1375 + }, + { + "epoch": 1.3533464566929134, + "grad_norm": 9.851486206054688, + "learning_rate": 9.234718176321586e-07, + "loss": 0.4369, + "step": 1376 + }, + { + "epoch": 1.3543307086614174, + "grad_norm": 11.24117660522461, + "learning_rate": 9.230630636741358e-07, + "loss": 0.7483, + "step": 1377 + }, + { + "epoch": 1.3543307086614174, + "eval_Qnli-dev_cosine_accuracy": 0.671875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8206880688667297, + "eval_Qnli-dev_cosine_ap": 0.7020016169571128, + "eval_Qnli-dev_cosine_f1": 0.6984667802385007, + "eval_Qnli-dev_cosine_f1_threshold": 0.7731959819793701, + "eval_Qnli-dev_cosine_precision": 0.584045584045584, + "eval_Qnli-dev_cosine_recall": 0.8686440677966102, + "eval_Qnli-dev_dot_accuracy": 0.669921875, + "eval_Qnli-dev_dot_accuracy_threshold": 637.1887817382812, + "eval_Qnli-dev_dot_ap": 0.70064192309746, + "eval_Qnli-dev_dot_f1": 0.696095076400679, + "eval_Qnli-dev_dot_f1_threshold": 588.2445068359375, + "eval_Qnli-dev_dot_precision": 0.5807365439093485, + "eval_Qnli-dev_dot_recall": 0.8686440677966102, + "eval_Qnli-dev_euclidean_accuracy": 0.671875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 16.55626106262207, + "eval_Qnli-dev_euclidean_ap": 0.7017081005305794, + "eval_Qnli-dev_euclidean_f1": 0.6958041958041957, + "eval_Qnli-dev_euclidean_f1_threshold": 18.35430335998535, + "eval_Qnli-dev_euclidean_precision": 0.5922619047619048, + "eval_Qnli-dev_euclidean_recall": 0.8432203389830508, + "eval_Qnli-dev_manhattan_accuracy": 0.693359375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 337.3518981933594, + "eval_Qnli-dev_manhattan_ap": 0.7143798529397983, + "eval_Qnli-dev_manhattan_f1": 0.6998284734133791, + "eval_Qnli-dev_manhattan_f1_threshold": 390.21343994140625, + "eval_Qnli-dev_manhattan_precision": 0.5878962536023055, + "eval_Qnli-dev_manhattan_recall": 0.864406779661017, + "eval_Qnli-dev_max_accuracy": 0.693359375, + "eval_Qnli-dev_max_accuracy_threshold": 637.1887817382812, + "eval_Qnli-dev_max_ap": 0.7143798529397983, + "eval_Qnli-dev_max_f1": 0.6998284734133791, + "eval_Qnli-dev_max_f1_threshold": 588.2445068359375, + "eval_Qnli-dev_max_precision": 0.5922619047619048, + "eval_Qnli-dev_max_recall": 0.8686440677966102, + "eval_allNLI-dev_cosine_accuracy": 0.689453125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9173126816749573, + "eval_allNLI-dev_cosine_ap": 0.5006976649965474, + "eval_allNLI-dev_cosine_f1": 0.5413223140495869, + "eval_allNLI-dev_cosine_f1_threshold": 0.848089337348938, + "eval_allNLI-dev_cosine_precision": 0.4212218649517685, + "eval_allNLI-dev_cosine_recall": 0.7572254335260116, + "eval_allNLI-dev_dot_accuracy": 0.69140625, + "eval_allNLI-dev_dot_accuracy_threshold": 701.9290771484375, + "eval_allNLI-dev_dot_ap": 0.5021573167737639, + "eval_allNLI-dev_dot_f1": 0.5424430641821947, + "eval_allNLI-dev_dot_f1_threshold": 649.3282470703125, + "eval_allNLI-dev_dot_precision": 0.42258064516129035, + "eval_allNLI-dev_dot_recall": 0.7572254335260116, + "eval_allNLI-dev_euclidean_accuracy": 0.6875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.926703453063965, + "eval_allNLI-dev_euclidean_ap": 0.5003891082478613, + "eval_allNLI-dev_euclidean_f1": 0.5413223140495869, + "eval_allNLI-dev_euclidean_f1_threshold": 15.256593704223633, + "eval_allNLI-dev_euclidean_precision": 0.4212218649517685, + "eval_allNLI-dev_euclidean_recall": 0.7572254335260116, + "eval_allNLI-dev_manhattan_accuracy": 0.6875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 229.8043212890625, + "eval_allNLI-dev_manhattan_ap": 0.4991680725185854, + "eval_allNLI-dev_manhattan_f1": 0.5409090909090909, + "eval_allNLI-dev_manhattan_f1_threshold": 306.140380859375, + "eval_allNLI-dev_manhattan_precision": 0.44569288389513106, + "eval_allNLI-dev_manhattan_recall": 0.6878612716763006, + "eval_allNLI-dev_max_accuracy": 0.69140625, + "eval_allNLI-dev_max_accuracy_threshold": 701.9290771484375, + "eval_allNLI-dev_max_ap": 0.5021573167737639, + "eval_allNLI-dev_max_f1": 0.5424430641821947, + "eval_allNLI-dev_max_f1_threshold": 649.3282470703125, + "eval_allNLI-dev_max_precision": 0.44569288389513106, + "eval_allNLI-dev_max_recall": 0.7572254335260116, + "eval_loss": 1.1422334909439087, + "eval_runtime": 54.1068, + "eval_samples_per_second": 30.754, + "eval_sequential_score": 0.7143798529397983, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.559231077840739, + "eval_sts-test_pearson_dot": 0.5569194332314316, + "eval_sts-test_pearson_euclidean": 0.5735407858028017, + "eval_sts-test_pearson_manhattan": 0.5839195979621228, + "eval_sts-test_pearson_max": 0.5839195979621228, + "eval_sts-test_spearman_cosine": 0.5765562772319613, + "eval_sts-test_spearman_dot": 0.575029274690764, + "eval_sts-test_spearman_euclidean": 0.5768639421540271, + "eval_sts-test_spearman_manhattan": 0.5881996156695143, + "eval_sts-test_spearman_max": 0.5881996156695143, + "step": 1377 + }, + { + "epoch": 1.3553149606299213, + "grad_norm": 14.724981307983398, + "learning_rate": 9.226533122939772e-07, + "loss": 1.2169, + "step": 1378 + }, + { + "epoch": 1.3562992125984252, + "grad_norm": 20.81302261352539, + "learning_rate": 9.222425644615421e-07, + "loss": 1.2488, + "step": 1379 + }, + { + "epoch": 1.3572834645669292, + "grad_norm": 21.09238052368164, + "learning_rate": 9.218308211490479e-07, + "loss": 1.641, + "step": 1380 + }, + { + "epoch": 1.358267716535433, + "grad_norm": 11.956694602966309, + "learning_rate": 9.214180833310684e-07, + "loss": 0.7456, + "step": 1381 + }, + { + "epoch": 1.359251968503937, + "grad_norm": 16.93556022644043, + "learning_rate": 9.210043519845313e-07, + "loss": 1.1028, + "step": 1382 + }, + { + "epoch": 1.360236220472441, + "grad_norm": 16.941574096679688, + "learning_rate": 9.205896280887163e-07, + "loss": 1.0972, + "step": 1383 + }, + { + "epoch": 1.361220472440945, + "grad_norm": 11.039828300476074, + "learning_rate": 9.201739126252518e-07, + "loss": 0.6608, + "step": 1384 + }, + { + "epoch": 1.3622047244094488, + "grad_norm": 10.035400390625, + "learning_rate": 9.197572065781133e-07, + "loss": 0.4861, + "step": 1385 + }, + { + "epoch": 1.3631889763779528, + "grad_norm": 14.977723121643066, + "learning_rate": 9.193395109336216e-07, + "loss": 0.9011, + "step": 1386 + }, + { + "epoch": 1.3641732283464567, + "grad_norm": 16.243133544921875, + "learning_rate": 9.189208266804389e-07, + "loss": 1.1415, + "step": 1387 + }, + { + "epoch": 1.3651574803149606, + "grad_norm": 11.963306427001953, + "learning_rate": 9.185011548095682e-07, + "loss": 0.5773, + "step": 1388 + }, + { + "epoch": 1.3661417322834646, + "grad_norm": 11.360391616821289, + "learning_rate": 9.180804963143496e-07, + "loss": 0.6052, + "step": 1389 + }, + { + "epoch": 1.3671259842519685, + "grad_norm": 12.168638229370117, + "learning_rate": 9.176588521904588e-07, + "loss": 0.7691, + "step": 1390 + }, + { + "epoch": 1.3681102362204725, + "grad_norm": 7.507850170135498, + "learning_rate": 9.172362234359041e-07, + "loss": 0.3499, + "step": 1391 + }, + { + "epoch": 1.3690944881889764, + "grad_norm": 17.015483856201172, + "learning_rate": 9.168126110510246e-07, + "loss": 1.0002, + "step": 1392 + }, + { + "epoch": 1.3700787401574803, + "grad_norm": 11.774580001831055, + "learning_rate": 9.163880160384877e-07, + "loss": 0.8713, + "step": 1393 + }, + { + "epoch": 1.3710629921259843, + "grad_norm": 10.965306282043457, + "learning_rate": 9.159624394032864e-07, + "loss": 0.7215, + "step": 1394 + }, + { + "epoch": 1.3720472440944882, + "grad_norm": 17.416032791137695, + "learning_rate": 9.155358821527371e-07, + "loss": 0.954, + "step": 1395 + }, + { + "epoch": 1.3730314960629921, + "grad_norm": 15.789299011230469, + "learning_rate": 9.151083452964775e-07, + "loss": 1.0503, + "step": 1396 + }, + { + "epoch": 1.374015748031496, + "grad_norm": 12.647138595581055, + "learning_rate": 9.146798298464642e-07, + "loss": 1.0006, + "step": 1397 + }, + { + "epoch": 1.375, + "grad_norm": 11.720510482788086, + "learning_rate": 9.14250336816969e-07, + "loss": 0.8779, + "step": 1398 + }, + { + "epoch": 1.375984251968504, + "grad_norm": 9.829345703125, + "learning_rate": 9.13819867224579e-07, + "loss": 0.3058, + "step": 1399 + }, + { + "epoch": 1.3769685039370079, + "grad_norm": 9.576454162597656, + "learning_rate": 9.133884220881917e-07, + "loss": 0.4527, + "step": 1400 + }, + { + "epoch": 1.3779527559055118, + "grad_norm": 14.037468910217285, + "learning_rate": 9.129560024290145e-07, + "loss": 1.17, + "step": 1401 + }, + { + "epoch": 1.3789370078740157, + "grad_norm": 8.34862232208252, + "learning_rate": 9.125226092705603e-07, + "loss": 0.4068, + "step": 1402 + }, + { + "epoch": 1.3799212598425197, + "grad_norm": 10.172402381896973, + "learning_rate": 9.120882436386474e-07, + "loss": 0.5578, + "step": 1403 + }, + { + "epoch": 1.3809055118110236, + "grad_norm": 10.482789993286133, + "learning_rate": 9.116529065613954e-07, + "loss": 0.5727, + "step": 1404 + }, + { + "epoch": 1.3818897637795275, + "grad_norm": 12.011210441589355, + "learning_rate": 9.112165990692231e-07, + "loss": 1.0053, + "step": 1405 + }, + { + "epoch": 1.3828740157480315, + "grad_norm": 10.86913013458252, + "learning_rate": 9.107793221948464e-07, + "loss": 0.7875, + "step": 1406 + }, + { + "epoch": 1.3838582677165354, + "grad_norm": 17.681474685668945, + "learning_rate": 9.103410769732755e-07, + "loss": 1.2545, + "step": 1407 + }, + { + "epoch": 1.3848425196850394, + "grad_norm": 13.744503021240234, + "learning_rate": 9.099018644418133e-07, + "loss": 1.0035, + "step": 1408 + }, + { + "epoch": 1.3858267716535433, + "grad_norm": 12.040654182434082, + "learning_rate": 9.094616856400512e-07, + "loss": 0.7365, + "step": 1409 + }, + { + "epoch": 1.3868110236220472, + "grad_norm": 7.282107353210449, + "learning_rate": 9.090205416098685e-07, + "loss": 0.3283, + "step": 1410 + }, + { + "epoch": 1.3877952755905512, + "grad_norm": 12.776232719421387, + "learning_rate": 9.08578433395429e-07, + "loss": 0.839, + "step": 1411 + }, + { + "epoch": 1.388779527559055, + "grad_norm": 14.230220794677734, + "learning_rate": 9.081353620431783e-07, + "loss": 0.5744, + "step": 1412 + }, + { + "epoch": 1.389763779527559, + "grad_norm": 13.209874153137207, + "learning_rate": 9.07691328601842e-07, + "loss": 0.6241, + "step": 1413 + }, + { + "epoch": 1.390748031496063, + "grad_norm": 7.803886413574219, + "learning_rate": 9.072463341224232e-07, + "loss": 0.3192, + "step": 1414 + }, + { + "epoch": 1.391732283464567, + "grad_norm": 19.139183044433594, + "learning_rate": 9.068003796581991e-07, + "loss": 1.1741, + "step": 1415 + }, + { + "epoch": 1.3927165354330708, + "grad_norm": 8.626065254211426, + "learning_rate": 9.063534662647195e-07, + "loss": 0.4036, + "step": 1416 + }, + { + "epoch": 1.3937007874015748, + "grad_norm": 13.161855697631836, + "learning_rate": 9.059055949998039e-07, + "loss": 1.107, + "step": 1417 + }, + { + "epoch": 1.3946850393700787, + "grad_norm": 10.349092483520508, + "learning_rate": 9.05456766923539e-07, + "loss": 0.5198, + "step": 1418 + }, + { + "epoch": 1.3956692913385826, + "grad_norm": 11.868388175964355, + "learning_rate": 9.05006983098276e-07, + "loss": 0.5921, + "step": 1419 + }, + { + "epoch": 1.3966535433070866, + "grad_norm": 8.070558547973633, + "learning_rate": 9.045562445886289e-07, + "loss": 0.3738, + "step": 1420 + }, + { + "epoch": 1.3976377952755905, + "grad_norm": 16.018455505371094, + "learning_rate": 9.041045524614707e-07, + "loss": 0.8777, + "step": 1421 + }, + { + "epoch": 1.3986220472440944, + "grad_norm": 9.561680793762207, + "learning_rate": 9.03651907785932e-07, + "loss": 0.4179, + "step": 1422 + }, + { + "epoch": 1.3996062992125984, + "grad_norm": 13.034375190734863, + "learning_rate": 9.03198311633398e-07, + "loss": 1.0642, + "step": 1423 + }, + { + "epoch": 1.4005905511811023, + "grad_norm": 9.478670120239258, + "learning_rate": 9.027437650775055e-07, + "loss": 0.4586, + "step": 1424 + }, + { + "epoch": 1.4015748031496063, + "grad_norm": 8.93643569946289, + "learning_rate": 9.022882691941417e-07, + "loss": 0.5648, + "step": 1425 + }, + { + "epoch": 1.4025590551181102, + "grad_norm": 9.910337448120117, + "learning_rate": 9.018318250614402e-07, + "loss": 0.551, + "step": 1426 + }, + { + "epoch": 1.4035433070866141, + "grad_norm": 14.57641887664795, + "learning_rate": 9.01374433759779e-07, + "loss": 0.9472, + "step": 1427 + }, + { + "epoch": 1.404527559055118, + "grad_norm": 9.881360054016113, + "learning_rate": 9.009160963717784e-07, + "loss": 0.3759, + "step": 1428 + }, + { + "epoch": 1.405511811023622, + "grad_norm": 12.745662689208984, + "learning_rate": 9.004568139822978e-07, + "loss": 0.8034, + "step": 1429 + }, + { + "epoch": 1.406496062992126, + "grad_norm": 23.145109176635742, + "learning_rate": 8.999965876784333e-07, + "loss": 1.5744, + "step": 1430 + }, + { + "epoch": 1.4074803149606299, + "grad_norm": 8.001358032226562, + "learning_rate": 8.995354185495155e-07, + "loss": 0.2946, + "step": 1431 + }, + { + "epoch": 1.4084645669291338, + "grad_norm": 12.089577674865723, + "learning_rate": 8.990733076871063e-07, + "loss": 0.6559, + "step": 1432 + }, + { + "epoch": 1.4094488188976377, + "grad_norm": 11.743132591247559, + "learning_rate": 8.986102561849967e-07, + "loss": 0.5264, + "step": 1433 + }, + { + "epoch": 1.4104330708661417, + "grad_norm": 10.895527839660645, + "learning_rate": 8.981462651392041e-07, + "loss": 0.6314, + "step": 1434 + }, + { + "epoch": 1.4114173228346456, + "grad_norm": 17.545785903930664, + "learning_rate": 8.9768133564797e-07, + "loss": 1.0635, + "step": 1435 + }, + { + "epoch": 1.4124015748031495, + "grad_norm": 8.993785858154297, + "learning_rate": 8.972154688117568e-07, + "loss": 0.373, + "step": 1436 + }, + { + "epoch": 1.4133858267716535, + "grad_norm": 20.556406021118164, + "learning_rate": 8.967486657332462e-07, + "loss": 1.1461, + "step": 1437 + }, + { + "epoch": 1.4143700787401574, + "grad_norm": 18.048484802246094, + "learning_rate": 8.962809275173347e-07, + "loss": 0.8015, + "step": 1438 + }, + { + "epoch": 1.4153543307086613, + "grad_norm": 8.574625015258789, + "learning_rate": 8.958122552711336e-07, + "loss": 0.4479, + "step": 1439 + }, + { + "epoch": 1.4163385826771653, + "grad_norm": 9.441163063049316, + "learning_rate": 8.953426501039641e-07, + "loss": 0.3387, + "step": 1440 + }, + { + "epoch": 1.4173228346456692, + "grad_norm": 21.563499450683594, + "learning_rate": 8.94872113127356e-07, + "loss": 0.7423, + "step": 1441 + }, + { + "epoch": 1.4183070866141732, + "grad_norm": 14.528879165649414, + "learning_rate": 8.944006454550443e-07, + "loss": 0.9856, + "step": 1442 + }, + { + "epoch": 1.419291338582677, + "grad_norm": 9.45664119720459, + "learning_rate": 8.939282482029671e-07, + "loss": 0.449, + "step": 1443 + }, + { + "epoch": 1.420275590551181, + "grad_norm": 14.329258918762207, + "learning_rate": 8.934549224892629e-07, + "loss": 0.6372, + "step": 1444 + }, + { + "epoch": 1.421259842519685, + "grad_norm": 8.746792793273926, + "learning_rate": 8.929806694342677e-07, + "loss": 0.3884, + "step": 1445 + }, + { + "epoch": 1.422244094488189, + "grad_norm": 10.391413688659668, + "learning_rate": 8.925054901605124e-07, + "loss": 0.5394, + "step": 1446 + }, + { + "epoch": 1.423228346456693, + "grad_norm": 15.459757804870605, + "learning_rate": 8.9202938579272e-07, + "loss": 0.8704, + "step": 1447 + }, + { + "epoch": 1.4242125984251968, + "grad_norm": 9.648519515991211, + "learning_rate": 8.915523574578038e-07, + "loss": 0.6108, + "step": 1448 + }, + { + "epoch": 1.425196850393701, + "grad_norm": 19.38627052307129, + "learning_rate": 8.910744062848633e-07, + "loss": 1.1582, + "step": 1449 + }, + { + "epoch": 1.4261811023622046, + "grad_norm": 21.816274642944336, + "learning_rate": 8.90595533405183e-07, + "loss": 1.5682, + "step": 1450 + }, + { + "epoch": 1.4271653543307088, + "grad_norm": 10.449719429016113, + "learning_rate": 8.901157399522286e-07, + "loss": 0.4988, + "step": 1451 + }, + { + "epoch": 1.4281496062992125, + "grad_norm": 11.805849075317383, + "learning_rate": 8.896350270616449e-07, + "loss": 0.8312, + "step": 1452 + }, + { + "epoch": 1.4291338582677167, + "grad_norm": 20.020519256591797, + "learning_rate": 8.891533958712529e-07, + "loss": 0.9784, + "step": 1453 + }, + { + "epoch": 1.4301181102362204, + "grad_norm": 11.389593124389648, + "learning_rate": 8.88670847521047e-07, + "loss": 0.6923, + "step": 1454 + }, + { + "epoch": 1.4311023622047245, + "grad_norm": 10.024092674255371, + "learning_rate": 8.881873831531928e-07, + "loss": 0.4661, + "step": 1455 + }, + { + "epoch": 1.4320866141732282, + "grad_norm": 6.586660385131836, + "learning_rate": 8.877030039120242e-07, + "loss": 0.229, + "step": 1456 + }, + { + "epoch": 1.4330708661417324, + "grad_norm": 11.63663387298584, + "learning_rate": 8.872177109440398e-07, + "loss": 0.6479, + "step": 1457 + }, + { + "epoch": 1.4340551181102361, + "grad_norm": 11.15836238861084, + "learning_rate": 8.867315053979016e-07, + "loss": 0.6628, + "step": 1458 + }, + { + "epoch": 1.4350393700787403, + "grad_norm": 13.676897048950195, + "learning_rate": 8.862443884244314e-07, + "loss": 0.7929, + "step": 1459 + }, + { + "epoch": 1.436023622047244, + "grad_norm": 6.395401477813721, + "learning_rate": 8.857563611766085e-07, + "loss": 0.2601, + "step": 1460 + }, + { + "epoch": 1.4370078740157481, + "grad_norm": 11.027069091796875, + "learning_rate": 8.852674248095663e-07, + "loss": 0.5944, + "step": 1461 + }, + { + "epoch": 1.4379921259842519, + "grad_norm": 14.573845863342285, + "learning_rate": 8.847775804805907e-07, + "loss": 0.9913, + "step": 1462 + }, + { + "epoch": 1.438976377952756, + "grad_norm": 9.79361629486084, + "learning_rate": 8.84286829349116e-07, + "loss": 0.4411, + "step": 1463 + }, + { + "epoch": 1.4399606299212597, + "grad_norm": 7.446544647216797, + "learning_rate": 8.837951725767234e-07, + "loss": 0.3742, + "step": 1464 + }, + { + "epoch": 1.4409448818897639, + "grad_norm": 10.105347633361816, + "learning_rate": 8.833026113271374e-07, + "loss": 0.398, + "step": 1465 + }, + { + "epoch": 1.4419291338582676, + "grad_norm": 12.443421363830566, + "learning_rate": 8.828091467662234e-07, + "loss": 0.579, + "step": 1466 + }, + { + "epoch": 1.4429133858267718, + "grad_norm": 14.672910690307617, + "learning_rate": 8.82314780061985e-07, + "loss": 0.9689, + "step": 1467 + }, + { + "epoch": 1.4438976377952755, + "grad_norm": 11.923147201538086, + "learning_rate": 8.81819512384561e-07, + "loss": 0.5551, + "step": 1468 + }, + { + "epoch": 1.4448818897637796, + "grad_norm": 11.983245849609375, + "learning_rate": 8.813233449062229e-07, + "loss": 0.4934, + "step": 1469 + }, + { + "epoch": 1.4458661417322833, + "grad_norm": 16.73973274230957, + "learning_rate": 8.808262788013718e-07, + "loss": 0.9175, + "step": 1470 + }, + { + "epoch": 1.4468503937007875, + "grad_norm": 23.527509689331055, + "learning_rate": 8.803283152465357e-07, + "loss": 1.5047, + "step": 1471 + }, + { + "epoch": 1.4478346456692912, + "grad_norm": 10.171259880065918, + "learning_rate": 8.798294554203674e-07, + "loss": 0.5172, + "step": 1472 + }, + { + "epoch": 1.4488188976377954, + "grad_norm": 22.97465705871582, + "learning_rate": 8.793297005036402e-07, + "loss": 1.2806, + "step": 1473 + }, + { + "epoch": 1.449803149606299, + "grad_norm": 11.81129264831543, + "learning_rate": 8.788290516792471e-07, + "loss": 0.8314, + "step": 1474 + }, + { + "epoch": 1.4507874015748032, + "grad_norm": 14.46216869354248, + "learning_rate": 8.783275101321962e-07, + "loss": 0.5415, + "step": 1475 + }, + { + "epoch": 1.451771653543307, + "grad_norm": 13.006589889526367, + "learning_rate": 8.778250770496084e-07, + "loss": 0.5999, + "step": 1476 + }, + { + "epoch": 1.452755905511811, + "grad_norm": 12.510464668273926, + "learning_rate": 8.773217536207158e-07, + "loss": 0.7689, + "step": 1477 + }, + { + "epoch": 1.453740157480315, + "grad_norm": 19.87265968322754, + "learning_rate": 8.76817541036857e-07, + "loss": 1.1297, + "step": 1478 + }, + { + "epoch": 1.454724409448819, + "grad_norm": 10.278497695922852, + "learning_rate": 8.763124404914754e-07, + "loss": 0.3801, + "step": 1479 + }, + { + "epoch": 1.455708661417323, + "grad_norm": 13.605413436889648, + "learning_rate": 8.758064531801166e-07, + "loss": 0.3896, + "step": 1480 + }, + { + "epoch": 1.4566929133858268, + "grad_norm": 10.33932876586914, + "learning_rate": 8.752995803004247e-07, + "loss": 0.4557, + "step": 1481 + }, + { + "epoch": 1.4576771653543308, + "grad_norm": 10.69277286529541, + "learning_rate": 8.747918230521398e-07, + "loss": 0.5844, + "step": 1482 + }, + { + "epoch": 1.4586614173228347, + "grad_norm": 15.579158782958984, + "learning_rate": 8.742831826370956e-07, + "loss": 0.8164, + "step": 1483 + }, + { + "epoch": 1.4596456692913387, + "grad_norm": 9.021865844726562, + "learning_rate": 8.737736602592158e-07, + "loss": 0.4885, + "step": 1484 + }, + { + "epoch": 1.4606299212598426, + "grad_norm": 11.235892295837402, + "learning_rate": 8.732632571245121e-07, + "loss": 0.9719, + "step": 1485 + }, + { + "epoch": 1.4616141732283465, + "grad_norm": 13.27570629119873, + "learning_rate": 8.727519744410806e-07, + "loss": 0.6612, + "step": 1486 + }, + { + "epoch": 1.4625984251968505, + "grad_norm": 11.570671081542969, + "learning_rate": 8.722398134190995e-07, + "loss": 0.4546, + "step": 1487 + }, + { + "epoch": 1.4635826771653544, + "grad_norm": 17.243309020996094, + "learning_rate": 8.717267752708257e-07, + "loss": 1.1661, + "step": 1488 + }, + { + "epoch": 1.4645669291338583, + "grad_norm": 14.900501251220703, + "learning_rate": 8.712128612105924e-07, + "loss": 0.6853, + "step": 1489 + }, + { + "epoch": 1.4655511811023623, + "grad_norm": 9.174758911132812, + "learning_rate": 8.706980724548058e-07, + "loss": 0.3457, + "step": 1490 + }, + { + "epoch": 1.4665354330708662, + "grad_norm": 10.846331596374512, + "learning_rate": 8.701824102219428e-07, + "loss": 0.5259, + "step": 1491 + }, + { + "epoch": 1.4675196850393701, + "grad_norm": 12.600656509399414, + "learning_rate": 8.696658757325474e-07, + "loss": 0.8553, + "step": 1492 + }, + { + "epoch": 1.468503937007874, + "grad_norm": 10.123764991760254, + "learning_rate": 8.691484702092286e-07, + "loss": 0.3868, + "step": 1493 + }, + { + "epoch": 1.469488188976378, + "grad_norm": 8.318851470947266, + "learning_rate": 8.686301948766568e-07, + "loss": 0.2448, + "step": 1494 + }, + { + "epoch": 1.470472440944882, + "grad_norm": 8.971074104309082, + "learning_rate": 8.681110509615612e-07, + "loss": 0.435, + "step": 1495 + }, + { + "epoch": 1.4714566929133859, + "grad_norm": 13.754014015197754, + "learning_rate": 8.675910396927269e-07, + "loss": 0.6314, + "step": 1496 + }, + { + "epoch": 1.4724409448818898, + "grad_norm": 8.66455364227295, + "learning_rate": 8.670701623009921e-07, + "loss": 0.5653, + "step": 1497 + }, + { + "epoch": 1.4734251968503937, + "grad_norm": 15.178778648376465, + "learning_rate": 8.66548420019245e-07, + "loss": 0.8, + "step": 1498 + }, + { + "epoch": 1.4744094488188977, + "grad_norm": 10.51429557800293, + "learning_rate": 8.66025814082421e-07, + "loss": 0.5797, + "step": 1499 + }, + { + "epoch": 1.4753937007874016, + "grad_norm": 16.99217414855957, + "learning_rate": 8.655023457274994e-07, + "loss": 0.827, + "step": 1500 + }, + { + "epoch": 1.4763779527559056, + "grad_norm": 14.463443756103516, + "learning_rate": 8.649780161935012e-07, + "loss": 1.0582, + "step": 1501 + }, + { + "epoch": 1.4773622047244095, + "grad_norm": 9.892260551452637, + "learning_rate": 8.644528267214859e-07, + "loss": 0.2981, + "step": 1502 + }, + { + "epoch": 1.4783464566929134, + "grad_norm": 10.980031967163086, + "learning_rate": 8.639267785545478e-07, + "loss": 0.8084, + "step": 1503 + }, + { + "epoch": 1.4793307086614174, + "grad_norm": 11.92680835723877, + "learning_rate": 8.63399872937814e-07, + "loss": 1.1578, + "step": 1504 + }, + { + "epoch": 1.4803149606299213, + "grad_norm": 14.131094932556152, + "learning_rate": 8.628721111184414e-07, + "loss": 0.7456, + "step": 1505 + }, + { + "epoch": 1.4812992125984252, + "grad_norm": 13.81746768951416, + "learning_rate": 8.623434943456132e-07, + "loss": 0.8266, + "step": 1506 + }, + { + "epoch": 1.4822834645669292, + "grad_norm": 9.297046661376953, + "learning_rate": 8.61814023870536e-07, + "loss": 0.3407, + "step": 1507 + }, + { + "epoch": 1.483267716535433, + "grad_norm": 13.54496955871582, + "learning_rate": 8.612837009464377e-07, + "loss": 0.7785, + "step": 1508 + }, + { + "epoch": 1.484251968503937, + "grad_norm": 16.752914428710938, + "learning_rate": 8.607525268285634e-07, + "loss": 1.2844, + "step": 1509 + }, + { + "epoch": 1.485236220472441, + "grad_norm": 14.00351333618164, + "learning_rate": 8.602205027741731e-07, + "loss": 0.6928, + "step": 1510 + }, + { + "epoch": 1.486220472440945, + "grad_norm": 12.860536575317383, + "learning_rate": 8.596876300425384e-07, + "loss": 0.3945, + "step": 1511 + }, + { + "epoch": 1.4872047244094488, + "grad_norm": 10.744647026062012, + "learning_rate": 8.591539098949401e-07, + "loss": 0.5651, + "step": 1512 + }, + { + "epoch": 1.4881889763779528, + "grad_norm": 9.299052238464355, + "learning_rate": 8.586193435946641e-07, + "loss": 0.4731, + "step": 1513 + }, + { + "epoch": 1.4891732283464567, + "grad_norm": 7.104952812194824, + "learning_rate": 8.580839324069998e-07, + "loss": 0.249, + "step": 1514 + }, + { + "epoch": 1.4901574803149606, + "grad_norm": 11.674582481384277, + "learning_rate": 8.57547677599236e-07, + "loss": 0.4961, + "step": 1515 + }, + { + "epoch": 1.4911417322834646, + "grad_norm": 11.296091079711914, + "learning_rate": 8.570105804406583e-07, + "loss": 0.6058, + "step": 1516 + }, + { + "epoch": 1.4921259842519685, + "grad_norm": 15.509420394897461, + "learning_rate": 8.564726422025462e-07, + "loss": 1.0121, + "step": 1517 + }, + { + "epoch": 1.4931102362204725, + "grad_norm": 9.842504501342773, + "learning_rate": 8.5593386415817e-07, + "loss": 0.4994, + "step": 1518 + }, + { + "epoch": 1.4940944881889764, + "grad_norm": 11.032633781433105, + "learning_rate": 8.553942475827879e-07, + "loss": 0.3958, + "step": 1519 + }, + { + "epoch": 1.4950787401574803, + "grad_norm": 13.140924453735352, + "learning_rate": 8.548537937536425e-07, + "loss": 0.5271, + "step": 1520 + }, + { + "epoch": 1.4960629921259843, + "grad_norm": 7.572324275970459, + "learning_rate": 8.543125039499582e-07, + "loss": 0.3546, + "step": 1521 + }, + { + "epoch": 1.4970472440944882, + "grad_norm": 11.842230796813965, + "learning_rate": 8.537703794529387e-07, + "loss": 0.5617, + "step": 1522 + }, + { + "epoch": 1.4980314960629921, + "grad_norm": 9.245935440063477, + "learning_rate": 8.532274215457625e-07, + "loss": 0.4117, + "step": 1523 + }, + { + "epoch": 1.499015748031496, + "grad_norm": 19.322467803955078, + "learning_rate": 8.526836315135815e-07, + "loss": 1.4661, + "step": 1524 + }, + { + "epoch": 1.5, + "grad_norm": 14.873554229736328, + "learning_rate": 8.521390106435168e-07, + "loss": 0.8395, + "step": 1525 + }, + { + "epoch": 1.500984251968504, + "grad_norm": 9.65881633758545, + "learning_rate": 8.515935602246558e-07, + "loss": 0.4478, + "step": 1526 + }, + { + "epoch": 1.5019685039370079, + "grad_norm": 10.6986665725708, + "learning_rate": 8.510472815480498e-07, + "loss": 0.5156, + "step": 1527 + }, + { + "epoch": 1.5029527559055118, + "grad_norm": 10.98793888092041, + "learning_rate": 8.505001759067106e-07, + "loss": 0.6582, + "step": 1528 + }, + { + "epoch": 1.5039370078740157, + "grad_norm": 7.674596786499023, + "learning_rate": 8.499522445956071e-07, + "loss": 0.3326, + "step": 1529 + }, + { + "epoch": 1.5049212598425197, + "grad_norm": 9.579343795776367, + "learning_rate": 8.494034889116626e-07, + "loss": 0.6277, + "step": 1530 + }, + { + "epoch": 1.5049212598425197, + "eval_Qnli-dev_cosine_accuracy": 0.66796875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7962973117828369, + "eval_Qnli-dev_cosine_ap": 0.691759185255588, + "eval_Qnli-dev_cosine_f1": 0.6839546191247974, + "eval_Qnli-dev_cosine_f1_threshold": 0.7521535158157349, + "eval_Qnli-dev_cosine_precision": 0.5538057742782152, + "eval_Qnli-dev_cosine_recall": 0.8940677966101694, + "eval_Qnli-dev_dot_accuracy": 0.662109375, + "eval_Qnli-dev_dot_accuracy_threshold": 606.8631591796875, + "eval_Qnli-dev_dot_ap": 0.6893728512808852, + "eval_Qnli-dev_dot_f1": 0.6848816029143898, + "eval_Qnli-dev_dot_f1_threshold": 601.61083984375, + "eval_Qnli-dev_dot_precision": 0.6006389776357828, + "eval_Qnli-dev_dot_recall": 0.7966101694915254, + "eval_Qnli-dev_euclidean_accuracy": 0.671875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.649578094482422, + "eval_Qnli-dev_euclidean_ap": 0.6921233720273546, + "eval_Qnli-dev_euclidean_f1": 0.6839546191247974, + "eval_Qnli-dev_euclidean_f1_threshold": 19.418928146362305, + "eval_Qnli-dev_euclidean_precision": 0.5538057742782152, + "eval_Qnli-dev_euclidean_recall": 0.8940677966101694, + "eval_Qnli-dev_manhattan_accuracy": 0.669921875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 355.29461669921875, + "eval_Qnli-dev_manhattan_ap": 0.7063039458792412, + "eval_Qnli-dev_manhattan_f1": 0.6957983193277311, + "eval_Qnli-dev_manhattan_f1_threshold": 397.1031799316406, + "eval_Qnli-dev_manhattan_precision": 0.5766016713091922, + "eval_Qnli-dev_manhattan_recall": 0.8771186440677966, + "eval_Qnli-dev_max_accuracy": 0.671875, + "eval_Qnli-dev_max_accuracy_threshold": 606.8631591796875, + "eval_Qnli-dev_max_ap": 0.7063039458792412, + "eval_Qnli-dev_max_f1": 0.6957983193277311, + "eval_Qnli-dev_max_f1_threshold": 601.61083984375, + "eval_Qnli-dev_max_precision": 0.6006389776357828, + "eval_Qnli-dev_max_recall": 0.8940677966101694, + "eval_allNLI-dev_cosine_accuracy": 0.689453125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9331935048103333, + "eval_allNLI-dev_cosine_ap": 0.5030724645677387, + "eval_allNLI-dev_cosine_f1": 0.5440613026819924, + "eval_allNLI-dev_cosine_f1_threshold": 0.819710373878479, + "eval_allNLI-dev_cosine_precision": 0.4068767908309456, + "eval_allNLI-dev_cosine_recall": 0.8208092485549133, + "eval_allNLI-dev_dot_accuracy": 0.689453125, + "eval_allNLI-dev_dot_accuracy_threshold": 707.578857421875, + "eval_allNLI-dev_dot_ap": 0.5031528402578558, + "eval_allNLI-dev_dot_f1": 0.543726235741445, + "eval_allNLI-dev_dot_f1_threshold": 624.6593017578125, + "eval_allNLI-dev_dot_precision": 0.40509915014164305, + "eval_allNLI-dev_dot_recall": 0.8265895953757225, + "eval_allNLI-dev_euclidean_accuracy": 0.689453125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.111281394958496, + "eval_allNLI-dev_euclidean_ap": 0.5029907933861366, + "eval_allNLI-dev_euclidean_f1": 0.5461538461538462, + "eval_allNLI-dev_euclidean_f1_threshold": 16.565919876098633, + "eval_allNLI-dev_euclidean_precision": 0.4092219020172911, + "eval_allNLI-dev_euclidean_recall": 0.8208092485549133, + "eval_allNLI-dev_manhattan_accuracy": 0.685546875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 224.79913330078125, + "eval_allNLI-dev_manhattan_ap": 0.5004588997249526, + "eval_allNLI-dev_manhattan_f1": 0.5419103313840157, + "eval_allNLI-dev_manhattan_f1_threshold": 342.75079345703125, + "eval_allNLI-dev_manhattan_precision": 0.4088235294117647, + "eval_allNLI-dev_manhattan_recall": 0.8034682080924855, + "eval_allNLI-dev_max_accuracy": 0.689453125, + "eval_allNLI-dev_max_accuracy_threshold": 707.578857421875, + "eval_allNLI-dev_max_ap": 0.5031528402578558, + "eval_allNLI-dev_max_f1": 0.5461538461538462, + "eval_allNLI-dev_max_f1_threshold": 624.6593017578125, + "eval_allNLI-dev_max_precision": 0.4092219020172911, + "eval_allNLI-dev_max_recall": 0.8265895953757225, + "eval_loss": 1.0736732482910156, + "eval_runtime": 54.2804, + "eval_samples_per_second": 30.656, + "eval_sequential_score": 0.7063039458792412, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.6063324046452339, + "eval_sts-test_pearson_dot": 0.6046695712154431, + "eval_sts-test_pearson_euclidean": 0.6274252948479226, + "eval_sts-test_pearson_manhattan": 0.6331295272217956, + "eval_sts-test_pearson_max": 0.6331295272217956, + "eval_sts-test_spearman_cosine": 0.6293434778801531, + "eval_sts-test_spearman_dot": 0.6275092131782424, + "eval_sts-test_spearman_euclidean": 0.6293739534891606, + "eval_sts-test_spearman_manhattan": 0.6344549222050173, + "eval_sts-test_spearman_max": 0.6344549222050173, + "step": 1530 + }, + { + "epoch": 1.5059055118110236, + "grad_norm": 12.969947814941406, + "learning_rate": 8.488539101537518e-07, + "loss": 0.6134, + "step": 1531 + }, + { + "epoch": 1.5068897637795275, + "grad_norm": 16.989253997802734, + "learning_rate": 8.483035096226975e-07, + "loss": 0.8329, + "step": 1532 + }, + { + "epoch": 1.5078740157480315, + "grad_norm": 9.056514739990234, + "learning_rate": 8.477522886212673e-07, + "loss": 0.458, + "step": 1533 + }, + { + "epoch": 1.5088582677165354, + "grad_norm": 18.400096893310547, + "learning_rate": 8.472002484541711e-07, + "loss": 0.7997, + "step": 1534 + }, + { + "epoch": 1.5098425196850394, + "grad_norm": 14.392437934875488, + "learning_rate": 8.466473904280579e-07, + "loss": 0.9097, + "step": 1535 + }, + { + "epoch": 1.5108267716535433, + "grad_norm": 9.243882179260254, + "learning_rate": 8.46093715851512e-07, + "loss": 0.4139, + "step": 1536 + }, + { + "epoch": 1.5118110236220472, + "grad_norm": 21.380508422851562, + "learning_rate": 8.455392260350507e-07, + "loss": 1.1263, + "step": 1537 + }, + { + "epoch": 1.5127952755905512, + "grad_norm": 13.089934349060059, + "learning_rate": 8.449839222911214e-07, + "loss": 0.4691, + "step": 1538 + }, + { + "epoch": 1.513779527559055, + "grad_norm": 8.922409057617188, + "learning_rate": 8.444278059340971e-07, + "loss": 0.4162, + "step": 1539 + }, + { + "epoch": 1.514763779527559, + "grad_norm": 10.464616775512695, + "learning_rate": 8.438708782802748e-07, + "loss": 0.4039, + "step": 1540 + }, + { + "epoch": 1.515748031496063, + "grad_norm": 11.656524658203125, + "learning_rate": 8.43313140647872e-07, + "loss": 0.6962, + "step": 1541 + }, + { + "epoch": 1.516732283464567, + "grad_norm": 9.229491233825684, + "learning_rate": 8.427545943570226e-07, + "loss": 0.4833, + "step": 1542 + }, + { + "epoch": 1.5177165354330708, + "grad_norm": 13.765993118286133, + "learning_rate": 8.421952407297752e-07, + "loss": 0.8962, + "step": 1543 + }, + { + "epoch": 1.5187007874015748, + "grad_norm": 5.461718559265137, + "learning_rate": 8.416350810900892e-07, + "loss": 0.1782, + "step": 1544 + }, + { + "epoch": 1.5196850393700787, + "grad_norm": 11.364346504211426, + "learning_rate": 8.410741167638317e-07, + "loss": 0.7229, + "step": 1545 + }, + { + "epoch": 1.5206692913385826, + "grad_norm": 8.603964805603027, + "learning_rate": 8.405123490787741e-07, + "loss": 0.3197, + "step": 1546 + }, + { + "epoch": 1.5216535433070866, + "grad_norm": 12.854615211486816, + "learning_rate": 8.3994977936459e-07, + "loss": 0.7508, + "step": 1547 + }, + { + "epoch": 1.5226377952755905, + "grad_norm": 12.71799087524414, + "learning_rate": 8.393864089528514e-07, + "loss": 0.6946, + "step": 1548 + }, + { + "epoch": 1.5236220472440944, + "grad_norm": 7.034970760345459, + "learning_rate": 8.388222391770245e-07, + "loss": 0.3118, + "step": 1549 + }, + { + "epoch": 1.5246062992125984, + "grad_norm": 16.605329513549805, + "learning_rate": 8.382572713724683e-07, + "loss": 0.8608, + "step": 1550 + }, + { + "epoch": 1.5255905511811023, + "grad_norm": 15.885974884033203, + "learning_rate": 8.376915068764308e-07, + "loss": 0.7907, + "step": 1551 + }, + { + "epoch": 1.5265748031496063, + "grad_norm": 22.729400634765625, + "learning_rate": 8.371249470280453e-07, + "loss": 1.2424, + "step": 1552 + }, + { + "epoch": 1.5275590551181102, + "grad_norm": 10.601911544799805, + "learning_rate": 8.36557593168328e-07, + "loss": 0.5366, + "step": 1553 + }, + { + "epoch": 1.5285433070866141, + "grad_norm": 10.668947219848633, + "learning_rate": 8.359894466401743e-07, + "loss": 0.5525, + "step": 1554 + }, + { + "epoch": 1.529527559055118, + "grad_norm": 7.3187150955200195, + "learning_rate": 8.354205087883556e-07, + "loss": 0.2904, + "step": 1555 + }, + { + "epoch": 1.530511811023622, + "grad_norm": 13.080991744995117, + "learning_rate": 8.348507809595169e-07, + "loss": 0.7339, + "step": 1556 + }, + { + "epoch": 1.531496062992126, + "grad_norm": 7.837886333465576, + "learning_rate": 8.342802645021727e-07, + "loss": 0.4061, + "step": 1557 + }, + { + "epoch": 1.53248031496063, + "grad_norm": 9.239282608032227, + "learning_rate": 8.337089607667039e-07, + "loss": 0.3988, + "step": 1558 + }, + { + "epoch": 1.5334645669291338, + "grad_norm": 14.300249099731445, + "learning_rate": 8.331368711053552e-07, + "loss": 0.6063, + "step": 1559 + }, + { + "epoch": 1.534448818897638, + "grad_norm": 13.190995216369629, + "learning_rate": 8.325639968722313e-07, + "loss": 0.7394, + "step": 1560 + }, + { + "epoch": 1.5354330708661417, + "grad_norm": 22.914588928222656, + "learning_rate": 8.319903394232942e-07, + "loss": 1.2118, + "step": 1561 + }, + { + "epoch": 1.5364173228346458, + "grad_norm": 12.508071899414062, + "learning_rate": 8.314159001163596e-07, + "loss": 0.5494, + "step": 1562 + }, + { + "epoch": 1.5374015748031495, + "grad_norm": 10.416210174560547, + "learning_rate": 8.308406803110936e-07, + "loss": 0.4957, + "step": 1563 + }, + { + "epoch": 1.5383858267716537, + "grad_norm": 12.602110862731934, + "learning_rate": 8.302646813690101e-07, + "loss": 0.628, + "step": 1564 + }, + { + "epoch": 1.5393700787401574, + "grad_norm": 10.616961479187012, + "learning_rate": 8.296879046534668e-07, + "loss": 0.454, + "step": 1565 + }, + { + "epoch": 1.5403543307086616, + "grad_norm": 10.201837539672852, + "learning_rate": 8.291103515296628e-07, + "loss": 0.4168, + "step": 1566 + }, + { + "epoch": 1.5413385826771653, + "grad_norm": 10.088976860046387, + "learning_rate": 8.285320233646345e-07, + "loss": 0.4472, + "step": 1567 + }, + { + "epoch": 1.5423228346456694, + "grad_norm": 11.572919845581055, + "learning_rate": 8.279529215272526e-07, + "loss": 0.4394, + "step": 1568 + }, + { + "epoch": 1.5433070866141732, + "grad_norm": 24.046588897705078, + "learning_rate": 8.273730473882198e-07, + "loss": 1.1084, + "step": 1569 + }, + { + "epoch": 1.5442913385826773, + "grad_norm": 10.44725513458252, + "learning_rate": 8.267924023200662e-07, + "loss": 0.3262, + "step": 1570 + }, + { + "epoch": 1.545275590551181, + "grad_norm": 8.531761169433594, + "learning_rate": 8.262109876971466e-07, + "loss": 0.3507, + "step": 1571 + }, + { + "epoch": 1.5462598425196852, + "grad_norm": 16.767059326171875, + "learning_rate": 8.256288048956379e-07, + "loss": 1.0028, + "step": 1572 + }, + { + "epoch": 1.547244094488189, + "grad_norm": 9.508438110351562, + "learning_rate": 8.250458552935343e-07, + "loss": 0.5218, + "step": 1573 + }, + { + "epoch": 1.548228346456693, + "grad_norm": 13.773395538330078, + "learning_rate": 8.244621402706462e-07, + "loss": 0.9723, + "step": 1574 + }, + { + "epoch": 1.5492125984251968, + "grad_norm": 14.080069541931152, + "learning_rate": 8.238776612085943e-07, + "loss": 0.5616, + "step": 1575 + }, + { + "epoch": 1.550196850393701, + "grad_norm": 11.193357467651367, + "learning_rate": 8.232924194908088e-07, + "loss": 0.4887, + "step": 1576 + }, + { + "epoch": 1.5511811023622046, + "grad_norm": 9.106607437133789, + "learning_rate": 8.227064165025247e-07, + "loss": 0.3899, + "step": 1577 + }, + { + "epoch": 1.5521653543307088, + "grad_norm": 8.196263313293457, + "learning_rate": 8.221196536307788e-07, + "loss": 0.3545, + "step": 1578 + }, + { + "epoch": 1.5531496062992125, + "grad_norm": 9.231218338012695, + "learning_rate": 8.215321322644067e-07, + "loss": 0.3574, + "step": 1579 + }, + { + "epoch": 1.5541338582677167, + "grad_norm": 22.20115852355957, + "learning_rate": 8.209438537940393e-07, + "loss": 1.3911, + "step": 1580 + }, + { + "epoch": 1.5551181102362204, + "grad_norm": 11.824051856994629, + "learning_rate": 8.203548196120994e-07, + "loss": 0.7395, + "step": 1581 + }, + { + "epoch": 1.5561023622047245, + "grad_norm": 9.031845092773438, + "learning_rate": 8.197650311127981e-07, + "loss": 0.2491, + "step": 1582 + }, + { + "epoch": 1.5570866141732282, + "grad_norm": 10.274392127990723, + "learning_rate": 8.191744896921331e-07, + "loss": 0.6825, + "step": 1583 + }, + { + "epoch": 1.5580708661417324, + "grad_norm": 5.4450154304504395, + "learning_rate": 8.185831967478831e-07, + "loss": 0.1434, + "step": 1584 + }, + { + "epoch": 1.5590551181102361, + "grad_norm": 11.918195724487305, + "learning_rate": 8.179911536796062e-07, + "loss": 0.3764, + "step": 1585 + }, + { + "epoch": 1.5600393700787403, + "grad_norm": 11.091309547424316, + "learning_rate": 8.173983618886358e-07, + "loss": 0.5378, + "step": 1586 + }, + { + "epoch": 1.561023622047244, + "grad_norm": 8.367671966552734, + "learning_rate": 8.168048227780775e-07, + "loss": 0.2975, + "step": 1587 + }, + { + "epoch": 1.5620078740157481, + "grad_norm": 15.36435604095459, + "learning_rate": 8.162105377528057e-07, + "loss": 1.0245, + "step": 1588 + }, + { + "epoch": 1.5629921259842519, + "grad_norm": 14.115655899047852, + "learning_rate": 8.156155082194607e-07, + "loss": 0.6128, + "step": 1589 + }, + { + "epoch": 1.563976377952756, + "grad_norm": 17.05628776550293, + "learning_rate": 8.150197355864446e-07, + "loss": 0.89, + "step": 1590 + }, + { + "epoch": 1.5649606299212597, + "grad_norm": 14.739027976989746, + "learning_rate": 8.144232212639184e-07, + "loss": 0.8377, + "step": 1591 + }, + { + "epoch": 1.5659448818897639, + "grad_norm": 21.410964965820312, + "learning_rate": 8.138259666637987e-07, + "loss": 0.7711, + "step": 1592 + }, + { + "epoch": 1.5669291338582676, + "grad_norm": 14.888060569763184, + "learning_rate": 8.132279731997544e-07, + "loss": 0.8045, + "step": 1593 + }, + { + "epoch": 1.5679133858267718, + "grad_norm": 11.131561279296875, + "learning_rate": 8.126292422872034e-07, + "loss": 0.5162, + "step": 1594 + }, + { + "epoch": 1.5688976377952755, + "grad_norm": 12.947890281677246, + "learning_rate": 8.120297753433087e-07, + "loss": 0.5009, + "step": 1595 + }, + { + "epoch": 1.5698818897637796, + "grad_norm": 16.48666763305664, + "learning_rate": 8.114295737869759e-07, + "loss": 0.6454, + "step": 1596 + }, + { + "epoch": 1.5708661417322833, + "grad_norm": 13.440455436706543, + "learning_rate": 8.108286390388487e-07, + "loss": 0.4929, + "step": 1597 + }, + { + "epoch": 1.5718503937007875, + "grad_norm": 9.928903579711914, + "learning_rate": 8.102269725213071e-07, + "loss": 0.4363, + "step": 1598 + }, + { + "epoch": 1.5728346456692912, + "grad_norm": 13.543412208557129, + "learning_rate": 8.096245756584625e-07, + "loss": 0.4686, + "step": 1599 + }, + { + "epoch": 1.5738188976377954, + "grad_norm": 9.794742584228516, + "learning_rate": 8.090214498761551e-07, + "loss": 0.3621, + "step": 1600 + }, + { + "epoch": 1.574803149606299, + "grad_norm": 10.056382179260254, + "learning_rate": 8.084175966019507e-07, + "loss": 0.5088, + "step": 1601 + }, + { + "epoch": 1.5757874015748032, + "grad_norm": 12.101970672607422, + "learning_rate": 8.078130172651368e-07, + "loss": 0.5757, + "step": 1602 + }, + { + "epoch": 1.576771653543307, + "grad_norm": 11.040324211120605, + "learning_rate": 8.072077132967193e-07, + "loss": 0.5269, + "step": 1603 + }, + { + "epoch": 1.577755905511811, + "grad_norm": 14.213335037231445, + "learning_rate": 8.066016861294198e-07, + "loss": 0.7907, + "step": 1604 + }, + { + "epoch": 1.5787401574803148, + "grad_norm": 12.529051780700684, + "learning_rate": 8.059949371976711e-07, + "loss": 0.5693, + "step": 1605 + }, + { + "epoch": 1.579724409448819, + "grad_norm": 8.595606803894043, + "learning_rate": 8.053874679376147e-07, + "loss": 0.4048, + "step": 1606 + }, + { + "epoch": 1.5807086614173227, + "grad_norm": 9.17170524597168, + "learning_rate": 8.047792797870968e-07, + "loss": 0.4727, + "step": 1607 + }, + { + "epoch": 1.5816929133858268, + "grad_norm": 10.802497863769531, + "learning_rate": 8.041703741856657e-07, + "loss": 0.5753, + "step": 1608 + }, + { + "epoch": 1.5826771653543306, + "grad_norm": 9.498733520507812, + "learning_rate": 8.035607525745672e-07, + "loss": 0.554, + "step": 1609 + }, + { + "epoch": 1.5836614173228347, + "grad_norm": 10.088098526000977, + "learning_rate": 8.029504163967423e-07, + "loss": 0.5813, + "step": 1610 + }, + { + "epoch": 1.5846456692913384, + "grad_norm": 31.178342819213867, + "learning_rate": 8.023393670968235e-07, + "loss": 2.0048, + "step": 1611 + }, + { + "epoch": 1.5856299212598426, + "grad_norm": 20.6037540435791, + "learning_rate": 8.017276061211303e-07, + "loss": 0.9611, + "step": 1612 + }, + { + "epoch": 1.5866141732283463, + "grad_norm": 11.645776748657227, + "learning_rate": 8.011151349176681e-07, + "loss": 0.7653, + "step": 1613 + }, + { + "epoch": 1.5875984251968505, + "grad_norm": 14.203062057495117, + "learning_rate": 8.005019549361222e-07, + "loss": 0.8667, + "step": 1614 + }, + { + "epoch": 1.5885826771653542, + "grad_norm": 12.750677108764648, + "learning_rate": 7.998880676278558e-07, + "loss": 0.7168, + "step": 1615 + }, + { + "epoch": 1.5895669291338583, + "grad_norm": 9.95725154876709, + "learning_rate": 7.992734744459068e-07, + "loss": 0.6818, + "step": 1616 + }, + { + "epoch": 1.590551181102362, + "grad_norm": 12.161149024963379, + "learning_rate": 7.986581768449833e-07, + "loss": 0.8315, + "step": 1617 + }, + { + "epoch": 1.5915354330708662, + "grad_norm": 17.93672752380371, + "learning_rate": 7.98042176281461e-07, + "loss": 0.9919, + "step": 1618 + }, + { + "epoch": 1.59251968503937, + "grad_norm": 17.684009552001953, + "learning_rate": 7.974254742133795e-07, + "loss": 0.7954, + "step": 1619 + }, + { + "epoch": 1.593503937007874, + "grad_norm": 19.807653427124023, + "learning_rate": 7.968080721004386e-07, + "loss": 0.9298, + "step": 1620 + }, + { + "epoch": 1.594488188976378, + "grad_norm": 9.452728271484375, + "learning_rate": 7.961899714039954e-07, + "loss": 0.3301, + "step": 1621 + }, + { + "epoch": 1.595472440944882, + "grad_norm": 11.764351844787598, + "learning_rate": 7.955711735870601e-07, + "loss": 0.468, + "step": 1622 + }, + { + "epoch": 1.5964566929133859, + "grad_norm": 18.4998836517334, + "learning_rate": 7.949516801142932e-07, + "loss": 1.0635, + "step": 1623 + }, + { + "epoch": 1.5974409448818898, + "grad_norm": 11.700940132141113, + "learning_rate": 7.943314924520017e-07, + "loss": 0.6376, + "step": 1624 + }, + { + "epoch": 1.5984251968503937, + "grad_norm": 12.373494148254395, + "learning_rate": 7.93710612068136e-07, + "loss": 0.8462, + "step": 1625 + }, + { + "epoch": 1.5994094488188977, + "grad_norm": 11.433577537536621, + "learning_rate": 7.930890404322857e-07, + "loss": 0.6499, + "step": 1626 + }, + { + "epoch": 1.6003937007874016, + "grad_norm": 11.250558853149414, + "learning_rate": 7.924667790156766e-07, + "loss": 0.5233, + "step": 1627 + }, + { + "epoch": 1.6013779527559056, + "grad_norm": 13.41759204864502, + "learning_rate": 7.918438292911674e-07, + "loss": 0.5169, + "step": 1628 + }, + { + "epoch": 1.6023622047244095, + "grad_norm": 17.72492218017578, + "learning_rate": 7.912201927332461e-07, + "loss": 0.8332, + "step": 1629 + }, + { + "epoch": 1.6033464566929134, + "grad_norm": 6.448797225952148, + "learning_rate": 7.905958708180258e-07, + "loss": 0.1964, + "step": 1630 + }, + { + "epoch": 1.6043307086614174, + "grad_norm": 17.034406661987305, + "learning_rate": 7.899708650232426e-07, + "loss": 0.914, + "step": 1631 + }, + { + "epoch": 1.6053149606299213, + "grad_norm": 11.307120323181152, + "learning_rate": 7.893451768282508e-07, + "loss": 0.4956, + "step": 1632 + }, + { + "epoch": 1.6062992125984252, + "grad_norm": 9.852806091308594, + "learning_rate": 7.8871880771402e-07, + "loss": 0.5823, + "step": 1633 + }, + { + "epoch": 1.6072834645669292, + "grad_norm": 13.381124496459961, + "learning_rate": 7.880917591631312e-07, + "loss": 0.9373, + "step": 1634 + }, + { + "epoch": 1.608267716535433, + "grad_norm": 14.188089370727539, + "learning_rate": 7.874640326597741e-07, + "loss": 0.879, + "step": 1635 + }, + { + "epoch": 1.609251968503937, + "grad_norm": 11.136621475219727, + "learning_rate": 7.868356296897431e-07, + "loss": 0.8167, + "step": 1636 + }, + { + "epoch": 1.610236220472441, + "grad_norm": 19.828542709350586, + "learning_rate": 7.862065517404334e-07, + "loss": 0.8329, + "step": 1637 + }, + { + "epoch": 1.611220472440945, + "grad_norm": 10.303707122802734, + "learning_rate": 7.855768003008376e-07, + "loss": 0.5512, + "step": 1638 + }, + { + "epoch": 1.6122047244094488, + "grad_norm": 11.766749382019043, + "learning_rate": 7.849463768615432e-07, + "loss": 0.6024, + "step": 1639 + }, + { + "epoch": 1.6131889763779528, + "grad_norm": 15.314605712890625, + "learning_rate": 7.843152829147278e-07, + "loss": 0.966, + "step": 1640 + }, + { + "epoch": 1.6141732283464567, + "grad_norm": 11.736525535583496, + "learning_rate": 7.836835199541562e-07, + "loss": 0.6611, + "step": 1641 + }, + { + "epoch": 1.6151574803149606, + "grad_norm": 4.6410369873046875, + "learning_rate": 7.830510894751764e-07, + "loss": 0.1169, + "step": 1642 + }, + { + "epoch": 1.6161417322834646, + "grad_norm": 9.750079154968262, + "learning_rate": 7.824179929747168e-07, + "loss": 0.4613, + "step": 1643 + }, + { + "epoch": 1.6171259842519685, + "grad_norm": 17.277965545654297, + "learning_rate": 7.817842319512818e-07, + "loss": 1.3485, + "step": 1644 + }, + { + "epoch": 1.6181102362204725, + "grad_norm": 10.674585342407227, + "learning_rate": 7.811498079049492e-07, + "loss": 0.4776, + "step": 1645 + }, + { + "epoch": 1.6190944881889764, + "grad_norm": 14.96063232421875, + "learning_rate": 7.805147223373658e-07, + "loss": 1.061, + "step": 1646 + }, + { + "epoch": 1.6200787401574803, + "grad_norm": 8.888649940490723, + "learning_rate": 7.798789767517444e-07, + "loss": 0.4102, + "step": 1647 + }, + { + "epoch": 1.6210629921259843, + "grad_norm": 11.162688255310059, + "learning_rate": 7.792425726528596e-07, + "loss": 0.5991, + "step": 1648 + }, + { + "epoch": 1.6220472440944882, + "grad_norm": 12.890787124633789, + "learning_rate": 7.786055115470454e-07, + "loss": 0.9039, + "step": 1649 + }, + { + "epoch": 1.6230314960629921, + "grad_norm": 11.972373962402344, + "learning_rate": 7.779677949421902e-07, + "loss": 0.7915, + "step": 1650 + }, + { + "epoch": 1.624015748031496, + "grad_norm": 9.724061965942383, + "learning_rate": 7.77329424347734e-07, + "loss": 0.5604, + "step": 1651 + }, + { + "epoch": 1.625, + "grad_norm": 8.689451217651367, + "learning_rate": 7.766904012746653e-07, + "loss": 0.2933, + "step": 1652 + }, + { + "epoch": 1.625984251968504, + "grad_norm": 10.402259826660156, + "learning_rate": 7.760507272355165e-07, + "loss": 0.6606, + "step": 1653 + }, + { + "epoch": 1.6269685039370079, + "grad_norm": 11.47604751586914, + "learning_rate": 7.754104037443609e-07, + "loss": 0.6412, + "step": 1654 + }, + { + "epoch": 1.6279527559055118, + "grad_norm": 9.512560844421387, + "learning_rate": 7.747694323168091e-07, + "loss": 0.5141, + "step": 1655 + }, + { + "epoch": 1.6289370078740157, + "grad_norm": 14.507397651672363, + "learning_rate": 7.741278144700052e-07, + "loss": 1.0262, + "step": 1656 + }, + { + "epoch": 1.6299212598425197, + "grad_norm": 10.463685989379883, + "learning_rate": 7.734855517226234e-07, + "loss": 0.535, + "step": 1657 + }, + { + "epoch": 1.6309055118110236, + "grad_norm": 13.42270565032959, + "learning_rate": 7.728426455948644e-07, + "loss": 0.6027, + "step": 1658 + }, + { + "epoch": 1.6318897637795275, + "grad_norm": 9.59408950805664, + "learning_rate": 7.721990976084517e-07, + "loss": 0.6127, + "step": 1659 + }, + { + "epoch": 1.6328740157480315, + "grad_norm": 17.450523376464844, + "learning_rate": 7.715549092866282e-07, + "loss": 1.2897, + "step": 1660 + }, + { + "epoch": 1.6338582677165354, + "grad_norm": 8.003664016723633, + "learning_rate": 7.70910082154152e-07, + "loss": 0.3319, + "step": 1661 + }, + { + "epoch": 1.6348425196850394, + "grad_norm": 11.838500022888184, + "learning_rate": 7.702646177372938e-07, + "loss": 1.0125, + "step": 1662 + }, + { + "epoch": 1.6358267716535433, + "grad_norm": 19.441804885864258, + "learning_rate": 7.696185175638325e-07, + "loss": 1.0042, + "step": 1663 + }, + { + "epoch": 1.6368110236220472, + "grad_norm": 13.763493537902832, + "learning_rate": 7.689717831630513e-07, + "loss": 0.6483, + "step": 1664 + }, + { + "epoch": 1.6377952755905512, + "grad_norm": 14.2445068359375, + "learning_rate": 7.683244160657353e-07, + "loss": 0.6933, + "step": 1665 + }, + { + "epoch": 1.638779527559055, + "grad_norm": 5.94786262512207, + "learning_rate": 7.676764178041668e-07, + "loss": 0.2073, + "step": 1666 + }, + { + "epoch": 1.639763779527559, + "grad_norm": 9.012726783752441, + "learning_rate": 7.670277899121223e-07, + "loss": 0.3838, + "step": 1667 + }, + { + "epoch": 1.640748031496063, + "grad_norm": 11.752018928527832, + "learning_rate": 7.663785339248683e-07, + "loss": 0.6676, + "step": 1668 + }, + { + "epoch": 1.641732283464567, + "grad_norm": 11.327059745788574, + "learning_rate": 7.657286513791578e-07, + "loss": 0.5122, + "step": 1669 + }, + { + "epoch": 1.6427165354330708, + "grad_norm": 12.427508354187012, + "learning_rate": 7.650781438132276e-07, + "loss": 0.8016, + "step": 1670 + }, + { + "epoch": 1.6437007874015748, + "grad_norm": 8.388118743896484, + "learning_rate": 7.644270127667932e-07, + "loss": 0.3227, + "step": 1671 + }, + { + "epoch": 1.6446850393700787, + "grad_norm": 3.641969919204712, + "learning_rate": 7.637752597810456e-07, + "loss": 0.1016, + "step": 1672 + }, + { + "epoch": 1.6456692913385826, + "grad_norm": 7.827102184295654, + "learning_rate": 7.631228863986491e-07, + "loss": 0.1843, + "step": 1673 + }, + { + "epoch": 1.6466535433070866, + "grad_norm": 11.204242706298828, + "learning_rate": 7.624698941637352e-07, + "loss": 0.4915, + "step": 1674 + }, + { + "epoch": 1.6476377952755905, + "grad_norm": 12.071660995483398, + "learning_rate": 7.618162846219006e-07, + "loss": 0.8414, + "step": 1675 + }, + { + "epoch": 1.6486220472440944, + "grad_norm": 10.194080352783203, + "learning_rate": 7.611620593202037e-07, + "loss": 0.493, + "step": 1676 + }, + { + "epoch": 1.6496062992125984, + "grad_norm": 12.145285606384277, + "learning_rate": 7.605072198071592e-07, + "loss": 0.9144, + "step": 1677 + }, + { + "epoch": 1.6505905511811023, + "grad_norm": 24.41216468811035, + "learning_rate": 7.598517676327368e-07, + "loss": 1.6354, + "step": 1678 + }, + { + "epoch": 1.6515748031496063, + "grad_norm": 23.160465240478516, + "learning_rate": 7.591957043483555e-07, + "loss": 2.1757, + "step": 1679 + }, + { + "epoch": 1.6525590551181102, + "grad_norm": 10.578500747680664, + "learning_rate": 7.585390315068812e-07, + "loss": 0.6866, + "step": 1680 + }, + { + "epoch": 1.6535433070866141, + "grad_norm": 10.035861015319824, + "learning_rate": 7.578817506626224e-07, + "loss": 0.4759, + "step": 1681 + }, + { + "epoch": 1.654527559055118, + "grad_norm": 9.562938690185547, + "learning_rate": 7.572238633713268e-07, + "loss": 0.4671, + "step": 1682 + }, + { + "epoch": 1.655511811023622, + "grad_norm": 9.952853202819824, + "learning_rate": 7.565653711901778e-07, + "loss": 0.474, + "step": 1683 + }, + { + "epoch": 1.655511811023622, + "eval_Qnli-dev_cosine_accuracy": 0.662109375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8541384935379028, + "eval_Qnli-dev_cosine_ap": 0.6939321999394883, + "eval_Qnli-dev_cosine_f1": 0.6912, + "eval_Qnli-dev_cosine_f1_threshold": 0.748054027557373, + "eval_Qnli-dev_cosine_precision": 0.5552699228791774, + "eval_Qnli-dev_cosine_recall": 0.9152542372881356, + "eval_Qnli-dev_dot_accuracy": 0.662109375, + "eval_Qnli-dev_dot_accuracy_threshold": 652.1033935546875, + "eval_Qnli-dev_dot_ap": 0.6900299415325823, + "eval_Qnli-dev_dot_f1": 0.6900958466453675, + "eval_Qnli-dev_dot_f1_threshold": 568.281982421875, + "eval_Qnli-dev_dot_precision": 0.5538461538461539, + "eval_Qnli-dev_dot_recall": 0.9152542372881356, + "eval_Qnli-dev_euclidean_accuracy": 0.66015625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 14.833984375, + "eval_Qnli-dev_euclidean_ap": 0.6946097416292264, + "eval_Qnli-dev_euclidean_f1": 0.6921850079744816, + "eval_Qnli-dev_euclidean_f1_threshold": 19.65032196044922, + "eval_Qnli-dev_euclidean_precision": 0.5549872122762148, + "eval_Qnli-dev_euclidean_recall": 0.9194915254237288, + "eval_Qnli-dev_manhattan_accuracy": 0.673828125, + "eval_Qnli-dev_manhattan_accuracy_threshold": 335.0054016113281, + "eval_Qnli-dev_manhattan_ap": 0.7052862991616473, + "eval_Qnli-dev_manhattan_f1": 0.6913183279742765, + "eval_Qnli-dev_manhattan_f1_threshold": 413.606201171875, + "eval_Qnli-dev_manhattan_precision": 0.5569948186528497, + "eval_Qnli-dev_manhattan_recall": 0.9110169491525424, + "eval_Qnli-dev_max_accuracy": 0.673828125, + "eval_Qnli-dev_max_accuracy_threshold": 652.1033935546875, + "eval_Qnli-dev_max_ap": 0.7052862991616473, + "eval_Qnli-dev_max_f1": 0.6921850079744816, + "eval_Qnli-dev_max_f1_threshold": 568.281982421875, + "eval_Qnli-dev_max_precision": 0.5569948186528497, + "eval_Qnli-dev_max_recall": 0.9194915254237288, + "eval_allNLI-dev_cosine_accuracy": 0.693359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9193407297134399, + "eval_allNLI-dev_cosine_ap": 0.5107511056200436, + "eval_allNLI-dev_cosine_f1": 0.5541666666666667, + "eval_allNLI-dev_cosine_f1_threshold": 0.8247472047805786, + "eval_allNLI-dev_cosine_precision": 0.43322475570032576, + "eval_allNLI-dev_cosine_recall": 0.7687861271676301, + "eval_allNLI-dev_dot_accuracy": 0.6953125, + "eval_allNLI-dev_dot_accuracy_threshold": 701.783447265625, + "eval_allNLI-dev_dot_ap": 0.5115874668538737, + "eval_allNLI-dev_dot_f1": 0.551440329218107, + "eval_allNLI-dev_dot_f1_threshold": 627.7216186523438, + "eval_allNLI-dev_dot_precision": 0.4281150159744409, + "eval_allNLI-dev_dot_recall": 0.7745664739884393, + "eval_allNLI-dev_euclidean_accuracy": 0.693359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.089791297912598, + "eval_allNLI-dev_euclidean_ap": 0.5111656459885842, + "eval_allNLI-dev_euclidean_f1": 0.5541666666666667, + "eval_allNLI-dev_euclidean_f1_threshold": 16.34630584716797, + "eval_allNLI-dev_euclidean_precision": 0.43322475570032576, + "eval_allNLI-dev_euclidean_recall": 0.7687861271676301, + "eval_allNLI-dev_manhattan_accuracy": 0.69140625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 234.15560913085938, + "eval_allNLI-dev_manhattan_ap": 0.5103208400600096, + "eval_allNLI-dev_manhattan_f1": 0.5557809330628803, + "eval_allNLI-dev_manhattan_f1_threshold": 347.42694091796875, + "eval_allNLI-dev_manhattan_precision": 0.428125, + "eval_allNLI-dev_manhattan_recall": 0.791907514450867, + "eval_allNLI-dev_max_accuracy": 0.6953125, + "eval_allNLI-dev_max_accuracy_threshold": 701.783447265625, + "eval_allNLI-dev_max_ap": 0.5115874668538737, + "eval_allNLI-dev_max_f1": 0.5557809330628803, + "eval_allNLI-dev_max_f1_threshold": 627.7216186523438, + "eval_allNLI-dev_max_precision": 0.43322475570032576, + "eval_allNLI-dev_max_recall": 0.791907514450867, + "eval_loss": 1.046770691871643, + "eval_runtime": 54.2587, + "eval_samples_per_second": 30.668, + "eval_sequential_score": 0.7052862991616473, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.6387787653991626, + "eval_sts-test_pearson_dot": 0.6375419066754788, + "eval_sts-test_pearson_euclidean": 0.6534286404135119, + "eval_sts-test_pearson_manhattan": 0.6562338816714135, + "eval_sts-test_pearson_max": 0.6562338816714135, + "eval_sts-test_spearman_cosine": 0.6501734447586552, + "eval_sts-test_spearman_dot": 0.6498578588633231, + "eval_sts-test_spearman_euclidean": 0.650259948946925, + "eval_sts-test_spearman_manhattan": 0.6551447281328417, + "eval_sts-test_spearman_max": 0.6551447281328417, + "step": 1683 + }, + { + "epoch": 1.656496062992126, + "grad_norm": 7.453148365020752, + "learning_rate": 7.559062756777898e-07, + "loss": 0.3317, + "step": 1684 + }, + { + "epoch": 1.65748031496063, + "grad_norm": 14.089736938476562, + "learning_rate": 7.552465783942058e-07, + "loss": 0.6944, + "step": 1685 + }, + { + "epoch": 1.6584645669291338, + "grad_norm": 6.532533168792725, + "learning_rate": 7.545862809008931e-07, + "loss": 0.2652, + "step": 1686 + }, + { + "epoch": 1.659448818897638, + "grad_norm": 9.175405502319336, + "learning_rate": 7.539253847607397e-07, + "loss": 0.4397, + "step": 1687 + }, + { + "epoch": 1.6604330708661417, + "grad_norm": 13.5653076171875, + "learning_rate": 7.532638915380503e-07, + "loss": 1.0538, + "step": 1688 + }, + { + "epoch": 1.6614173228346458, + "grad_norm": 7.647918224334717, + "learning_rate": 7.526018027985434e-07, + "loss": 0.3413, + "step": 1689 + }, + { + "epoch": 1.6624015748031495, + "grad_norm": 9.469629287719727, + "learning_rate": 7.519391201093462e-07, + "loss": 0.4697, + "step": 1690 + }, + { + "epoch": 1.6633858267716537, + "grad_norm": 9.352253913879395, + "learning_rate": 7.512758450389925e-07, + "loss": 0.4027, + "step": 1691 + }, + { + "epoch": 1.6643700787401574, + "grad_norm": 10.696208000183105, + "learning_rate": 7.50611979157418e-07, + "loss": 0.5004, + "step": 1692 + }, + { + "epoch": 1.6653543307086616, + "grad_norm": 8.486995697021484, + "learning_rate": 7.499475240359567e-07, + "loss": 0.402, + "step": 1693 + }, + { + "epoch": 1.6663385826771653, + "grad_norm": 8.808696746826172, + "learning_rate": 7.492824812473376e-07, + "loss": 0.336, + "step": 1694 + }, + { + "epoch": 1.6673228346456694, + "grad_norm": 10.125554084777832, + "learning_rate": 7.4861685236568e-07, + "loss": 0.4602, + "step": 1695 + }, + { + "epoch": 1.6683070866141732, + "grad_norm": 13.361590385437012, + "learning_rate": 7.479506389664914e-07, + "loss": 0.8332, + "step": 1696 + }, + { + "epoch": 1.6692913385826773, + "grad_norm": 17.709083557128906, + "learning_rate": 7.47283842626662e-07, + "loss": 0.701, + "step": 1697 + }, + { + "epoch": 1.670275590551181, + "grad_norm": 8.722396850585938, + "learning_rate": 7.466164649244625e-07, + "loss": 0.2903, + "step": 1698 + }, + { + "epoch": 1.6712598425196852, + "grad_norm": 9.08720588684082, + "learning_rate": 7.45948507439539e-07, + "loss": 0.3562, + "step": 1699 + }, + { + "epoch": 1.672244094488189, + "grad_norm": 8.78695011138916, + "learning_rate": 7.452799717529102e-07, + "loss": 0.4957, + "step": 1700 + }, + { + "epoch": 1.673228346456693, + "grad_norm": 11.732802391052246, + "learning_rate": 7.446108594469634e-07, + "loss": 0.5276, + "step": 1701 + }, + { + "epoch": 1.6742125984251968, + "grad_norm": 14.943826675415039, + "learning_rate": 7.439411721054508e-07, + "loss": 1.0879, + "step": 1702 + }, + { + "epoch": 1.675196850393701, + "grad_norm": 10.825159072875977, + "learning_rate": 7.432709113134856e-07, + "loss": 0.4225, + "step": 1703 + }, + { + "epoch": 1.6761811023622046, + "grad_norm": 14.165266990661621, + "learning_rate": 7.42600078657538e-07, + "loss": 0.7649, + "step": 1704 + }, + { + "epoch": 1.6771653543307088, + "grad_norm": 13.455033302307129, + "learning_rate": 7.419286757254324e-07, + "loss": 0.8458, + "step": 1705 + }, + { + "epoch": 1.6781496062992125, + "grad_norm": 8.739083290100098, + "learning_rate": 7.412567041063424e-07, + "loss": 0.377, + "step": 1706 + }, + { + "epoch": 1.6791338582677167, + "grad_norm": 11.673897743225098, + "learning_rate": 7.40584165390788e-07, + "loss": 0.4533, + "step": 1707 + }, + { + "epoch": 1.6801181102362204, + "grad_norm": 12.348942756652832, + "learning_rate": 7.399110611706316e-07, + "loss": 0.6464, + "step": 1708 + }, + { + "epoch": 1.6811023622047245, + "grad_norm": 8.441234588623047, + "learning_rate": 7.392373930390738e-07, + "loss": 0.3329, + "step": 1709 + }, + { + "epoch": 1.6820866141732282, + "grad_norm": 9.699177742004395, + "learning_rate": 7.3856316259065e-07, + "loss": 0.5933, + "step": 1710 + }, + { + "epoch": 1.6830708661417324, + "grad_norm": 9.690885543823242, + "learning_rate": 7.378883714212268e-07, + "loss": 0.2693, + "step": 1711 + }, + { + "epoch": 1.6840551181102361, + "grad_norm": 9.51298713684082, + "learning_rate": 7.372130211279977e-07, + "loss": 0.3935, + "step": 1712 + }, + { + "epoch": 1.6850393700787403, + "grad_norm": 21.50560760498047, + "learning_rate": 7.365371133094799e-07, + "loss": 0.8808, + "step": 1713 + }, + { + "epoch": 1.686023622047244, + "grad_norm": 8.320510864257812, + "learning_rate": 7.3586064956551e-07, + "loss": 0.3107, + "step": 1714 + }, + { + "epoch": 1.6870078740157481, + "grad_norm": 13.256457328796387, + "learning_rate": 7.351836314972408e-07, + "loss": 0.7207, + "step": 1715 + }, + { + "epoch": 1.6879921259842519, + "grad_norm": 9.8543119430542, + "learning_rate": 7.345060607071365e-07, + "loss": 0.2448, + "step": 1716 + }, + { + "epoch": 1.688976377952756, + "grad_norm": 16.452669143676758, + "learning_rate": 7.338279387989703e-07, + "loss": 0.5973, + "step": 1717 + }, + { + "epoch": 1.6899606299212597, + "grad_norm": 12.07260513305664, + "learning_rate": 7.331492673778193e-07, + "loss": 0.7099, + "step": 1718 + }, + { + "epoch": 1.6909448818897639, + "grad_norm": 12.174388885498047, + "learning_rate": 7.324700480500615e-07, + "loss": 0.7829, + "step": 1719 + }, + { + "epoch": 1.6919291338582676, + "grad_norm": 17.407093048095703, + "learning_rate": 7.317902824233719e-07, + "loss": 1.0199, + "step": 1720 + }, + { + "epoch": 1.6929133858267718, + "grad_norm": 11.410693168640137, + "learning_rate": 7.311099721067183e-07, + "loss": 0.5131, + "step": 1721 + }, + { + "epoch": 1.6938976377952755, + "grad_norm": 10.559515953063965, + "learning_rate": 7.304291187103578e-07, + "loss": 0.3552, + "step": 1722 + }, + { + "epoch": 1.6948818897637796, + "grad_norm": 7.444983005523682, + "learning_rate": 7.297477238458333e-07, + "loss": 0.2427, + "step": 1723 + }, + { + "epoch": 1.6958661417322833, + "grad_norm": 10.84617805480957, + "learning_rate": 7.290657891259688e-07, + "loss": 0.5793, + "step": 1724 + }, + { + "epoch": 1.6968503937007875, + "grad_norm": 11.741364479064941, + "learning_rate": 7.283833161648664e-07, + "loss": 0.6017, + "step": 1725 + }, + { + "epoch": 1.6978346456692912, + "grad_norm": 12.906566619873047, + "learning_rate": 7.277003065779022e-07, + "loss": 0.868, + "step": 1726 + }, + { + "epoch": 1.6988188976377954, + "grad_norm": 9.38211441040039, + "learning_rate": 7.270167619817222e-07, + "loss": 0.437, + "step": 1727 + }, + { + "epoch": 1.699803149606299, + "grad_norm": 12.161310195922852, + "learning_rate": 7.26332683994239e-07, + "loss": 0.6072, + "step": 1728 + }, + { + "epoch": 1.7007874015748032, + "grad_norm": 9.744953155517578, + "learning_rate": 7.256480742346278e-07, + "loss": 0.4196, + "step": 1729 + }, + { + "epoch": 1.701771653543307, + "grad_norm": 11.653627395629883, + "learning_rate": 7.249629343233221e-07, + "loss": 0.5175, + "step": 1730 + }, + { + "epoch": 1.702755905511811, + "grad_norm": 7.9525628089904785, + "learning_rate": 7.242772658820105e-07, + "loss": 0.3439, + "step": 1731 + }, + { + "epoch": 1.7037401574803148, + "grad_norm": 8.413471221923828, + "learning_rate": 7.235910705336327e-07, + "loss": 0.3967, + "step": 1732 + }, + { + "epoch": 1.704724409448819, + "grad_norm": 15.535460472106934, + "learning_rate": 7.229043499023753e-07, + "loss": 0.6726, + "step": 1733 + }, + { + "epoch": 1.7057086614173227, + "grad_norm": 23.099462509155273, + "learning_rate": 7.222171056136683e-07, + "loss": 1.0673, + "step": 1734 + }, + { + "epoch": 1.7066929133858268, + "grad_norm": 20.861698150634766, + "learning_rate": 7.215293392941814e-07, + "loss": 1.5876, + "step": 1735 + }, + { + "epoch": 1.7076771653543306, + "grad_norm": 6.357061386108398, + "learning_rate": 7.208410525718195e-07, + "loss": 0.3725, + "step": 1736 + }, + { + "epoch": 1.7086614173228347, + "grad_norm": 9.408873558044434, + "learning_rate": 7.201522470757196e-07, + "loss": 0.644, + "step": 1737 + }, + { + "epoch": 1.7096456692913384, + "grad_norm": 13.452515602111816, + "learning_rate": 7.194629244362462e-07, + "loss": 0.7511, + "step": 1738 + }, + { + "epoch": 1.7106299212598426, + "grad_norm": 11.06453800201416, + "learning_rate": 7.187730862849886e-07, + "loss": 0.4075, + "step": 1739 + }, + { + "epoch": 1.7116141732283463, + "grad_norm": 7.75278377532959, + "learning_rate": 7.180827342547556e-07, + "loss": 0.325, + "step": 1740 + }, + { + "epoch": 1.7125984251968505, + "grad_norm": 18.88231086730957, + "learning_rate": 7.173918699795726e-07, + "loss": 0.6408, + "step": 1741 + }, + { + "epoch": 1.7135826771653542, + "grad_norm": 11.604063987731934, + "learning_rate": 7.167004950946774e-07, + "loss": 0.5693, + "step": 1742 + }, + { + "epoch": 1.7145669291338583, + "grad_norm": 12.89757251739502, + "learning_rate": 7.160086112365163e-07, + "loss": 0.755, + "step": 1743 + }, + { + "epoch": 1.715551181102362, + "grad_norm": 9.268168449401855, + "learning_rate": 7.153162200427404e-07, + "loss": 0.3953, + "step": 1744 + }, + { + "epoch": 1.7165354330708662, + "grad_norm": 7.196349620819092, + "learning_rate": 7.146233231522018e-07, + "loss": 0.2483, + "step": 1745 + }, + { + "epoch": 1.71751968503937, + "grad_norm": 18.42107582092285, + "learning_rate": 7.139299222049493e-07, + "loss": 0.8284, + "step": 1746 + }, + { + "epoch": 1.718503937007874, + "grad_norm": 16.944561004638672, + "learning_rate": 7.132360188422249e-07, + "loss": 0.8637, + "step": 1747 + }, + { + "epoch": 1.719488188976378, + "grad_norm": 15.024608612060547, + "learning_rate": 7.125416147064597e-07, + "loss": 0.809, + "step": 1748 + }, + { + "epoch": 1.720472440944882, + "grad_norm": 10.315958976745605, + "learning_rate": 7.118467114412702e-07, + "loss": 0.4039, + "step": 1749 + }, + { + "epoch": 1.7214566929133859, + "grad_norm": 14.782455444335938, + "learning_rate": 7.111513106914544e-07, + "loss": 0.7835, + "step": 1750 + }, + { + "epoch": 1.7224409448818898, + "grad_norm": 7.03788423538208, + "learning_rate": 7.104554141029874e-07, + "loss": 0.2731, + "step": 1751 + }, + { + "epoch": 1.7234251968503937, + "grad_norm": 17.06793785095215, + "learning_rate": 7.097590233230186e-07, + "loss": 1.1285, + "step": 1752 + }, + { + "epoch": 1.7244094488188977, + "grad_norm": 12.640857696533203, + "learning_rate": 7.090621399998665e-07, + "loss": 0.8982, + "step": 1753 + }, + { + "epoch": 1.7253937007874016, + "grad_norm": 10.769369125366211, + "learning_rate": 7.083647657830155e-07, + "loss": 0.5333, + "step": 1754 + }, + { + "epoch": 1.7263779527559056, + "grad_norm": 4.927238941192627, + "learning_rate": 7.076669023231126e-07, + "loss": 0.1284, + "step": 1755 + }, + { + "epoch": 1.7273622047244095, + "grad_norm": 12.758343696594238, + "learning_rate": 7.069685512719619e-07, + "loss": 0.6596, + "step": 1756 + }, + { + "epoch": 1.7283464566929134, + "grad_norm": 9.049027442932129, + "learning_rate": 7.06269714282522e-07, + "loss": 0.3537, + "step": 1757 + }, + { + "epoch": 1.7293307086614174, + "grad_norm": 8.433937072753906, + "learning_rate": 7.055703930089018e-07, + "loss": 0.4507, + "step": 1758 + }, + { + "epoch": 1.7303149606299213, + "grad_norm": 16.079524993896484, + "learning_rate": 7.048705891063562e-07, + "loss": 0.7807, + "step": 1759 + }, + { + "epoch": 1.7312992125984252, + "grad_norm": 9.565560340881348, + "learning_rate": 7.04170304231283e-07, + "loss": 0.3956, + "step": 1760 + }, + { + "epoch": 1.7322834645669292, + "grad_norm": 5.272612571716309, + "learning_rate": 7.034695400412177e-07, + "loss": 0.1616, + "step": 1761 + }, + { + "epoch": 1.733267716535433, + "grad_norm": 9.687300682067871, + "learning_rate": 7.027682981948306e-07, + "loss": 0.4198, + "step": 1762 + }, + { + "epoch": 1.734251968503937, + "grad_norm": 11.882638931274414, + "learning_rate": 7.020665803519229e-07, + "loss": 0.5712, + "step": 1763 + }, + { + "epoch": 1.735236220472441, + "grad_norm": 12.776277542114258, + "learning_rate": 7.013643881734222e-07, + "loss": 0.8004, + "step": 1764 + }, + { + "epoch": 1.736220472440945, + "grad_norm": 10.282646179199219, + "learning_rate": 7.006617233213787e-07, + "loss": 0.493, + "step": 1765 + }, + { + "epoch": 1.7372047244094488, + "grad_norm": 9.987165451049805, + "learning_rate": 6.999585874589616e-07, + "loss": 0.3876, + "step": 1766 + }, + { + "epoch": 1.7381889763779528, + "grad_norm": 8.786449432373047, + "learning_rate": 6.992549822504548e-07, + "loss": 0.2701, + "step": 1767 + }, + { + "epoch": 1.7391732283464567, + "grad_norm": 17.34662437438965, + "learning_rate": 6.985509093612534e-07, + "loss": 1.1244, + "step": 1768 + }, + { + "epoch": 1.7401574803149606, + "grad_norm": 13.91609001159668, + "learning_rate": 6.978463704578589e-07, + "loss": 0.9328, + "step": 1769 + }, + { + "epoch": 1.7411417322834646, + "grad_norm": 12.426129341125488, + "learning_rate": 6.971413672078766e-07, + "loss": 0.6236, + "step": 1770 + }, + { + "epoch": 1.7421259842519685, + "grad_norm": 16.320554733276367, + "learning_rate": 6.9643590128001e-07, + "loss": 0.5158, + "step": 1771 + }, + { + "epoch": 1.7431102362204725, + "grad_norm": 7.571167469024658, + "learning_rate": 6.957299743440589e-07, + "loss": 0.2337, + "step": 1772 + }, + { + "epoch": 1.7440944881889764, + "grad_norm": 10.513776779174805, + "learning_rate": 6.950235880709131e-07, + "loss": 0.5936, + "step": 1773 + }, + { + "epoch": 1.7450787401574803, + "grad_norm": 12.005305290222168, + "learning_rate": 6.943167441325504e-07, + "loss": 0.6437, + "step": 1774 + }, + { + "epoch": 1.7460629921259843, + "grad_norm": 20.60370445251465, + "learning_rate": 6.936094442020314e-07, + "loss": 0.9568, + "step": 1775 + }, + { + "epoch": 1.7470472440944882, + "grad_norm": 12.10549545288086, + "learning_rate": 6.929016899534963e-07, + "loss": 0.5351, + "step": 1776 + }, + { + "epoch": 1.7480314960629921, + "grad_norm": 12.539772987365723, + "learning_rate": 6.921934830621603e-07, + "loss": 0.6413, + "step": 1777 + }, + { + "epoch": 1.749015748031496, + "grad_norm": 12.27780818939209, + "learning_rate": 6.914848252043106e-07, + "loss": 0.7472, + "step": 1778 + }, + { + "epoch": 1.75, + "grad_norm": 14.630345344543457, + "learning_rate": 6.907757180573013e-07, + "loss": 0.6276, + "step": 1779 + }, + { + "epoch": 1.750984251968504, + "grad_norm": 11.79141616821289, + "learning_rate": 6.9006616329955e-07, + "loss": 0.4015, + "step": 1780 + }, + { + "epoch": 1.7519685039370079, + "grad_norm": 9.032299041748047, + "learning_rate": 6.893561626105338e-07, + "loss": 0.3602, + "step": 1781 + }, + { + "epoch": 1.7529527559055118, + "grad_norm": 22.263681411743164, + "learning_rate": 6.886457176707855e-07, + "loss": 1.1816, + "step": 1782 + }, + { + "epoch": 1.7539370078740157, + "grad_norm": 11.743998527526855, + "learning_rate": 6.879348301618894e-07, + "loss": 0.4165, + "step": 1783 + }, + { + "epoch": 1.7549212598425197, + "grad_norm": 19.353025436401367, + "learning_rate": 6.872235017664769e-07, + "loss": 0.6603, + "step": 1784 + }, + { + "epoch": 1.7559055118110236, + "grad_norm": 13.100909233093262, + "learning_rate": 6.865117341682234e-07, + "loss": 0.628, + "step": 1785 + }, + { + "epoch": 1.7568897637795275, + "grad_norm": 20.460329055786133, + "learning_rate": 6.857995290518439e-07, + "loss": 0.8104, + "step": 1786 + }, + { + "epoch": 1.7578740157480315, + "grad_norm": 19.350191116333008, + "learning_rate": 6.850868881030883e-07, + "loss": 0.7918, + "step": 1787 + }, + { + "epoch": 1.7588582677165354, + "grad_norm": 11.969427108764648, + "learning_rate": 6.843738130087393e-07, + "loss": 0.6674, + "step": 1788 + }, + { + "epoch": 1.7598425196850394, + "grad_norm": 13.915922164916992, + "learning_rate": 6.836603054566059e-07, + "loss": 0.696, + "step": 1789 + }, + { + "epoch": 1.7608267716535433, + "grad_norm": 16.64994239807129, + "learning_rate": 6.829463671355216e-07, + "loss": 0.6102, + "step": 1790 + }, + { + "epoch": 1.7618110236220472, + "grad_norm": 12.086501121520996, + "learning_rate": 6.822319997353391e-07, + "loss": 0.6323, + "step": 1791 + }, + { + "epoch": 1.7627952755905512, + "grad_norm": 10.079103469848633, + "learning_rate": 6.815172049469268e-07, + "loss": 0.447, + "step": 1792 + }, + { + "epoch": 1.763779527559055, + "grad_norm": 13.176106452941895, + "learning_rate": 6.808019844621649e-07, + "loss": 0.5301, + "step": 1793 + }, + { + "epoch": 1.764763779527559, + "grad_norm": 10.617973327636719, + "learning_rate": 6.800863399739409e-07, + "loss": 0.4242, + "step": 1794 + }, + { + "epoch": 1.765748031496063, + "grad_norm": 16.50306510925293, + "learning_rate": 6.793702731761459e-07, + "loss": 1.0912, + "step": 1795 + }, + { + "epoch": 1.766732283464567, + "grad_norm": 10.760600090026855, + "learning_rate": 6.786537857636707e-07, + "loss": 0.5947, + "step": 1796 + }, + { + "epoch": 1.7677165354330708, + "grad_norm": 17.52591323852539, + "learning_rate": 6.77936879432402e-07, + "loss": 0.9659, + "step": 1797 + }, + { + "epoch": 1.7687007874015748, + "grad_norm": 10.315180778503418, + "learning_rate": 6.772195558792175e-07, + "loss": 0.3915, + "step": 1798 + }, + { + "epoch": 1.7696850393700787, + "grad_norm": 16.027626037597656, + "learning_rate": 6.765018168019825e-07, + "loss": 1.0406, + "step": 1799 + }, + { + "epoch": 1.7706692913385826, + "grad_norm": 11.071070671081543, + "learning_rate": 6.757836638995461e-07, + "loss": 0.601, + "step": 1800 + }, + { + "epoch": 1.7716535433070866, + "grad_norm": 7.411619663238525, + "learning_rate": 6.75065098871737e-07, + "loss": 0.2213, + "step": 1801 + }, + { + "epoch": 1.7726377952755905, + "grad_norm": 13.08354663848877, + "learning_rate": 6.74346123419359e-07, + "loss": 0.6773, + "step": 1802 + }, + { + "epoch": 1.7736220472440944, + "grad_norm": 14.372445106506348, + "learning_rate": 6.736267392441874e-07, + "loss": 0.5883, + "step": 1803 + }, + { + "epoch": 1.7746062992125984, + "grad_norm": 10.676488876342773, + "learning_rate": 6.729069480489652e-07, + "loss": 0.4667, + "step": 1804 + }, + { + "epoch": 1.7755905511811023, + "grad_norm": 12.284120559692383, + "learning_rate": 6.721867515373987e-07, + "loss": 0.6663, + "step": 1805 + }, + { + "epoch": 1.7765748031496063, + "grad_norm": 10.81378173828125, + "learning_rate": 6.714661514141535e-07, + "loss": 0.6057, + "step": 1806 + }, + { + "epoch": 1.7775590551181102, + "grad_norm": 12.88852310180664, + "learning_rate": 6.707451493848502e-07, + "loss": 0.8443, + "step": 1807 + }, + { + "epoch": 1.7785433070866141, + "grad_norm": 11.20924186706543, + "learning_rate": 6.700237471560614e-07, + "loss": 0.6473, + "step": 1808 + }, + { + "epoch": 1.779527559055118, + "grad_norm": 6.5453691482543945, + "learning_rate": 6.693019464353065e-07, + "loss": 0.2648, + "step": 1809 + }, + { + "epoch": 1.780511811023622, + "grad_norm": 11.30556583404541, + "learning_rate": 6.685797489310481e-07, + "loss": 0.5918, + "step": 1810 + }, + { + "epoch": 1.781496062992126, + "grad_norm": 15.30104923248291, + "learning_rate": 6.67857156352688e-07, + "loss": 0.6399, + "step": 1811 + }, + { + "epoch": 1.78248031496063, + "grad_norm": 15.234845161437988, + "learning_rate": 6.671341704105632e-07, + "loss": 0.5961, + "step": 1812 + }, + { + "epoch": 1.7834645669291338, + "grad_norm": 13.376496315002441, + "learning_rate": 6.664107928159419e-07, + "loss": 0.2637, + "step": 1813 + }, + { + "epoch": 1.784448818897638, + "grad_norm": 18.91242790222168, + "learning_rate": 6.656870252810187e-07, + "loss": 0.8332, + "step": 1814 + }, + { + "epoch": 1.7854330708661417, + "grad_norm": 6.986865043640137, + "learning_rate": 6.649628695189123e-07, + "loss": 0.3121, + "step": 1815 + }, + { + "epoch": 1.7864173228346458, + "grad_norm": 17.72435188293457, + "learning_rate": 6.642383272436589e-07, + "loss": 0.9966, + "step": 1816 + }, + { + "epoch": 1.7874015748031495, + "grad_norm": 13.330727577209473, + "learning_rate": 6.635134001702107e-07, + "loss": 0.9226, + "step": 1817 + }, + { + "epoch": 1.7883858267716537, + "grad_norm": 3.9042019844055176, + "learning_rate": 6.627880900144303e-07, + "loss": 0.1044, + "step": 1818 + }, + { + "epoch": 1.7893700787401574, + "grad_norm": 12.28075885772705, + "learning_rate": 6.620623984930869e-07, + "loss": 0.67, + "step": 1819 + }, + { + "epoch": 1.7903543307086616, + "grad_norm": 9.03951358795166, + "learning_rate": 6.613363273238527e-07, + "loss": 0.4401, + "step": 1820 + }, + { + "epoch": 1.7913385826771653, + "grad_norm": 9.525883674621582, + "learning_rate": 6.606098782252978e-07, + "loss": 0.4222, + "step": 1821 + }, + { + "epoch": 1.7923228346456694, + "grad_norm": 10.233338356018066, + "learning_rate": 6.598830529168877e-07, + "loss": 0.5033, + "step": 1822 + }, + { + "epoch": 1.7933070866141732, + "grad_norm": 13.113160133361816, + "learning_rate": 6.591558531189781e-07, + "loss": 0.7639, + "step": 1823 + }, + { + "epoch": 1.7942913385826773, + "grad_norm": 10.392827987670898, + "learning_rate": 6.584282805528106e-07, + "loss": 0.5791, + "step": 1824 + }, + { + "epoch": 1.795275590551181, + "grad_norm": 13.654716491699219, + "learning_rate": 6.577003369405099e-07, + "loss": 0.7615, + "step": 1825 + }, + { + "epoch": 1.7962598425196852, + "grad_norm": 9.4904146194458, + "learning_rate": 6.569720240050784e-07, + "loss": 0.26, + "step": 1826 + }, + { + "epoch": 1.797244094488189, + "grad_norm": 7.8181071281433105, + "learning_rate": 6.562433434703928e-07, + "loss": 0.2372, + "step": 1827 + }, + { + "epoch": 1.798228346456693, + "grad_norm": 11.520416259765625, + "learning_rate": 6.555142970612001e-07, + "loss": 0.3546, + "step": 1828 + }, + { + "epoch": 1.7992125984251968, + "grad_norm": 10.961482048034668, + "learning_rate": 6.54784886503113e-07, + "loss": 0.4989, + "step": 1829 + }, + { + "epoch": 1.800196850393701, + "grad_norm": 19.619661331176758, + "learning_rate": 6.540551135226061e-07, + "loss": 1.1705, + "step": 1830 + }, + { + "epoch": 1.8011811023622046, + "grad_norm": 20.04231071472168, + "learning_rate": 6.533249798470123e-07, + "loss": 1.1289, + "step": 1831 + }, + { + "epoch": 1.8021653543307088, + "grad_norm": 14.368165969848633, + "learning_rate": 6.525944872045175e-07, + "loss": 0.8638, + "step": 1832 + }, + { + "epoch": 1.8031496062992125, + "grad_norm": 8.188169479370117, + "learning_rate": 6.51863637324158e-07, + "loss": 0.2951, + "step": 1833 + }, + { + "epoch": 1.8041338582677167, + "grad_norm": 7.987644672393799, + "learning_rate": 6.511324319358154e-07, + "loss": 0.3667, + "step": 1834 + }, + { + "epoch": 1.8051181102362204, + "grad_norm": 11.989052772521973, + "learning_rate": 6.504008727702123e-07, + "loss": 0.5948, + "step": 1835 + }, + { + "epoch": 1.8061023622047245, + "grad_norm": 6.342445373535156, + "learning_rate": 6.496689615589096e-07, + "loss": 0.1921, + "step": 1836 + }, + { + "epoch": 1.8061023622047245, + "eval_Qnli-dev_cosine_accuracy": 0.6640625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8278301954269409, + "eval_Qnli-dev_cosine_ap": 0.6921415813364429, + "eval_Qnli-dev_cosine_f1": 0.687813021702838, + "eval_Qnli-dev_cosine_f1_threshold": 0.7467654943466187, + "eval_Qnli-dev_cosine_precision": 0.5674931129476584, + "eval_Qnli-dev_cosine_recall": 0.8728813559322034, + "eval_Qnli-dev_dot_accuracy": 0.658203125, + "eval_Qnli-dev_dot_accuracy_threshold": 638.0369262695312, + "eval_Qnli-dev_dot_ap": 0.6907455906812887, + "eval_Qnli-dev_dot_f1": 0.6901172529313233, + "eval_Qnli-dev_dot_f1_threshold": 568.925537109375, + "eval_Qnli-dev_dot_precision": 0.5706371191135734, + "eval_Qnli-dev_dot_recall": 0.8728813559322034, + "eval_Qnli-dev_euclidean_accuracy": 0.6640625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 16.232084274291992, + "eval_Qnli-dev_euclidean_ap": 0.692218626949551, + "eval_Qnli-dev_euclidean_f1": 0.6877076411960132, + "eval_Qnli-dev_euclidean_f1_threshold": 19.771434783935547, + "eval_Qnli-dev_euclidean_precision": 0.5655737704918032, + "eval_Qnli-dev_euclidean_recall": 0.8771186440677966, + "eval_Qnli-dev_manhattan_accuracy": 0.6796875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 370.8492431640625, + "eval_Qnli-dev_manhattan_ap": 0.7016599415151726, + "eval_Qnli-dev_manhattan_f1": 0.6907216494845361, + "eval_Qnli-dev_manhattan_f1_threshold": 403.894775390625, + "eval_Qnli-dev_manhattan_precision": 0.5809248554913294, + "eval_Qnli-dev_manhattan_recall": 0.8516949152542372, + "eval_Qnli-dev_max_accuracy": 0.6796875, + "eval_Qnli-dev_max_accuracy_threshold": 638.0369262695312, + "eval_Qnli-dev_max_ap": 0.7016599415151726, + "eval_Qnli-dev_max_f1": 0.6907216494845361, + "eval_Qnli-dev_max_f1_threshold": 568.925537109375, + "eval_Qnli-dev_max_precision": 0.5809248554913294, + "eval_Qnli-dev_max_recall": 0.8771186440677966, + "eval_allNLI-dev_cosine_accuracy": 0.6953125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9156147241592407, + "eval_allNLI-dev_cosine_ap": 0.5190278899962391, + "eval_allNLI-dev_cosine_f1": 0.5562632696390658, + "eval_allNLI-dev_cosine_f1_threshold": 0.83030104637146, + "eval_allNLI-dev_cosine_precision": 0.4395973154362416, + "eval_allNLI-dev_cosine_recall": 0.7572254335260116, + "eval_allNLI-dev_dot_accuracy": 0.6953125, + "eval_allNLI-dev_dot_accuracy_threshold": 709.8850708007812, + "eval_allNLI-dev_dot_ap": 0.5167210503115699, + "eval_allNLI-dev_dot_f1": 0.5520169851380042, + "eval_allNLI-dev_dot_f1_threshold": 634.1051635742188, + "eval_allNLI-dev_dot_precision": 0.436241610738255, + "eval_allNLI-dev_dot_recall": 0.7514450867052023, + "eval_allNLI-dev_euclidean_accuracy": 0.6953125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.66873550415039, + "eval_allNLI-dev_euclidean_ap": 0.5195587301363868, + "eval_allNLI-dev_euclidean_f1": 0.5562632696390658, + "eval_allNLI-dev_euclidean_f1_threshold": 16.106643676757812, + "eval_allNLI-dev_euclidean_precision": 0.4395973154362416, + "eval_allNLI-dev_euclidean_recall": 0.7572254335260116, + "eval_allNLI-dev_manhattan_accuracy": 0.69140625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 259.05010986328125, + "eval_allNLI-dev_manhattan_ap": 0.515787932229368, + "eval_allNLI-dev_manhattan_f1": 0.5493562231759657, + "eval_allNLI-dev_manhattan_f1_threshold": 338.73870849609375, + "eval_allNLI-dev_manhattan_precision": 0.43686006825938567, + "eval_allNLI-dev_manhattan_recall": 0.7398843930635838, + "eval_allNLI-dev_max_accuracy": 0.6953125, + "eval_allNLI-dev_max_accuracy_threshold": 709.8850708007812, + "eval_allNLI-dev_max_ap": 0.5195587301363868, + "eval_allNLI-dev_max_f1": 0.5562632696390658, + "eval_allNLI-dev_max_f1_threshold": 634.1051635742188, + "eval_allNLI-dev_max_precision": 0.4395973154362416, + "eval_allNLI-dev_max_recall": 0.7572254335260116, + "eval_loss": 0.9842851758003235, + "eval_runtime": 54.3651, + "eval_samples_per_second": 30.608, + "eval_sequential_score": 0.7016599415151726, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.6309265304897379, + "eval_sts-test_pearson_dot": 0.6292250868831678, + "eval_sts-test_pearson_euclidean": 0.6445377889980463, + "eval_sts-test_pearson_manhattan": 0.6504294362294498, + "eval_sts-test_pearson_max": 0.6504294362294498, + "eval_sts-test_spearman_cosine": 0.642238464347562, + "eval_sts-test_spearman_dot": 0.6404100620608937, + "eval_sts-test_spearman_euclidean": 0.6423653341736068, + "eval_sts-test_spearman_manhattan": 0.6474620105822118, + "eval_sts-test_spearman_max": 0.6474620105822118, + "step": 1836 + }, + { + "epoch": 1.8070866141732282, + "grad_norm": 8.733999252319336, + "learning_rate": 6.489367000343005e-07, + "loss": 0.3578, + "step": 1837 + }, + { + "epoch": 1.8080708661417324, + "grad_norm": 13.05897331237793, + "learning_rate": 6.482040899296081e-07, + "loss": 0.5384, + "step": 1838 + }, + { + "epoch": 1.8090551181102361, + "grad_norm": 8.08520793914795, + "learning_rate": 6.474711329788802e-07, + "loss": 0.3188, + "step": 1839 + }, + { + "epoch": 1.8100393700787403, + "grad_norm": 4.597148895263672, + "learning_rate": 6.467378309169858e-07, + "loss": 0.1175, + "step": 1840 + }, + { + "epoch": 1.811023622047244, + "grad_norm": 9.91828727722168, + "learning_rate": 6.460041854796105e-07, + "loss": 0.403, + "step": 1841 + }, + { + "epoch": 1.8120078740157481, + "grad_norm": 6.756091594696045, + "learning_rate": 6.452701984032526e-07, + "loss": 0.2802, + "step": 1842 + }, + { + "epoch": 1.8129921259842519, + "grad_norm": 17.230257034301758, + "learning_rate": 6.445358714252195e-07, + "loss": 0.8946, + "step": 1843 + }, + { + "epoch": 1.813976377952756, + "grad_norm": 14.641213417053223, + "learning_rate": 6.438012062836226e-07, + "loss": 0.678, + "step": 1844 + }, + { + "epoch": 1.8149606299212597, + "grad_norm": 10.255255699157715, + "learning_rate": 6.430662047173742e-07, + "loss": 0.4228, + "step": 1845 + }, + { + "epoch": 1.8159448818897639, + "grad_norm": 13.459108352661133, + "learning_rate": 6.423308684661825e-07, + "loss": 0.848, + "step": 1846 + }, + { + "epoch": 1.8169291338582676, + "grad_norm": 13.858319282531738, + "learning_rate": 6.415951992705482e-07, + "loss": 0.9801, + "step": 1847 + }, + { + "epoch": 1.8179133858267718, + "grad_norm": 9.967562675476074, + "learning_rate": 6.408591988717594e-07, + "loss": 0.4032, + "step": 1848 + }, + { + "epoch": 1.8188976377952755, + "grad_norm": 11.916775703430176, + "learning_rate": 6.401228690118893e-07, + "loss": 0.6389, + "step": 1849 + }, + { + "epoch": 1.8198818897637796, + "grad_norm": 13.487015724182129, + "learning_rate": 6.3938621143379e-07, + "loss": 0.4744, + "step": 1850 + }, + { + "epoch": 1.8208661417322833, + "grad_norm": 6.505801200866699, + "learning_rate": 6.386492278810894e-07, + "loss": 0.2109, + "step": 1851 + }, + { + "epoch": 1.8218503937007875, + "grad_norm": 13.570316314697266, + "learning_rate": 6.37911920098187e-07, + "loss": 0.7483, + "step": 1852 + }, + { + "epoch": 1.8228346456692912, + "grad_norm": 13.152098655700684, + "learning_rate": 6.371742898302503e-07, + "loss": 0.7438, + "step": 1853 + }, + { + "epoch": 1.8238188976377954, + "grad_norm": 13.68462085723877, + "learning_rate": 6.364363388232091e-07, + "loss": 0.6909, + "step": 1854 + }, + { + "epoch": 1.824803149606299, + "grad_norm": 10.51680850982666, + "learning_rate": 6.356980688237535e-07, + "loss": 0.5785, + "step": 1855 + }, + { + "epoch": 1.8257874015748032, + "grad_norm": 12.95541000366211, + "learning_rate": 6.349594815793275e-07, + "loss": 0.6868, + "step": 1856 + }, + { + "epoch": 1.826771653543307, + "grad_norm": 10.486471176147461, + "learning_rate": 6.34220578838127e-07, + "loss": 0.3436, + "step": 1857 + }, + { + "epoch": 1.827755905511811, + "grad_norm": 14.526586532592773, + "learning_rate": 6.334813623490939e-07, + "loss": 1.1297, + "step": 1858 + }, + { + "epoch": 1.8287401574803148, + "grad_norm": 13.752555847167969, + "learning_rate": 6.327418338619132e-07, + "loss": 0.8586, + "step": 1859 + }, + { + "epoch": 1.829724409448819, + "grad_norm": 24.5888729095459, + "learning_rate": 6.320019951270082e-07, + "loss": 1.3802, + "step": 1860 + }, + { + "epoch": 1.8307086614173227, + "grad_norm": 12.717546463012695, + "learning_rate": 6.312618478955368e-07, + "loss": 0.4151, + "step": 1861 + }, + { + "epoch": 1.8316929133858268, + "grad_norm": 9.724681854248047, + "learning_rate": 6.305213939193864e-07, + "loss": 0.3279, + "step": 1862 + }, + { + "epoch": 1.8326771653543306, + "grad_norm": 9.508359909057617, + "learning_rate": 6.297806349511711e-07, + "loss": 0.6177, + "step": 1863 + }, + { + "epoch": 1.8336614173228347, + "grad_norm": 10.602931022644043, + "learning_rate": 6.290395727442269e-07, + "loss": 0.4888, + "step": 1864 + }, + { + "epoch": 1.8346456692913384, + "grad_norm": 12.10499095916748, + "learning_rate": 6.282982090526071e-07, + "loss": 0.3599, + "step": 1865 + }, + { + "epoch": 1.8356299212598426, + "grad_norm": 7.4049272537231445, + "learning_rate": 6.27556545631079e-07, + "loss": 0.2112, + "step": 1866 + }, + { + "epoch": 1.8366141732283463, + "grad_norm": 9.28160285949707, + "learning_rate": 6.268145842351191e-07, + "loss": 0.2462, + "step": 1867 + }, + { + "epoch": 1.8375984251968505, + "grad_norm": 14.189945220947266, + "learning_rate": 6.260723266209094e-07, + "loss": 0.9246, + "step": 1868 + }, + { + "epoch": 1.8385826771653542, + "grad_norm": 10.58701229095459, + "learning_rate": 6.253297745453329e-07, + "loss": 0.5992, + "step": 1869 + }, + { + "epoch": 1.8395669291338583, + "grad_norm": 16.576274871826172, + "learning_rate": 6.245869297659694e-07, + "loss": 0.7455, + "step": 1870 + }, + { + "epoch": 1.840551181102362, + "grad_norm": 14.601115226745605, + "learning_rate": 6.238437940410921e-07, + "loss": 0.4792, + "step": 1871 + }, + { + "epoch": 1.8415354330708662, + "grad_norm": 13.090888023376465, + "learning_rate": 6.231003691296622e-07, + "loss": 0.6048, + "step": 1872 + }, + { + "epoch": 1.84251968503937, + "grad_norm": 16.715045928955078, + "learning_rate": 6.223566567913256e-07, + "loss": 1.1413, + "step": 1873 + }, + { + "epoch": 1.843503937007874, + "grad_norm": 10.588872909545898, + "learning_rate": 6.216126587864084e-07, + "loss": 0.5035, + "step": 1874 + }, + { + "epoch": 1.844488188976378, + "grad_norm": 18.341047286987305, + "learning_rate": 6.208683768759136e-07, + "loss": 0.8205, + "step": 1875 + }, + { + "epoch": 1.845472440944882, + "grad_norm": 10.638150215148926, + "learning_rate": 6.201238128215152e-07, + "loss": 0.3922, + "step": 1876 + }, + { + "epoch": 1.8464566929133859, + "grad_norm": 10.543109893798828, + "learning_rate": 6.193789683855555e-07, + "loss": 0.4792, + "step": 1877 + }, + { + "epoch": 1.8474409448818898, + "grad_norm": 7.625102996826172, + "learning_rate": 6.186338453310404e-07, + "loss": 0.2323, + "step": 1878 + }, + { + "epoch": 1.8484251968503937, + "grad_norm": 21.77037811279297, + "learning_rate": 6.178884454216355e-07, + "loss": 1.0537, + "step": 1879 + }, + { + "epoch": 1.8494094488188977, + "grad_norm": 11.93779182434082, + "learning_rate": 6.171427704216609e-07, + "loss": 0.7624, + "step": 1880 + }, + { + "epoch": 1.8503937007874016, + "grad_norm": 13.156431198120117, + "learning_rate": 6.163968220960891e-07, + "loss": 0.6861, + "step": 1881 + }, + { + "epoch": 1.8513779527559056, + "grad_norm": 10.549013137817383, + "learning_rate": 6.156506022105382e-07, + "loss": 0.4958, + "step": 1882 + }, + { + "epoch": 1.8523622047244095, + "grad_norm": 9.736396789550781, + "learning_rate": 6.1490411253127e-07, + "loss": 0.4112, + "step": 1883 + }, + { + "epoch": 1.8533464566929134, + "grad_norm": 11.155174255371094, + "learning_rate": 6.141573548251847e-07, + "loss": 0.6358, + "step": 1884 + }, + { + "epoch": 1.8543307086614174, + "grad_norm": 13.777209281921387, + "learning_rate": 6.134103308598166e-07, + "loss": 0.7354, + "step": 1885 + }, + { + "epoch": 1.8553149606299213, + "grad_norm": 13.100604057312012, + "learning_rate": 6.126630424033303e-07, + "loss": 0.7199, + "step": 1886 + }, + { + "epoch": 1.8562992125984252, + "grad_norm": 12.631733894348145, + "learning_rate": 6.119154912245167e-07, + "loss": 0.8309, + "step": 1887 + }, + { + "epoch": 1.8572834645669292, + "grad_norm": 11.478351593017578, + "learning_rate": 6.111676790927884e-07, + "loss": 0.6902, + "step": 1888 + }, + { + "epoch": 1.858267716535433, + "grad_norm": 5.461014270782471, + "learning_rate": 6.104196077781757e-07, + "loss": 0.1763, + "step": 1889 + }, + { + "epoch": 1.859251968503937, + "grad_norm": 12.026067733764648, + "learning_rate": 6.096712790513222e-07, + "loss": 0.6163, + "step": 1890 + }, + { + "epoch": 1.860236220472441, + "grad_norm": 11.317314147949219, + "learning_rate": 6.08922694683481e-07, + "loss": 0.5629, + "step": 1891 + }, + { + "epoch": 1.861220472440945, + "grad_norm": 13.000326156616211, + "learning_rate": 6.081738564465103e-07, + "loss": 0.7518, + "step": 1892 + }, + { + "epoch": 1.8622047244094488, + "grad_norm": 13.375837326049805, + "learning_rate": 6.07424766112869e-07, + "loss": 0.6893, + "step": 1893 + }, + { + "epoch": 1.8631889763779528, + "grad_norm": 8.774999618530273, + "learning_rate": 6.066754254556127e-07, + "loss": 0.3519, + "step": 1894 + }, + { + "epoch": 1.8641732283464567, + "grad_norm": 5.026043891906738, + "learning_rate": 6.059258362483897e-07, + "loss": 0.121, + "step": 1895 + }, + { + "epoch": 1.8651574803149606, + "grad_norm": 14.182245254516602, + "learning_rate": 6.051760002654364e-07, + "loss": 0.5967, + "step": 1896 + }, + { + "epoch": 1.8661417322834646, + "grad_norm": 8.58065128326416, + "learning_rate": 6.044259192815735e-07, + "loss": 0.3536, + "step": 1897 + }, + { + "epoch": 1.8671259842519685, + "grad_norm": 11.549651145935059, + "learning_rate": 6.036755950722014e-07, + "loss": 0.6354, + "step": 1898 + }, + { + "epoch": 1.8681102362204725, + "grad_norm": 12.778814315795898, + "learning_rate": 6.029250294132965e-07, + "loss": 0.6448, + "step": 1899 + }, + { + "epoch": 1.8690944881889764, + "grad_norm": 9.952596664428711, + "learning_rate": 6.021742240814067e-07, + "loss": 0.3805, + "step": 1900 + }, + { + "epoch": 1.8700787401574803, + "grad_norm": 7.297644138336182, + "learning_rate": 6.014231808536463e-07, + "loss": 0.2854, + "step": 1901 + }, + { + "epoch": 1.8710629921259843, + "grad_norm": 22.722822189331055, + "learning_rate": 6.006719015076938e-07, + "loss": 1.1512, + "step": 1902 + }, + { + "epoch": 1.8720472440944882, + "grad_norm": 10.664011001586914, + "learning_rate": 5.999203878217861e-07, + "loss": 0.3825, + "step": 1903 + }, + { + "epoch": 1.8730314960629921, + "grad_norm": 9.881097793579102, + "learning_rate": 5.99168641574715e-07, + "loss": 0.4156, + "step": 1904 + }, + { + "epoch": 1.874015748031496, + "grad_norm": 8.037322044372559, + "learning_rate": 5.984166645458224e-07, + "loss": 0.2155, + "step": 1905 + }, + { + "epoch": 1.875, + "grad_norm": 21.00752830505371, + "learning_rate": 5.976644585149967e-07, + "loss": 0.9292, + "step": 1906 + }, + { + "epoch": 1.875984251968504, + "grad_norm": 15.958588600158691, + "learning_rate": 5.969120252626684e-07, + "loss": 0.6131, + "step": 1907 + }, + { + "epoch": 1.8769685039370079, + "grad_norm": 13.447348594665527, + "learning_rate": 5.961593665698055e-07, + "loss": 0.516, + "step": 1908 + }, + { + "epoch": 1.8779527559055118, + "grad_norm": 11.870959281921387, + "learning_rate": 5.954064842179099e-07, + "loss": 0.829, + "step": 1909 + }, + { + "epoch": 1.8789370078740157, + "grad_norm": 14.993748664855957, + "learning_rate": 5.946533799890127e-07, + "loss": 0.7785, + "step": 1910 + }, + { + "epoch": 1.8799212598425197, + "grad_norm": 10.761703491210938, + "learning_rate": 5.939000556656704e-07, + "loss": 0.4304, + "step": 1911 + }, + { + "epoch": 1.8809055118110236, + "grad_norm": 11.223014831542969, + "learning_rate": 5.931465130309604e-07, + "loss": 0.4163, + "step": 1912 + }, + { + "epoch": 1.8818897637795275, + "grad_norm": 10.408675193786621, + "learning_rate": 5.923927538684764e-07, + "loss": 0.4314, + "step": 1913 + }, + { + "epoch": 1.8828740157480315, + "grad_norm": 10.93159294128418, + "learning_rate": 5.916387799623253e-07, + "loss": 0.5409, + "step": 1914 + }, + { + "epoch": 1.8838582677165354, + "grad_norm": 5.786235809326172, + "learning_rate": 5.908845930971217e-07, + "loss": 0.1406, + "step": 1915 + }, + { + "epoch": 1.8848425196850394, + "grad_norm": 10.518360137939453, + "learning_rate": 5.901301950579845e-07, + "loss": 0.4662, + "step": 1916 + }, + { + "epoch": 1.8858267716535433, + "grad_norm": 12.29217529296875, + "learning_rate": 5.893755876305328e-07, + "loss": 0.5706, + "step": 1917 + }, + { + "epoch": 1.8868110236220472, + "grad_norm": 14.452948570251465, + "learning_rate": 5.886207726008805e-07, + "loss": 0.6524, + "step": 1918 + }, + { + "epoch": 1.8877952755905512, + "grad_norm": 13.0665283203125, + "learning_rate": 5.878657517556336e-07, + "loss": 0.5334, + "step": 1919 + }, + { + "epoch": 1.888779527559055, + "grad_norm": 9.819680213928223, + "learning_rate": 5.871105268818848e-07, + "loss": 0.2941, + "step": 1920 + }, + { + "epoch": 1.889763779527559, + "grad_norm": 13.505655288696289, + "learning_rate": 5.863550997672101e-07, + "loss": 1.0042, + "step": 1921 + }, + { + "epoch": 1.890748031496063, + "grad_norm": 11.612969398498535, + "learning_rate": 5.855994721996638e-07, + "loss": 0.6607, + "step": 1922 + }, + { + "epoch": 1.891732283464567, + "grad_norm": 13.760869979858398, + "learning_rate": 5.848436459677752e-07, + "loss": 0.6232, + "step": 1923 + }, + { + "epoch": 1.8927165354330708, + "grad_norm": 11.171491622924805, + "learning_rate": 5.840876228605432e-07, + "loss": 0.7116, + "step": 1924 + }, + { + "epoch": 1.8937007874015748, + "grad_norm": 10.05971908569336, + "learning_rate": 5.833314046674329e-07, + "loss": 0.5706, + "step": 1925 + }, + { + "epoch": 1.8946850393700787, + "grad_norm": 12.375988960266113, + "learning_rate": 5.825749931783715e-07, + "loss": 0.5863, + "step": 1926 + }, + { + "epoch": 1.8956692913385826, + "grad_norm": 7.188908576965332, + "learning_rate": 5.818183901837435e-07, + "loss": 0.2272, + "step": 1927 + }, + { + "epoch": 1.8966535433070866, + "grad_norm": 9.860791206359863, + "learning_rate": 5.810615974743864e-07, + "loss": 0.3337, + "step": 1928 + }, + { + "epoch": 1.8976377952755905, + "grad_norm": 11.774974822998047, + "learning_rate": 5.803046168415872e-07, + "loss": 0.7918, + "step": 1929 + }, + { + "epoch": 1.8986220472440944, + "grad_norm": 8.987882614135742, + "learning_rate": 5.795474500770772e-07, + "loss": 0.3992, + "step": 1930 + }, + { + "epoch": 1.8996062992125984, + "grad_norm": 7.925570487976074, + "learning_rate": 5.787900989730291e-07, + "loss": 0.2711, + "step": 1931 + }, + { + "epoch": 1.9005905511811023, + "grad_norm": 21.005599975585938, + "learning_rate": 5.780325653220509e-07, + "loss": 0.7199, + "step": 1932 + }, + { + "epoch": 1.9015748031496063, + "grad_norm": 9.796464920043945, + "learning_rate": 5.772748509171833e-07, + "loss": 0.3781, + "step": 1933 + }, + { + "epoch": 1.9025590551181102, + "grad_norm": 5.327077865600586, + "learning_rate": 5.765169575518948e-07, + "loss": 0.1524, + "step": 1934 + }, + { + "epoch": 1.9035433070866141, + "grad_norm": 3.7077765464782715, + "learning_rate": 5.757588870200772e-07, + "loss": 0.2347, + "step": 1935 + }, + { + "epoch": 1.904527559055118, + "grad_norm": 14.269500732421875, + "learning_rate": 5.750006411160421e-07, + "loss": 0.8348, + "step": 1936 + }, + { + "epoch": 1.905511811023622, + "grad_norm": 12.124266624450684, + "learning_rate": 5.742422216345157e-07, + "loss": 0.5498, + "step": 1937 + }, + { + "epoch": 1.906496062992126, + "grad_norm": 6.518699645996094, + "learning_rate": 5.734836303706352e-07, + "loss": 0.226, + "step": 1938 + }, + { + "epoch": 1.90748031496063, + "grad_norm": 12.600378036499023, + "learning_rate": 5.727248691199446e-07, + "loss": 0.719, + "step": 1939 + }, + { + "epoch": 1.9084645669291338, + "grad_norm": 6.58778190612793, + "learning_rate": 5.719659396783902e-07, + "loss": 0.2547, + "step": 1940 + }, + { + "epoch": 1.909448818897638, + "grad_norm": 11.544054985046387, + "learning_rate": 5.712068438423161e-07, + "loss": 0.5611, + "step": 1941 + }, + { + "epoch": 1.9104330708661417, + "grad_norm": 5.600453853607178, + "learning_rate": 5.704475834084608e-07, + "loss": 0.1426, + "step": 1942 + }, + { + "epoch": 1.9114173228346458, + "grad_norm": 14.013601303100586, + "learning_rate": 5.696881601739517e-07, + "loss": 0.743, + "step": 1943 + }, + { + "epoch": 1.9124015748031495, + "grad_norm": 12.552058219909668, + "learning_rate": 5.689285759363022e-07, + "loss": 0.5874, + "step": 1944 + }, + { + "epoch": 1.9133858267716537, + "grad_norm": 12.306783676147461, + "learning_rate": 5.681688324934064e-07, + "loss": 0.3334, + "step": 1945 + }, + { + "epoch": 1.9143700787401574, + "grad_norm": 12.582601547241211, + "learning_rate": 5.674089316435353e-07, + "loss": 0.4916, + "step": 1946 + }, + { + "epoch": 1.9153543307086616, + "grad_norm": 13.343052864074707, + "learning_rate": 5.666488751853321e-07, + "loss": 0.7915, + "step": 1947 + }, + { + "epoch": 1.9163385826771653, + "grad_norm": 4.297492027282715, + "learning_rate": 5.658886649178096e-07, + "loss": 0.1582, + "step": 1948 + }, + { + "epoch": 1.9173228346456694, + "grad_norm": 12.203116416931152, + "learning_rate": 5.651283026403427e-07, + "loss": 0.408, + "step": 1949 + }, + { + "epoch": 1.9183070866141732, + "grad_norm": 11.420016288757324, + "learning_rate": 5.64367790152668e-07, + "loss": 0.4246, + "step": 1950 + }, + { + "epoch": 1.9192913385826773, + "grad_norm": 11.461502075195312, + "learning_rate": 5.636071292548766e-07, + "loss": 0.5638, + "step": 1951 + }, + { + "epoch": 1.920275590551181, + "grad_norm": 10.541890144348145, + "learning_rate": 5.628463217474108e-07, + "loss": 0.4243, + "step": 1952 + }, + { + "epoch": 1.9212598425196852, + "grad_norm": 9.319199562072754, + "learning_rate": 5.620853694310604e-07, + "loss": 0.4148, + "step": 1953 + }, + { + "epoch": 1.922244094488189, + "grad_norm": 9.032517433166504, + "learning_rate": 5.613242741069578e-07, + "loss": 0.268, + "step": 1954 + }, + { + "epoch": 1.923228346456693, + "grad_norm": 6.920299053192139, + "learning_rate": 5.605630375765739e-07, + "loss": 0.2287, + "step": 1955 + }, + { + "epoch": 1.9242125984251968, + "grad_norm": 8.552000999450684, + "learning_rate": 5.598016616417136e-07, + "loss": 0.1981, + "step": 1956 + }, + { + "epoch": 1.925196850393701, + "grad_norm": 28.06435775756836, + "learning_rate": 5.590401481045118e-07, + "loss": 1.661, + "step": 1957 + }, + { + "epoch": 1.9261811023622046, + "grad_norm": 17.66850471496582, + "learning_rate": 5.582784987674296e-07, + "loss": 0.6172, + "step": 1958 + }, + { + "epoch": 1.9271653543307088, + "grad_norm": 10.603317260742188, + "learning_rate": 5.575167154332489e-07, + "loss": 0.7265, + "step": 1959 + }, + { + "epoch": 1.9281496062992125, + "grad_norm": 10.895527839660645, + "learning_rate": 5.567547999050692e-07, + "loss": 0.4045, + "step": 1960 + }, + { + "epoch": 1.9291338582677167, + "grad_norm": 12.122705459594727, + "learning_rate": 5.559927539863024e-07, + "loss": 0.3756, + "step": 1961 + }, + { + "epoch": 1.9301181102362204, + "grad_norm": 12.188031196594238, + "learning_rate": 5.552305794806699e-07, + "loss": 0.5601, + "step": 1962 + }, + { + "epoch": 1.9311023622047245, + "grad_norm": 8.183180809020996, + "learning_rate": 5.544682781921964e-07, + "loss": 0.2951, + "step": 1963 + }, + { + "epoch": 1.9320866141732282, + "grad_norm": 20.317163467407227, + "learning_rate": 5.537058519252076e-07, + "loss": 0.8282, + "step": 1964 + }, + { + "epoch": 1.9330708661417324, + "grad_norm": 10.017477989196777, + "learning_rate": 5.529433024843244e-07, + "loss": 0.3984, + "step": 1965 + }, + { + "epoch": 1.9340551181102361, + "grad_norm": 12.640592575073242, + "learning_rate": 5.521806316744594e-07, + "loss": 0.6055, + "step": 1966 + }, + { + "epoch": 1.9350393700787403, + "grad_norm": 10.652737617492676, + "learning_rate": 5.514178413008125e-07, + "loss": 0.5407, + "step": 1967 + }, + { + "epoch": 1.936023622047244, + "grad_norm": 7.8703837394714355, + "learning_rate": 5.506549331688669e-07, + "loss": 0.2173, + "step": 1968 + }, + { + "epoch": 1.9370078740157481, + "grad_norm": 16.282611846923828, + "learning_rate": 5.49891909084384e-07, + "loss": 0.5948, + "step": 1969 + }, + { + "epoch": 1.9379921259842519, + "grad_norm": 9.560543060302734, + "learning_rate": 5.491287708534e-07, + "loss": 0.4105, + "step": 1970 + }, + { + "epoch": 1.938976377952756, + "grad_norm": 6.578395843505859, + "learning_rate": 5.483655202822214e-07, + "loss": 0.1847, + "step": 1971 + }, + { + "epoch": 1.9399606299212597, + "grad_norm": 9.591692924499512, + "learning_rate": 5.4760215917742e-07, + "loss": 0.4112, + "step": 1972 + }, + { + "epoch": 1.9409448818897639, + "grad_norm": 5.674706935882568, + "learning_rate": 5.4683868934583e-07, + "loss": 0.1703, + "step": 1973 + }, + { + "epoch": 1.9419291338582676, + "grad_norm": 10.246027946472168, + "learning_rate": 5.460751125945424e-07, + "loss": 0.4364, + "step": 1974 + }, + { + "epoch": 1.9429133858267718, + "grad_norm": 15.110301971435547, + "learning_rate": 5.453114307309015e-07, + "loss": 0.4748, + "step": 1975 + }, + { + "epoch": 1.9438976377952755, + "grad_norm": 7.355463981628418, + "learning_rate": 5.445476455625002e-07, + "loss": 0.224, + "step": 1976 + }, + { + "epoch": 1.9448818897637796, + "grad_norm": 13.137723922729492, + "learning_rate": 5.437837588971761e-07, + "loss": 0.8408, + "step": 1977 + }, + { + "epoch": 1.9458661417322833, + "grad_norm": 14.898588180541992, + "learning_rate": 5.430197725430071e-07, + "loss": 0.5296, + "step": 1978 + }, + { + "epoch": 1.9468503937007875, + "grad_norm": 15.721670150756836, + "learning_rate": 5.422556883083069e-07, + "loss": 0.7688, + "step": 1979 + }, + { + "epoch": 1.9478346456692912, + "grad_norm": 13.495804786682129, + "learning_rate": 5.414915080016209e-07, + "loss": 1.0069, + "step": 1980 + }, + { + "epoch": 1.9488188976377954, + "grad_norm": 22.801025390625, + "learning_rate": 5.407272334317221e-07, + "loss": 1.1509, + "step": 1981 + }, + { + "epoch": 1.949803149606299, + "grad_norm": 14.018686294555664, + "learning_rate": 5.399628664076062e-07, + "loss": 0.5735, + "step": 1982 + }, + { + "epoch": 1.9507874015748032, + "grad_norm": 12.201386451721191, + "learning_rate": 5.391984087384882e-07, + "loss": 0.5137, + "step": 1983 + }, + { + "epoch": 1.951771653543307, + "grad_norm": 11.208210945129395, + "learning_rate": 5.384338622337973e-07, + "loss": 0.3373, + "step": 1984 + }, + { + "epoch": 1.952755905511811, + "grad_norm": 10.17148494720459, + "learning_rate": 5.37669228703173e-07, + "loss": 0.3546, + "step": 1985 + }, + { + "epoch": 1.9537401574803148, + "grad_norm": 12.111408233642578, + "learning_rate": 5.369045099564613e-07, + "loss": 0.5453, + "step": 1986 + }, + { + "epoch": 1.954724409448819, + "grad_norm": 18.0567569732666, + "learning_rate": 5.361397078037089e-07, + "loss": 0.8836, + "step": 1987 + }, + { + "epoch": 1.9557086614173227, + "grad_norm": 8.216286659240723, + "learning_rate": 5.353748240551612e-07, + "loss": 0.2828, + "step": 1988 + }, + { + "epoch": 1.9566929133858268, + "grad_norm": 9.901942253112793, + "learning_rate": 5.346098605212553e-07, + "loss": 0.3481, + "step": 1989 + }, + { + "epoch": 1.9566929133858268, + "eval_Qnli-dev_cosine_accuracy": 0.658203125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8023818731307983, + "eval_Qnli-dev_cosine_ap": 0.6886786340579703, + "eval_Qnli-dev_cosine_f1": 0.6916802610114192, + "eval_Qnli-dev_cosine_f1_threshold": 0.7310751676559448, + "eval_Qnli-dev_cosine_precision": 0.5623342175066313, + "eval_Qnli-dev_cosine_recall": 0.8983050847457628, + "eval_Qnli-dev_dot_accuracy": 0.66015625, + "eval_Qnli-dev_dot_accuracy_threshold": 615.7371826171875, + "eval_Qnli-dev_dot_ap": 0.6875993155505726, + "eval_Qnli-dev_dot_f1": 0.6916802610114192, + "eval_Qnli-dev_dot_f1_threshold": 558.3238525390625, + "eval_Qnli-dev_dot_precision": 0.5623342175066313, + "eval_Qnli-dev_dot_recall": 0.8983050847457628, + "eval_Qnli-dev_euclidean_accuracy": 0.662109375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.46044158935547, + "eval_Qnli-dev_euclidean_ap": 0.6887027301873778, + "eval_Qnli-dev_euclidean_f1": 0.6906710310965629, + "eval_Qnli-dev_euclidean_f1_threshold": 20.146350860595703, + "eval_Qnli-dev_euclidean_precision": 0.5626666666666666, + "eval_Qnli-dev_euclidean_recall": 0.8940677966101694, + "eval_Qnli-dev_manhattan_accuracy": 0.677734375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 363.1766052246094, + "eval_Qnli-dev_manhattan_ap": 0.7003574987662966, + "eval_Qnli-dev_manhattan_f1": 0.6951026856240127, + "eval_Qnli-dev_manhattan_f1_threshold": 436.8890075683594, + "eval_Qnli-dev_manhattan_precision": 0.5541561712846348, + "eval_Qnli-dev_manhattan_recall": 0.9322033898305084, + "eval_Qnli-dev_max_accuracy": 0.677734375, + "eval_Qnli-dev_max_accuracy_threshold": 615.7371826171875, + "eval_Qnli-dev_max_ap": 0.7003574987662966, + "eval_Qnli-dev_max_f1": 0.6951026856240127, + "eval_Qnli-dev_max_f1_threshold": 558.3238525390625, + "eval_Qnli-dev_max_precision": 0.5626666666666666, + "eval_Qnli-dev_max_recall": 0.9322033898305084, + "eval_allNLI-dev_cosine_accuracy": 0.701171875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9141938090324402, + "eval_allNLI-dev_cosine_ap": 0.5178189449379189, + "eval_allNLI-dev_cosine_f1": 0.5469728601252609, + "eval_allNLI-dev_cosine_f1_threshold": 0.8089991807937622, + "eval_allNLI-dev_cosine_precision": 0.42810457516339867, + "eval_allNLI-dev_cosine_recall": 0.7572254335260116, + "eval_allNLI-dev_dot_accuracy": 0.701171875, + "eval_allNLI-dev_dot_accuracy_threshold": 701.9563598632812, + "eval_allNLI-dev_dot_ap": 0.5185120023725823, + "eval_allNLI-dev_dot_f1": 0.5492662473794548, + "eval_allNLI-dev_dot_f1_threshold": 619.0263671875, + "eval_allNLI-dev_dot_precision": 0.4309210526315789, + "eval_allNLI-dev_dot_recall": 0.7572254335260116, + "eval_allNLI-dev_euclidean_accuracy": 0.701171875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.46473503112793, + "eval_allNLI-dev_euclidean_ap": 0.5178582923548423, + "eval_allNLI-dev_euclidean_f1": 0.5458422174840084, + "eval_allNLI-dev_euclidean_f1_threshold": 16.86907958984375, + "eval_allNLI-dev_euclidean_precision": 0.43243243243243246, + "eval_allNLI-dev_euclidean_recall": 0.7398843930635838, + "eval_allNLI-dev_manhattan_accuracy": 0.697265625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 243.82118225097656, + "eval_allNLI-dev_manhattan_ap": 0.5208045922745452, + "eval_allNLI-dev_manhattan_f1": 0.5536480686695279, + "eval_allNLI-dev_manhattan_f1_threshold": 353.4998474121094, + "eval_allNLI-dev_manhattan_precision": 0.4402730375426621, + "eval_allNLI-dev_manhattan_recall": 0.7456647398843931, + "eval_allNLI-dev_max_accuracy": 0.701171875, + "eval_allNLI-dev_max_accuracy_threshold": 701.9563598632812, + "eval_allNLI-dev_max_ap": 0.5208045922745452, + "eval_allNLI-dev_max_f1": 0.5536480686695279, + "eval_allNLI-dev_max_f1_threshold": 619.0263671875, + "eval_allNLI-dev_max_precision": 0.4402730375426621, + "eval_allNLI-dev_max_recall": 0.7572254335260116, + "eval_loss": 0.9706022143363953, + "eval_runtime": 54.1103, + "eval_samples_per_second": 30.752, + "eval_sequential_score": 0.7003574987662966, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.6620027075959773, + "eval_sts-test_pearson_dot": 0.6601968030872636, + "eval_sts-test_pearson_euclidean": 0.6760063290032935, + "eval_sts-test_pearson_manhattan": 0.6792900086102078, + "eval_sts-test_pearson_max": 0.6792900086102078, + "eval_sts-test_spearman_cosine": 0.6712350889173807, + "eval_sts-test_spearman_dot": 0.6700788953280725, + "eval_sts-test_spearman_euclidean": 0.6714677059740932, + "eval_sts-test_spearman_manhattan": 0.6752350909143282, + "eval_sts-test_spearman_max": 0.6752350909143282, + "step": 1989 + }, + { + "epoch": 1.9576771653543306, + "grad_norm": 10.558905601501465, + "learning_rate": 5.338448190126184e-07, + "loss": 0.5158, + "step": 1990 + }, + { + "epoch": 1.9586614173228347, + "grad_norm": 16.4892520904541, + "learning_rate": 5.330797013400614e-07, + "loss": 1.1472, + "step": 1991 + }, + { + "epoch": 1.9596456692913384, + "grad_norm": 13.120809555053711, + "learning_rate": 5.323145093145763e-07, + "loss": 0.5507, + "step": 1992 + }, + { + "epoch": 1.9606299212598426, + "grad_norm": 13.758267402648926, + "learning_rate": 5.3154924474733e-07, + "loss": 0.4904, + "step": 1993 + }, + { + "epoch": 1.9616141732283463, + "grad_norm": 10.835561752319336, + "learning_rate": 5.307839094496621e-07, + "loss": 0.2959, + "step": 1994 + }, + { + "epoch": 1.9625984251968505, + "grad_norm": 7.2871222496032715, + "learning_rate": 5.300185052330793e-07, + "loss": 0.1981, + "step": 1995 + }, + { + "epoch": 1.9635826771653542, + "grad_norm": 9.152851104736328, + "learning_rate": 5.292530339092513e-07, + "loss": 0.332, + "step": 1996 + }, + { + "epoch": 1.9645669291338583, + "grad_norm": 10.531471252441406, + "learning_rate": 5.284874972900064e-07, + "loss": 0.3562, + "step": 1997 + }, + { + "epoch": 1.965551181102362, + "grad_norm": 21.480953216552734, + "learning_rate": 5.277218971873281e-07, + "loss": 0.7395, + "step": 1998 + }, + { + "epoch": 1.9665354330708662, + "grad_norm": 10.496049880981445, + "learning_rate": 5.269562354133496e-07, + "loss": 0.5546, + "step": 1999 + }, + { + "epoch": 1.96751968503937, + "grad_norm": 12.824313163757324, + "learning_rate": 5.261905137803504e-07, + "loss": 0.6003, + "step": 2000 + }, + { + "epoch": 1.968503937007874, + "grad_norm": 14.431049346923828, + "learning_rate": 5.254247341007514e-07, + "loss": 0.784, + "step": 2001 + }, + { + "epoch": 1.969488188976378, + "grad_norm": 11.358637809753418, + "learning_rate": 5.246588981871113e-07, + "loss": 0.5101, + "step": 2002 + }, + { + "epoch": 1.970472440944882, + "grad_norm": 10.88247013092041, + "learning_rate": 5.238930078521215e-07, + "loss": 0.5691, + "step": 2003 + }, + { + "epoch": 1.9714566929133859, + "grad_norm": 6.141389846801758, + "learning_rate": 5.231270649086023e-07, + "loss": 0.2075, + "step": 2004 + }, + { + "epoch": 1.9724409448818898, + "grad_norm": 11.258003234863281, + "learning_rate": 5.22361071169499e-07, + "loss": 0.4202, + "step": 2005 + }, + { + "epoch": 1.9734251968503937, + "grad_norm": 9.83862590789795, + "learning_rate": 5.215950284478763e-07, + "loss": 0.3815, + "step": 2006 + }, + { + "epoch": 1.9744094488188977, + "grad_norm": 6.295627593994141, + "learning_rate": 5.208289385569154e-07, + "loss": 0.2197, + "step": 2007 + }, + { + "epoch": 1.9753937007874016, + "grad_norm": 8.904109954833984, + "learning_rate": 5.20062803309909e-07, + "loss": 0.3602, + "step": 2008 + }, + { + "epoch": 1.9763779527559056, + "grad_norm": 10.298103332519531, + "learning_rate": 5.192966245202571e-07, + "loss": 0.295, + "step": 2009 + }, + { + "epoch": 1.9773622047244095, + "grad_norm": 9.491402626037598, + "learning_rate": 5.18530404001463e-07, + "loss": 0.3633, + "step": 2010 + }, + { + "epoch": 1.9783464566929134, + "grad_norm": 9.721626281738281, + "learning_rate": 5.177641435671284e-07, + "loss": 0.4982, + "step": 2011 + }, + { + "epoch": 1.9793307086614174, + "grad_norm": 16.09177017211914, + "learning_rate": 5.169978450309498e-07, + "loss": 0.8649, + "step": 2012 + }, + { + "epoch": 1.9803149606299213, + "grad_norm": 14.924250602722168, + "learning_rate": 5.162315102067139e-07, + "loss": 0.8593, + "step": 2013 + }, + { + "epoch": 1.9812992125984252, + "grad_norm": 11.977181434631348, + "learning_rate": 5.154651409082927e-07, + "loss": 0.5689, + "step": 2014 + }, + { + "epoch": 1.9822834645669292, + "grad_norm": 20.18109893798828, + "learning_rate": 5.146987389496408e-07, + "loss": 0.9294, + "step": 2015 + }, + { + "epoch": 1.983267716535433, + "grad_norm": 9.459802627563477, + "learning_rate": 5.139323061447892e-07, + "loss": 0.3826, + "step": 2016 + }, + { + "epoch": 1.984251968503937, + "grad_norm": 8.496149063110352, + "learning_rate": 5.131658443078426e-07, + "loss": 0.3136, + "step": 2017 + }, + { + "epoch": 1.985236220472441, + "grad_norm": 11.517538070678711, + "learning_rate": 5.123993552529736e-07, + "loss": 0.6924, + "step": 2018 + }, + { + "epoch": 1.986220472440945, + "grad_norm": 11.961481094360352, + "learning_rate": 5.116328407944201e-07, + "loss": 0.6278, + "step": 2019 + }, + { + "epoch": 1.9872047244094488, + "grad_norm": 11.670297622680664, + "learning_rate": 5.108663027464795e-07, + "loss": 0.615, + "step": 2020 + }, + { + "epoch": 1.9881889763779528, + "grad_norm": 11.576616287231445, + "learning_rate": 5.100997429235055e-07, + "loss": 0.5819, + "step": 2021 + }, + { + "epoch": 1.9891732283464567, + "grad_norm": 9.68857192993164, + "learning_rate": 5.093331631399029e-07, + "loss": 0.5012, + "step": 2022 + }, + { + "epoch": 1.9901574803149606, + "grad_norm": 11.992721557617188, + "learning_rate": 5.085665652101244e-07, + "loss": 0.4416, + "step": 2023 + }, + { + "epoch": 1.9911417322834646, + "grad_norm": 9.953713417053223, + "learning_rate": 5.077999509486645e-07, + "loss": 0.4151, + "step": 2024 + }, + { + "epoch": 1.9921259842519685, + "grad_norm": 17.759462356567383, + "learning_rate": 5.070333221700575e-07, + "loss": 0.6352, + "step": 2025 + }, + { + "epoch": 1.9931102362204725, + "grad_norm": 4.798762321472168, + "learning_rate": 5.062666806888715e-07, + "loss": 0.1386, + "step": 2026 + }, + { + "epoch": 1.9940944881889764, + "grad_norm": 9.868159294128418, + "learning_rate": 5.055000283197051e-07, + "loss": 0.4451, + "step": 2027 + }, + { + "epoch": 1.9950787401574803, + "grad_norm": 6.54140567779541, + "learning_rate": 5.047333668771821e-07, + "loss": 0.2217, + "step": 2028 + }, + { + "epoch": 1.9960629921259843, + "grad_norm": 9.168428421020508, + "learning_rate": 5.03966698175948e-07, + "loss": 0.3559, + "step": 2029 + }, + { + "epoch": 1.9970472440944882, + "grad_norm": 6.069472312927246, + "learning_rate": 5.032000240306655e-07, + "loss": 0.3077, + "step": 2030 + }, + { + "epoch": 1.9980314960629921, + "grad_norm": 18.002717971801758, + "learning_rate": 5.024333462560104e-07, + "loss": 1.0025, + "step": 2031 + }, + { + "epoch": 1.999015748031496, + "grad_norm": 15.968050956726074, + "learning_rate": 5.016666666666668e-07, + "loss": 0.9233, + "step": 2032 + }, + { + "epoch": 2.0, + "grad_norm": 21.219402313232422, + "learning_rate": 5.008999870773232e-07, + "loss": 0.52, + "step": 2033 + }, + { + "epoch": 2.0009842519685037, + "grad_norm": 0.0, + "learning_rate": 5.001333093026681e-07, + "loss": 0.0, + "step": 2034 + }, + { + "epoch": 2.000984251968504, + "grad_norm": 19.841766357421875, + "learning_rate": 4.993666351573856e-07, + "loss": 1.0536, + "step": 2035 + }, + { + "epoch": 2.001968503937008, + "grad_norm": 12.783218383789062, + "learning_rate": 4.985999664561514e-07, + "loss": 0.5749, + "step": 2036 + }, + { + "epoch": 2.002952755905512, + "grad_norm": 21.010698318481445, + "learning_rate": 4.978333050136285e-07, + "loss": 0.9689, + "step": 2037 + }, + { + "epoch": 2.0039370078740157, + "grad_norm": 10.027554512023926, + "learning_rate": 4.970666526444619e-07, + "loss": 0.4037, + "step": 2038 + }, + { + "epoch": 2.00492125984252, + "grad_norm": 14.6770658493042, + "learning_rate": 4.96300011163276e-07, + "loss": 0.6314, + "step": 2039 + }, + { + "epoch": 2.0059055118110236, + "grad_norm": 6.943289279937744, + "learning_rate": 4.955333823846691e-07, + "loss": 0.1926, + "step": 2040 + }, + { + "epoch": 2.0068897637795278, + "grad_norm": 13.570602416992188, + "learning_rate": 4.947667681232094e-07, + "loss": 0.8417, + "step": 2041 + }, + { + "epoch": 2.0078740157480315, + "grad_norm": 12.345090866088867, + "learning_rate": 4.940001701934305e-07, + "loss": 0.6205, + "step": 2042 + }, + { + "epoch": 2.0088582677165356, + "grad_norm": 9.181805610656738, + "learning_rate": 4.932335904098282e-07, + "loss": 0.2801, + "step": 2043 + }, + { + "epoch": 2.0098425196850394, + "grad_norm": 17.511737823486328, + "learning_rate": 4.924670305868541e-07, + "loss": 0.6844, + "step": 2044 + }, + { + "epoch": 2.0108267716535435, + "grad_norm": 12.344738960266113, + "learning_rate": 4.917004925389137e-07, + "loss": 0.6534, + "step": 2045 + }, + { + "epoch": 2.0118110236220472, + "grad_norm": 14.163546562194824, + "learning_rate": 4.909339780803601e-07, + "loss": 0.5133, + "step": 2046 + }, + { + "epoch": 2.0127952755905514, + "grad_norm": 10.767459869384766, + "learning_rate": 4.901674890254911e-07, + "loss": 0.4383, + "step": 2047 + }, + { + "epoch": 2.013779527559055, + "grad_norm": 14.008377075195312, + "learning_rate": 4.894010271885444e-07, + "loss": 0.7045, + "step": 2048 + }, + { + "epoch": 2.0147637795275593, + "grad_norm": 11.19058895111084, + "learning_rate": 4.886345943836928e-07, + "loss": 0.4953, + "step": 2049 + }, + { + "epoch": 2.015748031496063, + "grad_norm": 6.641456127166748, + "learning_rate": 4.878681924250408e-07, + "loss": 0.2567, + "step": 2050 + }, + { + "epoch": 2.016732283464567, + "grad_norm": 8.315643310546875, + "learning_rate": 4.871018231266197e-07, + "loss": 0.4574, + "step": 2051 + }, + { + "epoch": 2.017716535433071, + "grad_norm": 9.133217811584473, + "learning_rate": 4.863354883023838e-07, + "loss": 0.3211, + "step": 2052 + }, + { + "epoch": 2.018700787401575, + "grad_norm": 15.33566951751709, + "learning_rate": 4.85569189766205e-07, + "loss": 0.7028, + "step": 2053 + }, + { + "epoch": 2.0196850393700787, + "grad_norm": 11.80440616607666, + "learning_rate": 4.848029293318705e-07, + "loss": 0.5107, + "step": 2054 + }, + { + "epoch": 2.020669291338583, + "grad_norm": 11.949298858642578, + "learning_rate": 4.840367088130764e-07, + "loss": 0.327, + "step": 2055 + }, + { + "epoch": 2.0216535433070866, + "grad_norm": 14.780426979064941, + "learning_rate": 4.832705300234245e-07, + "loss": 0.7758, + "step": 2056 + }, + { + "epoch": 2.0226377952755907, + "grad_norm": 8.046977996826172, + "learning_rate": 4.825043947764181e-07, + "loss": 0.2169, + "step": 2057 + }, + { + "epoch": 2.0236220472440944, + "grad_norm": 4.532480716705322, + "learning_rate": 4.817383048854572e-07, + "loss": 0.0995, + "step": 2058 + }, + { + "epoch": 2.0246062992125986, + "grad_norm": 10.037430763244629, + "learning_rate": 4.809722621638345e-07, + "loss": 0.4255, + "step": 2059 + }, + { + "epoch": 2.0255905511811023, + "grad_norm": 14.573445320129395, + "learning_rate": 4.802062684247311e-07, + "loss": 0.6974, + "step": 2060 + }, + { + "epoch": 2.0265748031496065, + "grad_norm": 9.875505447387695, + "learning_rate": 4.79440325481212e-07, + "loss": 0.3386, + "step": 2061 + }, + { + "epoch": 2.02755905511811, + "grad_norm": 11.617470741271973, + "learning_rate": 4.786744351462222e-07, + "loss": 0.4871, + "step": 2062 + }, + { + "epoch": 2.0285433070866143, + "grad_norm": 15.038314819335938, + "learning_rate": 4.779085992325821e-07, + "loss": 0.7656, + "step": 2063 + }, + { + "epoch": 2.029527559055118, + "grad_norm": 10.356962203979492, + "learning_rate": 4.771428195529833e-07, + "loss": 0.3288, + "step": 2064 + }, + { + "epoch": 2.030511811023622, + "grad_norm": 7.781601905822754, + "learning_rate": 4.763770979199841e-07, + "loss": 0.3246, + "step": 2065 + }, + { + "epoch": 2.031496062992126, + "grad_norm": 10.731339454650879, + "learning_rate": 4.756114361460057e-07, + "loss": 0.5359, + "step": 2066 + }, + { + "epoch": 2.03248031496063, + "grad_norm": 13.668349266052246, + "learning_rate": 4.748458360433273e-07, + "loss": 0.5731, + "step": 2067 + }, + { + "epoch": 2.033464566929134, + "grad_norm": 12.049600601196289, + "learning_rate": 4.7408029942408244e-07, + "loss": 0.6563, + "step": 2068 + }, + { + "epoch": 2.034448818897638, + "grad_norm": 10.32591724395752, + "learning_rate": 4.733148281002543e-07, + "loss": 0.434, + "step": 2069 + }, + { + "epoch": 2.0354330708661417, + "grad_norm": 14.998777389526367, + "learning_rate": 4.7254942388367147e-07, + "loss": 0.8234, + "step": 2070 + }, + { + "epoch": 2.036417322834646, + "grad_norm": 12.361584663391113, + "learning_rate": 4.717840885860036e-07, + "loss": 0.7725, + "step": 2071 + }, + { + "epoch": 2.0374015748031495, + "grad_norm": 9.835253715515137, + "learning_rate": 4.7101882401875745e-07, + "loss": 0.5631, + "step": 2072 + }, + { + "epoch": 2.0383858267716537, + "grad_norm": 7.7654709815979, + "learning_rate": 4.7025363199327213e-07, + "loss": 0.3587, + "step": 2073 + }, + { + "epoch": 2.0393700787401574, + "grad_norm": 11.994633674621582, + "learning_rate": 4.694885143207153e-07, + "loss": 0.6894, + "step": 2074 + }, + { + "epoch": 2.0403543307086616, + "grad_norm": 15.344325065612793, + "learning_rate": 4.687234728120783e-07, + "loss": 0.8536, + "step": 2075 + }, + { + "epoch": 2.0413385826771653, + "grad_norm": 6.544132232666016, + "learning_rate": 4.6795850927817256e-07, + "loss": 0.2449, + "step": 2076 + }, + { + "epoch": 2.0423228346456694, + "grad_norm": 12.171512603759766, + "learning_rate": 4.6719362552962457e-07, + "loss": 0.5641, + "step": 2077 + }, + { + "epoch": 2.043307086614173, + "grad_norm": 9.727547645568848, + "learning_rate": 4.664288233768724e-07, + "loss": 0.4836, + "step": 2078 + }, + { + "epoch": 2.0442913385826773, + "grad_norm": 17.973478317260742, + "learning_rate": 4.6566410463016044e-07, + "loss": 0.905, + "step": 2079 + }, + { + "epoch": 2.045275590551181, + "grad_norm": 10.42161750793457, + "learning_rate": 4.648994710995363e-07, + "loss": 0.4315, + "step": 2080 + }, + { + "epoch": 2.046259842519685, + "grad_norm": 10.346402168273926, + "learning_rate": 4.641349245948454e-07, + "loss": 0.3829, + "step": 2081 + }, + { + "epoch": 2.047244094488189, + "grad_norm": 8.672287940979004, + "learning_rate": 4.633704669257273e-07, + "loss": 0.366, + "step": 2082 + }, + { + "epoch": 2.048228346456693, + "grad_norm": 8.586081504821777, + "learning_rate": 4.626060999016115e-07, + "loss": 0.3392, + "step": 2083 + }, + { + "epoch": 2.0492125984251968, + "grad_norm": 11.210050582885742, + "learning_rate": 4.6184182533171257e-07, + "loss": 0.3751, + "step": 2084 + }, + { + "epoch": 2.050196850393701, + "grad_norm": 7.204981803894043, + "learning_rate": 4.6107764502502684e-07, + "loss": 0.2075, + "step": 2085 + }, + { + "epoch": 2.0511811023622046, + "grad_norm": 15.422869682312012, + "learning_rate": 4.603135607903266e-07, + "loss": 0.8522, + "step": 2086 + }, + { + "epoch": 2.052165354330709, + "grad_norm": 9.124427795410156, + "learning_rate": 4.5954957443615765e-07, + "loss": 0.4735, + "step": 2087 + }, + { + "epoch": 2.0531496062992125, + "grad_norm": 9.447388648986816, + "learning_rate": 4.5878568777083356e-07, + "loss": 0.4227, + "step": 2088 + }, + { + "epoch": 2.0541338582677167, + "grad_norm": 12.309717178344727, + "learning_rate": 4.5802190260243226e-07, + "loss": 0.5437, + "step": 2089 + }, + { + "epoch": 2.0551181102362204, + "grad_norm": 9.313911437988281, + "learning_rate": 4.572582207387912e-07, + "loss": 0.33, + "step": 2090 + }, + { + "epoch": 2.0561023622047245, + "grad_norm": 11.166501998901367, + "learning_rate": 4.5649464398750356e-07, + "loss": 0.3904, + "step": 2091 + }, + { + "epoch": 2.0570866141732282, + "grad_norm": 8.26967716217041, + "learning_rate": 4.557311741559135e-07, + "loss": 0.3672, + "step": 2092 + }, + { + "epoch": 2.0580708661417324, + "grad_norm": 16.655155181884766, + "learning_rate": 4.5496781305111215e-07, + "loss": 1.0232, + "step": 2093 + }, + { + "epoch": 2.059055118110236, + "grad_norm": 7.771897315979004, + "learning_rate": 4.5420456247993354e-07, + "loss": 0.3397, + "step": 2094 + }, + { + "epoch": 2.0600393700787403, + "grad_norm": 14.735974311828613, + "learning_rate": 4.5344142424894954e-07, + "loss": 0.6183, + "step": 2095 + }, + { + "epoch": 2.061023622047244, + "grad_norm": 7.392879009246826, + "learning_rate": 4.526784001644667e-07, + "loss": 0.282, + "step": 2096 + }, + { + "epoch": 2.062007874015748, + "grad_norm": 8.864421844482422, + "learning_rate": 4.5191549203252103e-07, + "loss": 0.2667, + "step": 2097 + }, + { + "epoch": 2.062992125984252, + "grad_norm": 10.327754974365234, + "learning_rate": 4.511527016588742e-07, + "loss": 0.5652, + "step": 2098 + }, + { + "epoch": 2.063976377952756, + "grad_norm": 14.026025772094727, + "learning_rate": 4.503900308490092e-07, + "loss": 1.0643, + "step": 2099 + }, + { + "epoch": 2.0649606299212597, + "grad_norm": 11.10537052154541, + "learning_rate": 4.496274814081259e-07, + "loss": 0.6203, + "step": 2100 + }, + { + "epoch": 2.065944881889764, + "grad_norm": 9.276914596557617, + "learning_rate": 4.4886505514113695e-07, + "loss": 0.4321, + "step": 2101 + }, + { + "epoch": 2.0669291338582676, + "grad_norm": 14.24113655090332, + "learning_rate": 4.481027538526636e-07, + "loss": 0.8522, + "step": 2102 + }, + { + "epoch": 2.0679133858267718, + "grad_norm": 12.71141529083252, + "learning_rate": 4.4734057934703095e-07, + "loss": 0.4217, + "step": 2103 + }, + { + "epoch": 2.0688976377952755, + "grad_norm": 17.116039276123047, + "learning_rate": 4.4657853342826435e-07, + "loss": 0.9329, + "step": 2104 + }, + { + "epoch": 2.0698818897637796, + "grad_norm": 9.301265716552734, + "learning_rate": 4.458166179000848e-07, + "loss": 0.4836, + "step": 2105 + }, + { + "epoch": 2.0708661417322833, + "grad_norm": 8.274540901184082, + "learning_rate": 4.4505483456590413e-07, + "loss": 0.1996, + "step": 2106 + }, + { + "epoch": 2.0718503937007875, + "grad_norm": 14.8345308303833, + "learning_rate": 4.442931852288219e-07, + "loss": 0.7449, + "step": 2107 + }, + { + "epoch": 2.072834645669291, + "grad_norm": 10.62204360961914, + "learning_rate": 4.435316716916202e-07, + "loss": 0.4782, + "step": 2108 + }, + { + "epoch": 2.0738188976377954, + "grad_norm": 6.196303367614746, + "learning_rate": 4.4277029575675983e-07, + "loss": 0.2508, + "step": 2109 + }, + { + "epoch": 2.074803149606299, + "grad_norm": 10.151144981384277, + "learning_rate": 4.4200905922637575e-07, + "loss": 0.1712, + "step": 2110 + }, + { + "epoch": 2.0757874015748032, + "grad_norm": 13.606684684753418, + "learning_rate": 4.4124796390227313e-07, + "loss": 0.9656, + "step": 2111 + }, + { + "epoch": 2.076771653543307, + "grad_norm": 11.874922752380371, + "learning_rate": 4.4048701158592277e-07, + "loss": 0.4058, + "step": 2112 + }, + { + "epoch": 2.077755905511811, + "grad_norm": 12.032085418701172, + "learning_rate": 4.3972620407845714e-07, + "loss": 0.3136, + "step": 2113 + }, + { + "epoch": 2.078740157480315, + "grad_norm": 12.020431518554688, + "learning_rate": 4.3896554318066556e-07, + "loss": 0.6674, + "step": 2114 + }, + { + "epoch": 2.079724409448819, + "grad_norm": 11.953283309936523, + "learning_rate": 4.3820503069299077e-07, + "loss": 0.5842, + "step": 2115 + }, + { + "epoch": 2.0807086614173227, + "grad_norm": 10.502331733703613, + "learning_rate": 4.3744466841552416e-07, + "loss": 0.5443, + "step": 2116 + }, + { + "epoch": 2.081692913385827, + "grad_norm": 12.981659889221191, + "learning_rate": 4.3668445814800137e-07, + "loss": 0.678, + "step": 2117 + }, + { + "epoch": 2.0826771653543306, + "grad_norm": 7.417089939117432, + "learning_rate": 4.359244016897984e-07, + "loss": 0.237, + "step": 2118 + }, + { + "epoch": 2.0836614173228347, + "grad_norm": 9.361095428466797, + "learning_rate": 4.3516450083992714e-07, + "loss": 0.4916, + "step": 2119 + }, + { + "epoch": 2.0846456692913384, + "grad_norm": 12.709978103637695, + "learning_rate": 4.344047573970313e-07, + "loss": 0.7586, + "step": 2120 + }, + { + "epoch": 2.0856299212598426, + "grad_norm": 13.901938438415527, + "learning_rate": 4.336451731593817e-07, + "loss": 0.5489, + "step": 2121 + }, + { + "epoch": 2.0866141732283463, + "grad_norm": 13.167049407958984, + "learning_rate": 4.328857499248727e-07, + "loss": 0.4912, + "step": 2122 + }, + { + "epoch": 2.0875984251968505, + "grad_norm": 13.484986305236816, + "learning_rate": 4.321264894910173e-07, + "loss": 0.7887, + "step": 2123 + }, + { + "epoch": 2.088582677165354, + "grad_norm": 11.721290588378906, + "learning_rate": 4.3136739365494334e-07, + "loss": 0.5329, + "step": 2124 + }, + { + "epoch": 2.0895669291338583, + "grad_norm": 5.6164093017578125, + "learning_rate": 4.306084642133889e-07, + "loss": 0.1016, + "step": 2125 + }, + { + "epoch": 2.090551181102362, + "grad_norm": 10.29283618927002, + "learning_rate": 4.2984970296269856e-07, + "loss": 0.3978, + "step": 2126 + }, + { + "epoch": 2.091535433070866, + "grad_norm": 12.174274444580078, + "learning_rate": 4.290911116988181e-07, + "loss": 0.4734, + "step": 2127 + }, + { + "epoch": 2.09251968503937, + "grad_norm": 8.400107383728027, + "learning_rate": 4.2833269221729153e-07, + "loss": 0.3846, + "step": 2128 + }, + { + "epoch": 2.093503937007874, + "grad_norm": 6.006407260894775, + "learning_rate": 4.275744463132564e-07, + "loss": 0.1579, + "step": 2129 + }, + { + "epoch": 2.094488188976378, + "grad_norm": 11.806699752807617, + "learning_rate": 4.268163757814388e-07, + "loss": 0.5939, + "step": 2130 + }, + { + "epoch": 2.095472440944882, + "grad_norm": 12.883883476257324, + "learning_rate": 4.2605848241615035e-07, + "loss": 0.5173, + "step": 2131 + }, + { + "epoch": 2.0964566929133857, + "grad_norm": 8.410216331481934, + "learning_rate": 4.2530076801128275e-07, + "loss": 0.3282, + "step": 2132 + }, + { + "epoch": 2.09744094488189, + "grad_norm": 11.412123680114746, + "learning_rate": 4.2454323436030457e-07, + "loss": 0.5281, + "step": 2133 + }, + { + "epoch": 2.0984251968503935, + "grad_norm": 10.594966888427734, + "learning_rate": 4.2378588325625627e-07, + "loss": 0.4023, + "step": 2134 + }, + { + "epoch": 2.0994094488188977, + "grad_norm": 8.408379554748535, + "learning_rate": 4.2302871649174646e-07, + "loss": 0.3061, + "step": 2135 + }, + { + "epoch": 2.1003937007874014, + "grad_norm": 8.434245109558105, + "learning_rate": 4.2227173585894724e-07, + "loss": 0.4158, + "step": 2136 + }, + { + "epoch": 2.1013779527559056, + "grad_norm": 11.54025936126709, + "learning_rate": 4.2151494314959015e-07, + "loss": 0.5952, + "step": 2137 + }, + { + "epoch": 2.1023622047244093, + "grad_norm": 10.130776405334473, + "learning_rate": 4.20758340154962e-07, + "loss": 0.3903, + "step": 2138 + }, + { + "epoch": 2.1033464566929134, + "grad_norm": 10.995735168457031, + "learning_rate": 4.200019286659007e-07, + "loss": 0.3577, + "step": 2139 + }, + { + "epoch": 2.104330708661417, + "grad_norm": 6.9837327003479, + "learning_rate": 4.192457104727905e-07, + "loss": 0.196, + "step": 2140 + }, + { + "epoch": 2.1053149606299213, + "grad_norm": 9.005616188049316, + "learning_rate": 4.184896873655584e-07, + "loss": 0.38, + "step": 2141 + }, + { + "epoch": 2.106299212598425, + "grad_norm": 15.84337329864502, + "learning_rate": 4.177338611336697e-07, + "loss": 0.6768, + "step": 2142 + }, + { + "epoch": 2.106299212598425, + "eval_Qnli-dev_cosine_accuracy": 0.671875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8106182813644409, + "eval_Qnli-dev_cosine_ap": 0.6921761205800739, + "eval_Qnli-dev_cosine_f1": 0.6859903381642513, + "eval_Qnli-dev_cosine_f1_threshold": 0.7057304382324219, + "eval_Qnli-dev_cosine_precision": 0.5532467532467532, + "eval_Qnli-dev_cosine_recall": 0.902542372881356, + "eval_Qnli-dev_dot_accuracy": 0.673828125, + "eval_Qnli-dev_dot_accuracy_threshold": 613.2439575195312, + "eval_Qnli-dev_dot_ap": 0.6888309953492245, + "eval_Qnli-dev_dot_f1": 0.685064935064935, + "eval_Qnli-dev_dot_f1_threshold": 541.5089721679688, + "eval_Qnli-dev_dot_precision": 0.5552631578947368, + "eval_Qnli-dev_dot_recall": 0.8940677966101694, + "eval_Qnli-dev_euclidean_accuracy": 0.673828125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.046558380126953, + "eval_Qnli-dev_euclidean_ap": 0.692753086975512, + "eval_Qnli-dev_euclidean_f1": 0.6838709677419356, + "eval_Qnli-dev_euclidean_f1_threshold": 21.16459083557129, + "eval_Qnli-dev_euclidean_precision": 0.5520833333333334, + "eval_Qnli-dev_euclidean_recall": 0.8983050847457628, + "eval_Qnli-dev_manhattan_accuracy": 0.6875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 371.18255615234375, + "eval_Qnli-dev_manhattan_ap": 0.7050924683830799, + "eval_Qnli-dev_manhattan_f1": 0.686106346483705, + "eval_Qnli-dev_manhattan_f1_threshold": 425.8172607421875, + "eval_Qnli-dev_manhattan_precision": 0.5763688760806917, + "eval_Qnli-dev_manhattan_recall": 0.847457627118644, + "eval_Qnli-dev_max_accuracy": 0.6875, + "eval_Qnli-dev_max_accuracy_threshold": 613.2439575195312, + "eval_Qnli-dev_max_ap": 0.7050924683830799, + "eval_Qnli-dev_max_f1": 0.686106346483705, + "eval_Qnli-dev_max_f1_threshold": 541.5089721679688, + "eval_Qnli-dev_max_precision": 0.5763688760806917, + "eval_Qnli-dev_max_recall": 0.902542372881356, + "eval_allNLI-dev_cosine_accuracy": 0.689453125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9146234393119812, + "eval_allNLI-dev_cosine_ap": 0.5060845831275185, + "eval_allNLI-dev_cosine_f1": 0.5430210325047802, + "eval_allNLI-dev_cosine_f1_threshold": 0.7860702276229858, + "eval_allNLI-dev_cosine_precision": 0.4057142857142857, + "eval_allNLI-dev_cosine_recall": 0.8208092485549133, + "eval_allNLI-dev_dot_accuracy": 0.689453125, + "eval_allNLI-dev_dot_accuracy_threshold": 700.458740234375, + "eval_allNLI-dev_dot_ap": 0.5043834893000478, + "eval_allNLI-dev_dot_f1": 0.5440613026819924, + "eval_allNLI-dev_dot_f1_threshold": 601.576904296875, + "eval_allNLI-dev_dot_precision": 0.4068767908309456, + "eval_allNLI-dev_dot_recall": 0.8208092485549133, + "eval_allNLI-dev_euclidean_accuracy": 0.689453125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.43293571472168, + "eval_allNLI-dev_euclidean_ap": 0.505756126438367, + "eval_allNLI-dev_euclidean_f1": 0.5430210325047802, + "eval_allNLI-dev_euclidean_f1_threshold": 18.10074234008789, + "eval_allNLI-dev_euclidean_precision": 0.4057142857142857, + "eval_allNLI-dev_euclidean_recall": 0.8208092485549133, + "eval_allNLI-dev_manhattan_accuracy": 0.681640625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 216.44842529296875, + "eval_allNLI-dev_manhattan_ap": 0.5042689696590352, + "eval_allNLI-dev_manhattan_f1": 0.5437616387337059, + "eval_allNLI-dev_manhattan_f1_threshold": 386.01776123046875, + "eval_allNLI-dev_manhattan_precision": 0.4010989010989011, + "eval_allNLI-dev_manhattan_recall": 0.8439306358381503, + "eval_allNLI-dev_max_accuracy": 0.689453125, + "eval_allNLI-dev_max_accuracy_threshold": 700.458740234375, + "eval_allNLI-dev_max_ap": 0.5060845831275185, + "eval_allNLI-dev_max_f1": 0.5440613026819924, + "eval_allNLI-dev_max_f1_threshold": 601.576904296875, + "eval_allNLI-dev_max_precision": 0.4068767908309456, + "eval_allNLI-dev_max_recall": 0.8439306358381503, + "eval_loss": 0.9628643989562988, + "eval_runtime": 54.1521, + "eval_samples_per_second": 30.728, + "eval_sequential_score": 0.7050924683830799, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.6202459810611565, + "eval_sts-test_pearson_dot": 0.6186786427489253, + "eval_sts-test_pearson_euclidean": 0.6307823993253823, + "eval_sts-test_pearson_manhattan": 0.63666039543781, + "eval_sts-test_pearson_max": 0.63666039543781, + "eval_sts-test_spearman_cosine": 0.6312596821416715, + "eval_sts-test_spearman_dot": 0.630321964210476, + "eval_sts-test_spearman_euclidean": 0.6316041773518624, + "eval_sts-test_spearman_manhattan": 0.6356551062842734, + "eval_sts-test_spearman_max": 0.6356551062842734, + "step": 2142 + }, + { + "epoch": 2.107283464566929, + "grad_norm": 11.44278335571289, + "learning_rate": 4.169782335661235e-07, + "loss": 0.5215, + "step": 2143 + }, + { + "epoch": 2.108267716535433, + "grad_norm": 10.876325607299805, + "learning_rate": 4.1622280645144883e-07, + "loss": 0.3134, + "step": 2144 + }, + { + "epoch": 2.109251968503937, + "grad_norm": 9.747419357299805, + "learning_rate": 4.1546758157770004e-07, + "loss": 0.3255, + "step": 2145 + }, + { + "epoch": 2.1102362204724407, + "grad_norm": 15.33596420288086, + "learning_rate": 4.1471256073245306e-07, + "loss": 0.8023, + "step": 2146 + }, + { + "epoch": 2.111220472440945, + "grad_norm": 23.52260398864746, + "learning_rate": 4.139577457028009e-07, + "loss": 1.2468, + "step": 2147 + }, + { + "epoch": 2.1122047244094486, + "grad_norm": 13.40678882598877, + "learning_rate": 4.13203138275349e-07, + "loss": 0.7636, + "step": 2148 + }, + { + "epoch": 2.1131889763779528, + "grad_norm": 12.064020156860352, + "learning_rate": 4.124487402362119e-07, + "loss": 0.3759, + "step": 2149 + }, + { + "epoch": 2.1141732283464565, + "grad_norm": 13.580906867980957, + "learning_rate": 4.116945533710084e-07, + "loss": 0.4563, + "step": 2150 + }, + { + "epoch": 2.1151574803149606, + "grad_norm": 16.30206871032715, + "learning_rate": 4.1094057946485717e-07, + "loss": 0.7122, + "step": 2151 + }, + { + "epoch": 2.1161417322834644, + "grad_norm": 11.582755088806152, + "learning_rate": 4.1018682030237324e-07, + "loss": 0.5501, + "step": 2152 + }, + { + "epoch": 2.1171259842519685, + "grad_norm": 9.636466979980469, + "learning_rate": 4.0943327766766314e-07, + "loss": 0.3651, + "step": 2153 + }, + { + "epoch": 2.1181102362204722, + "grad_norm": 10.754108428955078, + "learning_rate": 4.086799533443208e-07, + "loss": 0.3793, + "step": 2154 + }, + { + "epoch": 2.1190944881889764, + "grad_norm": 13.00988483428955, + "learning_rate": 4.0792684911542373e-07, + "loss": 0.6461, + "step": 2155 + }, + { + "epoch": 2.12007874015748, + "grad_norm": 11.099771499633789, + "learning_rate": 4.071739667635281e-07, + "loss": 0.5457, + "step": 2156 + }, + { + "epoch": 2.1210629921259843, + "grad_norm": 6.3068623542785645, + "learning_rate": 4.064213080706652e-07, + "loss": 0.1725, + "step": 2157 + }, + { + "epoch": 2.122047244094488, + "grad_norm": 6.355545520782471, + "learning_rate": 4.056688748183368e-07, + "loss": 0.227, + "step": 2158 + }, + { + "epoch": 2.123031496062992, + "grad_norm": 14.778956413269043, + "learning_rate": 4.049166687875111e-07, + "loss": 0.529, + "step": 2159 + }, + { + "epoch": 2.1240157480314963, + "grad_norm": 11.180927276611328, + "learning_rate": 4.0416469175861855e-07, + "loss": 0.3569, + "step": 2160 + }, + { + "epoch": 2.125, + "grad_norm": 8.885268211364746, + "learning_rate": 4.034129455115474e-07, + "loss": 0.3879, + "step": 2161 + }, + { + "epoch": 2.1259842519685037, + "grad_norm": 9.933160781860352, + "learning_rate": 4.0266143182563974e-07, + "loss": 0.3999, + "step": 2162 + }, + { + "epoch": 2.126968503937008, + "grad_norm": 16.638206481933594, + "learning_rate": 4.019101524796874e-07, + "loss": 0.3085, + "step": 2163 + }, + { + "epoch": 2.127952755905512, + "grad_norm": 17.310379028320312, + "learning_rate": 4.0115910925192705e-07, + "loss": 0.7374, + "step": 2164 + }, + { + "epoch": 2.1289370078740157, + "grad_norm": 10.217885971069336, + "learning_rate": 4.004083039200369e-07, + "loss": 0.2956, + "step": 2165 + }, + { + "epoch": 2.1299212598425195, + "grad_norm": 10.47181510925293, + "learning_rate": 3.99657738261132e-07, + "loss": 0.703, + "step": 2166 + }, + { + "epoch": 2.1309055118110236, + "grad_norm": 17.75139808654785, + "learning_rate": 3.9890741405176e-07, + "loss": 0.5395, + "step": 2167 + }, + { + "epoch": 2.1318897637795278, + "grad_norm": 12.336260795593262, + "learning_rate": 3.981573330678973e-07, + "loss": 0.5119, + "step": 2168 + }, + { + "epoch": 2.1328740157480315, + "grad_norm": 9.175243377685547, + "learning_rate": 3.9740749708494405e-07, + "loss": 0.3965, + "step": 2169 + }, + { + "epoch": 2.1338582677165356, + "grad_norm": 11.417914390563965, + "learning_rate": 3.96657907877721e-07, + "loss": 0.6092, + "step": 2170 + }, + { + "epoch": 2.1348425196850394, + "grad_norm": 10.03918170928955, + "learning_rate": 3.959085672204647e-07, + "loss": 0.4268, + "step": 2171 + }, + { + "epoch": 2.1358267716535435, + "grad_norm": 15.49741268157959, + "learning_rate": 3.951594768868233e-07, + "loss": 0.5907, + "step": 2172 + }, + { + "epoch": 2.1368110236220472, + "grad_norm": 10.845100402832031, + "learning_rate": 3.9441063864985255e-07, + "loss": 0.4374, + "step": 2173 + }, + { + "epoch": 2.1377952755905514, + "grad_norm": 7.834804058074951, + "learning_rate": 3.936620542820114e-07, + "loss": 0.2971, + "step": 2174 + }, + { + "epoch": 2.138779527559055, + "grad_norm": 10.065025329589844, + "learning_rate": 3.9291372555515794e-07, + "loss": 0.4096, + "step": 2175 + }, + { + "epoch": 2.1397637795275593, + "grad_norm": 10.566422462463379, + "learning_rate": 3.921656542405452e-07, + "loss": 0.627, + "step": 2176 + }, + { + "epoch": 2.140748031496063, + "grad_norm": 10.61729621887207, + "learning_rate": 3.9141784210881695e-07, + "loss": 0.4026, + "step": 2177 + }, + { + "epoch": 2.141732283464567, + "grad_norm": 10.74808120727539, + "learning_rate": 3.9067029093000335e-07, + "loss": 0.4056, + "step": 2178 + }, + { + "epoch": 2.142716535433071, + "grad_norm": 9.267818450927734, + "learning_rate": 3.899230024735171e-07, + "loss": 0.3453, + "step": 2179 + }, + { + "epoch": 2.143700787401575, + "grad_norm": 6.095523357391357, + "learning_rate": 3.8917597850814885e-07, + "loss": 0.2895, + "step": 2180 + }, + { + "epoch": 2.1446850393700787, + "grad_norm": 11.208436012268066, + "learning_rate": 3.884292208020635e-07, + "loss": 0.5194, + "step": 2181 + }, + { + "epoch": 2.145669291338583, + "grad_norm": 15.45287799835205, + "learning_rate": 3.876827311227954e-07, + "loss": 0.8513, + "step": 2182 + }, + { + "epoch": 2.1466535433070866, + "grad_norm": 9.89079761505127, + "learning_rate": 3.8693651123724455e-07, + "loss": 0.4421, + "step": 2183 + }, + { + "epoch": 2.1476377952755907, + "grad_norm": 8.460564613342285, + "learning_rate": 3.861905629116725e-07, + "loss": 0.486, + "step": 2184 + }, + { + "epoch": 2.1486220472440944, + "grad_norm": 12.875810623168945, + "learning_rate": 3.854448879116981e-07, + "loss": 1.1001, + "step": 2185 + }, + { + "epoch": 2.1496062992125986, + "grad_norm": 8.028280258178711, + "learning_rate": 3.84699488002293e-07, + "loss": 0.2871, + "step": 2186 + }, + { + "epoch": 2.1505905511811023, + "grad_norm": 13.88902759552002, + "learning_rate": 3.83954364947778e-07, + "loss": 0.7672, + "step": 2187 + }, + { + "epoch": 2.1515748031496065, + "grad_norm": 12.074015617370605, + "learning_rate": 3.832095205118183e-07, + "loss": 0.8712, + "step": 2188 + }, + { + "epoch": 2.15255905511811, + "grad_norm": 9.419001579284668, + "learning_rate": 3.8246495645742006e-07, + "loss": 0.2191, + "step": 2189 + }, + { + "epoch": 2.1535433070866143, + "grad_norm": 12.924214363098145, + "learning_rate": 3.8172067454692517e-07, + "loss": 0.617, + "step": 2190 + }, + { + "epoch": 2.154527559055118, + "grad_norm": 10.700763702392578, + "learning_rate": 3.8097667654200826e-07, + "loss": 0.4812, + "step": 2191 + }, + { + "epoch": 2.155511811023622, + "grad_norm": 12.375061988830566, + "learning_rate": 3.802329642036716e-07, + "loss": 0.5425, + "step": 2192 + }, + { + "epoch": 2.156496062992126, + "grad_norm": 11.396371841430664, + "learning_rate": 3.7948953929224155e-07, + "loss": 0.6849, + "step": 2193 + }, + { + "epoch": 2.15748031496063, + "grad_norm": 11.016363143920898, + "learning_rate": 3.787464035673641e-07, + "loss": 0.4039, + "step": 2194 + }, + { + "epoch": 2.158464566929134, + "grad_norm": 8.921324729919434, + "learning_rate": 3.780035587880007e-07, + "loss": 0.2696, + "step": 2195 + }, + { + "epoch": 2.159448818897638, + "grad_norm": 12.214681625366211, + "learning_rate": 3.7726100671242414e-07, + "loss": 0.6889, + "step": 2196 + }, + { + "epoch": 2.1604330708661417, + "grad_norm": 12.57135009765625, + "learning_rate": 3.765187490982144e-07, + "loss": 0.5543, + "step": 2197 + }, + { + "epoch": 2.161417322834646, + "grad_norm": 14.028517723083496, + "learning_rate": 3.757767877022545e-07, + "loss": 0.8829, + "step": 2198 + }, + { + "epoch": 2.1624015748031495, + "grad_norm": 15.3035249710083, + "learning_rate": 3.750351242807264e-07, + "loss": 0.7397, + "step": 2199 + }, + { + "epoch": 2.1633858267716537, + "grad_norm": 8.669970512390137, + "learning_rate": 3.7429376058910674e-07, + "loss": 0.4366, + "step": 2200 + }, + { + "epoch": 2.1643700787401574, + "grad_norm": 4.055088520050049, + "learning_rate": 3.735526983821624e-07, + "loss": 0.0933, + "step": 2201 + }, + { + "epoch": 2.1653543307086616, + "grad_norm": 8.365303039550781, + "learning_rate": 3.728119394139472e-07, + "loss": 0.285, + "step": 2202 + }, + { + "epoch": 2.1663385826771653, + "grad_norm": 9.966198921203613, + "learning_rate": 3.7207148543779677e-07, + "loss": 0.5301, + "step": 2203 + }, + { + "epoch": 2.1673228346456694, + "grad_norm": 14.930325508117676, + "learning_rate": 3.713313382063253e-07, + "loss": 0.6857, + "step": 2204 + }, + { + "epoch": 2.168307086614173, + "grad_norm": 11.771857261657715, + "learning_rate": 3.705914994714203e-07, + "loss": 0.7281, + "step": 2205 + }, + { + "epoch": 2.1692913385826773, + "grad_norm": 10.2894287109375, + "learning_rate": 3.6985197098423957e-07, + "loss": 0.4846, + "step": 2206 + }, + { + "epoch": 2.170275590551181, + "grad_norm": 9.719088554382324, + "learning_rate": 3.691127544952065e-07, + "loss": 0.3703, + "step": 2207 + }, + { + "epoch": 2.171259842519685, + "grad_norm": 12.411601066589355, + "learning_rate": 3.6837385175400597e-07, + "loss": 0.589, + "step": 2208 + }, + { + "epoch": 2.172244094488189, + "grad_norm": 10.245353698730469, + "learning_rate": 3.6763526450958005e-07, + "loss": 0.6338, + "step": 2209 + }, + { + "epoch": 2.173228346456693, + "grad_norm": 10.593074798583984, + "learning_rate": 3.668969945101245e-07, + "loss": 0.5374, + "step": 2210 + }, + { + "epoch": 2.1742125984251968, + "grad_norm": 6.299219131469727, + "learning_rate": 3.6615904350308343e-07, + "loss": 0.2392, + "step": 2211 + }, + { + "epoch": 2.175196850393701, + "grad_norm": 6.071145534515381, + "learning_rate": 3.6542141323514666e-07, + "loss": 0.2405, + "step": 2212 + }, + { + "epoch": 2.1761811023622046, + "grad_norm": 15.841425895690918, + "learning_rate": 3.6468410545224436e-07, + "loss": 0.8154, + "step": 2213 + }, + { + "epoch": 2.177165354330709, + "grad_norm": 11.276870727539062, + "learning_rate": 3.6394712189954365e-07, + "loss": 0.6533, + "step": 2214 + }, + { + "epoch": 2.1781496062992125, + "grad_norm": 12.441000938415527, + "learning_rate": 3.632104643214442e-07, + "loss": 0.5278, + "step": 2215 + }, + { + "epoch": 2.1791338582677167, + "grad_norm": 14.340779304504395, + "learning_rate": 3.62474134461574e-07, + "loss": 0.907, + "step": 2216 + }, + { + "epoch": 2.1801181102362204, + "grad_norm": 6.19671106338501, + "learning_rate": 3.6173813406278543e-07, + "loss": 0.1621, + "step": 2217 + }, + { + "epoch": 2.1811023622047245, + "grad_norm": 10.040095329284668, + "learning_rate": 3.610024648671511e-07, + "loss": 0.4738, + "step": 2218 + }, + { + "epoch": 2.1820866141732282, + "grad_norm": 16.7701416015625, + "learning_rate": 3.602671286159594e-07, + "loss": 0.8227, + "step": 2219 + }, + { + "epoch": 2.1830708661417324, + "grad_norm": 15.204830169677734, + "learning_rate": 3.5953212704971096e-07, + "loss": 0.668, + "step": 2220 + }, + { + "epoch": 2.184055118110236, + "grad_norm": 10.306490898132324, + "learning_rate": 3.5879746190811415e-07, + "loss": 0.5098, + "step": 2221 + }, + { + "epoch": 2.1850393700787403, + "grad_norm": 6.649641513824463, + "learning_rate": 3.5806313493008104e-07, + "loss": 0.2362, + "step": 2222 + }, + { + "epoch": 2.186023622047244, + "grad_norm": 13.032939910888672, + "learning_rate": 3.573291478537231e-07, + "loss": 0.8769, + "step": 2223 + }, + { + "epoch": 2.187007874015748, + "grad_norm": 9.772322654724121, + "learning_rate": 3.565955024163477e-07, + "loss": 0.479, + "step": 2224 + }, + { + "epoch": 2.187992125984252, + "grad_norm": 11.886706352233887, + "learning_rate": 3.5586220035445327e-07, + "loss": 0.6157, + "step": 2225 + }, + { + "epoch": 2.188976377952756, + "grad_norm": 8.725859642028809, + "learning_rate": 3.551292434037254e-07, + "loss": 0.3198, + "step": 2226 + }, + { + "epoch": 2.1899606299212597, + "grad_norm": 6.983432292938232, + "learning_rate": 3.54396633299033e-07, + "loss": 0.2217, + "step": 2227 + }, + { + "epoch": 2.190944881889764, + "grad_norm": 10.190155982971191, + "learning_rate": 3.5366437177442394e-07, + "loss": 0.5194, + "step": 2228 + }, + { + "epoch": 2.1919291338582676, + "grad_norm": 10.641023635864258, + "learning_rate": 3.529324605631211e-07, + "loss": 0.413, + "step": 2229 + }, + { + "epoch": 2.1929133858267718, + "grad_norm": 7.027321815490723, + "learning_rate": 3.522009013975184e-07, + "loss": 0.242, + "step": 2230 + }, + { + "epoch": 2.1938976377952755, + "grad_norm": 11.656682968139648, + "learning_rate": 3.514696960091756e-07, + "loss": 0.4697, + "step": 2231 + }, + { + "epoch": 2.1948818897637796, + "grad_norm": 8.944205284118652, + "learning_rate": 3.5073884612881616e-07, + "loss": 0.4982, + "step": 2232 + }, + { + "epoch": 2.1958661417322833, + "grad_norm": 9.124181747436523, + "learning_rate": 3.500083534863215e-07, + "loss": 0.4052, + "step": 2233 + }, + { + "epoch": 2.1968503937007875, + "grad_norm": 5.858971118927002, + "learning_rate": 3.492782198107275e-07, + "loss": 0.1339, + "step": 2234 + }, + { + "epoch": 2.197834645669291, + "grad_norm": 11.764811515808105, + "learning_rate": 3.485484468302207e-07, + "loss": 0.488, + "step": 2235 + }, + { + "epoch": 2.1988188976377954, + "grad_norm": 14.925984382629395, + "learning_rate": 3.4781903627213347e-07, + "loss": 0.7147, + "step": 2236 + }, + { + "epoch": 2.199803149606299, + "grad_norm": 10.232291221618652, + "learning_rate": 3.470899898629408e-07, + "loss": 0.6582, + "step": 2237 + }, + { + "epoch": 2.2007874015748032, + "grad_norm": 12.924386024475098, + "learning_rate": 3.463613093282552e-07, + "loss": 0.7598, + "step": 2238 + }, + { + "epoch": 2.201771653543307, + "grad_norm": 9.899399757385254, + "learning_rate": 3.456329963928237e-07, + "loss": 0.4642, + "step": 2239 + }, + { + "epoch": 2.202755905511811, + "grad_norm": 9.124329566955566, + "learning_rate": 3.4490505278052295e-07, + "loss": 0.4712, + "step": 2240 + }, + { + "epoch": 2.203740157480315, + "grad_norm": 7.940446376800537, + "learning_rate": 3.4417748021435563e-07, + "loss": 0.3164, + "step": 2241 + }, + { + "epoch": 2.204724409448819, + "grad_norm": 9.96937370300293, + "learning_rate": 3.4345028041644595e-07, + "loss": 0.4444, + "step": 2242 + }, + { + "epoch": 2.2057086614173227, + "grad_norm": 9.700082778930664, + "learning_rate": 3.427234551080358e-07, + "loss": 0.4983, + "step": 2243 + }, + { + "epoch": 2.206692913385827, + "grad_norm": 14.01784896850586, + "learning_rate": 3.41997006009481e-07, + "loss": 0.5059, + "step": 2244 + }, + { + "epoch": 2.2076771653543306, + "grad_norm": 19.83002281188965, + "learning_rate": 3.4127093484024656e-07, + "loss": 1.0876, + "step": 2245 + }, + { + "epoch": 2.2086614173228347, + "grad_norm": 5.920734882354736, + "learning_rate": 3.4054524331890316e-07, + "loss": 0.1743, + "step": 2246 + }, + { + "epoch": 2.2096456692913384, + "grad_norm": 11.091670036315918, + "learning_rate": 3.3981993316312277e-07, + "loss": 0.7035, + "step": 2247 + }, + { + "epoch": 2.2106299212598426, + "grad_norm": 8.12314224243164, + "learning_rate": 3.390950060896746e-07, + "loss": 0.2316, + "step": 2248 + }, + { + "epoch": 2.2116141732283463, + "grad_norm": 4.970713138580322, + "learning_rate": 3.3837046381442135e-07, + "loss": 0.1159, + "step": 2249 + }, + { + "epoch": 2.2125984251968505, + "grad_norm": 8.45492172241211, + "learning_rate": 3.376463080523147e-07, + "loss": 0.3093, + "step": 2250 + }, + { + "epoch": 2.213582677165354, + "grad_norm": 8.577178955078125, + "learning_rate": 3.369225405173918e-07, + "loss": 0.2272, + "step": 2251 + }, + { + "epoch": 2.2145669291338583, + "grad_norm": 13.585640907287598, + "learning_rate": 3.361991629227704e-07, + "loss": 0.8034, + "step": 2252 + }, + { + "epoch": 2.215551181102362, + "grad_norm": 11.29826831817627, + "learning_rate": 3.354761769806456e-07, + "loss": 0.4926, + "step": 2253 + }, + { + "epoch": 2.216535433070866, + "grad_norm": 8.566338539123535, + "learning_rate": 3.3475358440228547e-07, + "loss": 0.2633, + "step": 2254 + }, + { + "epoch": 2.21751968503937, + "grad_norm": 7.11183500289917, + "learning_rate": 3.340313868980271e-07, + "loss": 0.2974, + "step": 2255 + }, + { + "epoch": 2.218503937007874, + "grad_norm": 12.137798309326172, + "learning_rate": 3.3330958617727213e-07, + "loss": 0.6339, + "step": 2256 + }, + { + "epoch": 2.219488188976378, + "grad_norm": 13.24455738067627, + "learning_rate": 3.325881839484833e-07, + "loss": 0.4699, + "step": 2257 + }, + { + "epoch": 2.220472440944882, + "grad_norm": 10.560152053833008, + "learning_rate": 3.318671819191802e-07, + "loss": 0.5012, + "step": 2258 + }, + { + "epoch": 2.2214566929133857, + "grad_norm": 13.34184741973877, + "learning_rate": 3.3114658179593487e-07, + "loss": 0.5572, + "step": 2259 + }, + { + "epoch": 2.22244094488189, + "grad_norm": 16.4458065032959, + "learning_rate": 3.304263852843683e-07, + "loss": 0.9996, + "step": 2260 + }, + { + "epoch": 2.2234251968503935, + "grad_norm": 11.808619499206543, + "learning_rate": 3.297065940891462e-07, + "loss": 0.7209, + "step": 2261 + }, + { + "epoch": 2.2244094488188977, + "grad_norm": 9.634578704833984, + "learning_rate": 3.289872099139746e-07, + "loss": 0.3005, + "step": 2262 + }, + { + "epoch": 2.2253937007874014, + "grad_norm": 11.289079666137695, + "learning_rate": 3.282682344615966e-07, + "loss": 0.7379, + "step": 2263 + }, + { + "epoch": 2.2263779527559056, + "grad_norm": 5.66681432723999, + "learning_rate": 3.275496694337874e-07, + "loss": 0.1233, + "step": 2264 + }, + { + "epoch": 2.2273622047244093, + "grad_norm": 10.479454040527344, + "learning_rate": 3.2683151653135104e-07, + "loss": 0.4234, + "step": 2265 + }, + { + "epoch": 2.2283464566929134, + "grad_norm": 8.537384986877441, + "learning_rate": 3.261137774541161e-07, + "loss": 0.2912, + "step": 2266 + }, + { + "epoch": 2.229330708661417, + "grad_norm": 14.457018852233887, + "learning_rate": 3.2539645390093155e-07, + "loss": 0.5686, + "step": 2267 + }, + { + "epoch": 2.2303149606299213, + "grad_norm": 8.997966766357422, + "learning_rate": 3.2467954756966276e-07, + "loss": 0.3039, + "step": 2268 + }, + { + "epoch": 2.231299212598425, + "grad_norm": 9.181035995483398, + "learning_rate": 3.2396306015718775e-07, + "loss": 0.3685, + "step": 2269 + }, + { + "epoch": 2.232283464566929, + "grad_norm": 10.959527969360352, + "learning_rate": 3.2324699335939273e-07, + "loss": 0.4984, + "step": 2270 + }, + { + "epoch": 2.233267716535433, + "grad_norm": 6.941189765930176, + "learning_rate": 3.225313488711686e-07, + "loss": 0.2326, + "step": 2271 + }, + { + "epoch": 2.234251968503937, + "grad_norm": 13.026391983032227, + "learning_rate": 3.2181612838640673e-07, + "loss": 0.5876, + "step": 2272 + }, + { + "epoch": 2.2352362204724407, + "grad_norm": 10.644556999206543, + "learning_rate": 3.2110133359799457e-07, + "loss": 0.4002, + "step": 2273 + }, + { + "epoch": 2.236220472440945, + "grad_norm": 12.20460319519043, + "learning_rate": 3.203869661978121e-07, + "loss": 0.5853, + "step": 2274 + }, + { + "epoch": 2.2372047244094486, + "grad_norm": 8.994668006896973, + "learning_rate": 3.196730278767277e-07, + "loss": 0.2947, + "step": 2275 + }, + { + "epoch": 2.2381889763779528, + "grad_norm": 12.746809959411621, + "learning_rate": 3.189595203245944e-07, + "loss": 0.5385, + "step": 2276 + }, + { + "epoch": 2.2391732283464565, + "grad_norm": 8.311347961425781, + "learning_rate": 3.182464452302452e-07, + "loss": 0.3142, + "step": 2277 + }, + { + "epoch": 2.2401574803149606, + "grad_norm": 9.598487854003906, + "learning_rate": 3.175338042814898e-07, + "loss": 0.383, + "step": 2278 + }, + { + "epoch": 2.2411417322834644, + "grad_norm": 11.807975769042969, + "learning_rate": 3.168215991651101e-07, + "loss": 0.4688, + "step": 2279 + }, + { + "epoch": 2.2421259842519685, + "grad_norm": 8.984016418457031, + "learning_rate": 3.1610983156685664e-07, + "loss": 0.4124, + "step": 2280 + }, + { + "epoch": 2.2431102362204722, + "grad_norm": 15.54113483428955, + "learning_rate": 3.153985031714442e-07, + "loss": 0.538, + "step": 2281 + }, + { + "epoch": 2.2440944881889764, + "grad_norm": 12.840856552124023, + "learning_rate": 3.14687615662548e-07, + "loss": 0.6009, + "step": 2282 + }, + { + "epoch": 2.2450787401574805, + "grad_norm": 7.0070271492004395, + "learning_rate": 3.139771707227997e-07, + "loss": 0.2344, + "step": 2283 + }, + { + "epoch": 2.2460629921259843, + "grad_norm": 10.138955116271973, + "learning_rate": 3.1326717003378365e-07, + "loss": 0.3602, + "step": 2284 + }, + { + "epoch": 2.247047244094488, + "grad_norm": 11.75495719909668, + "learning_rate": 3.1255761527603233e-07, + "loss": 0.7008, + "step": 2285 + }, + { + "epoch": 2.248031496062992, + "grad_norm": 10.067399978637695, + "learning_rate": 3.1184850812902304e-07, + "loss": 0.2619, + "step": 2286 + }, + { + "epoch": 2.2490157480314963, + "grad_norm": 13.368266105651855, + "learning_rate": 3.111398502711733e-07, + "loss": 0.5225, + "step": 2287 + }, + { + "epoch": 2.25, + "grad_norm": 5.223156452178955, + "learning_rate": 3.104316433798374e-07, + "loss": 0.1657, + "step": 2288 + }, + { + "epoch": 2.2509842519685037, + "grad_norm": 10.531564712524414, + "learning_rate": 3.0972388913130215e-07, + "loss": 0.3965, + "step": 2289 + }, + { + "epoch": 2.251968503937008, + "grad_norm": 14.430265426635742, + "learning_rate": 3.0901658920078314e-07, + "loss": 0.8387, + "step": 2290 + }, + { + "epoch": 2.252952755905512, + "grad_norm": 13.575235366821289, + "learning_rate": 3.083097452624203e-07, + "loss": 0.6505, + "step": 2291 + }, + { + "epoch": 2.2539370078740157, + "grad_norm": 15.854157447814941, + "learning_rate": 3.076033589892745e-07, + "loss": 0.8453, + "step": 2292 + }, + { + "epoch": 2.2549212598425195, + "grad_norm": 10.565400123596191, + "learning_rate": 3.0689743205332345e-07, + "loss": 0.3865, + "step": 2293 + }, + { + "epoch": 2.2559055118110236, + "grad_norm": 11.114446640014648, + "learning_rate": 3.061919661254571e-07, + "loss": 0.7424, + "step": 2294 + }, + { + "epoch": 2.2568897637795278, + "grad_norm": 9.461173057556152, + "learning_rate": 3.054869628754748e-07, + "loss": 0.2301, + "step": 2295 + }, + { + "epoch": 2.2568897637795278, + "eval_Qnli-dev_cosine_accuracy": 0.67578125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8218872547149658, + "eval_Qnli-dev_cosine_ap": 0.6982356648607996, + "eval_Qnli-dev_cosine_f1": 0.6914378029079159, + "eval_Qnli-dev_cosine_f1_threshold": 0.7249837517738342, + "eval_Qnli-dev_cosine_precision": 0.5587467362924282, + "eval_Qnli-dev_cosine_recall": 0.9067796610169492, + "eval_Qnli-dev_dot_accuracy": 0.681640625, + "eval_Qnli-dev_dot_accuracy_threshold": 619.56884765625, + "eval_Qnli-dev_dot_ap": 0.694508632045775, + "eval_Qnli-dev_dot_f1": 0.6903225806451613, + "eval_Qnli-dev_dot_f1_threshold": 551.586669921875, + "eval_Qnli-dev_dot_precision": 0.5572916666666666, + "eval_Qnli-dev_dot_recall": 0.9067796610169492, + "eval_Qnli-dev_euclidean_accuracy": 0.677734375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 16.517976760864258, + "eval_Qnli-dev_euclidean_ap": 0.698306251672151, + "eval_Qnli-dev_euclidean_f1": 0.6914378029079159, + "eval_Qnli-dev_euclidean_f1_threshold": 20.45012664794922, + "eval_Qnli-dev_euclidean_precision": 0.5587467362924282, + "eval_Qnli-dev_euclidean_recall": 0.9067796610169492, + "eval_Qnli-dev_manhattan_accuracy": 0.685546875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 360.5780029296875, + "eval_Qnli-dev_manhattan_ap": 0.709700437823957, + "eval_Qnli-dev_manhattan_f1": 0.6902086677367576, + "eval_Qnli-dev_manhattan_f1_threshold": 433.43255615234375, + "eval_Qnli-dev_manhattan_precision": 0.5555555555555556, + "eval_Qnli-dev_manhattan_recall": 0.9110169491525424, + "eval_Qnli-dev_max_accuracy": 0.685546875, + "eval_Qnli-dev_max_accuracy_threshold": 619.56884765625, + "eval_Qnli-dev_max_ap": 0.709700437823957, + "eval_Qnli-dev_max_f1": 0.6914378029079159, + "eval_Qnli-dev_max_f1_threshold": 551.586669921875, + "eval_Qnli-dev_max_precision": 0.5587467362924282, + "eval_Qnli-dev_max_recall": 0.9110169491525424, + "eval_allNLI-dev_cosine_accuracy": 0.693359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9163593053817749, + "eval_allNLI-dev_cosine_ap": 0.5156280201012031, + "eval_allNLI-dev_cosine_f1": 0.5478424015009381, + "eval_allNLI-dev_cosine_f1_threshold": 0.7880935668945312, + "eval_allNLI-dev_cosine_precision": 0.40555555555555556, + "eval_allNLI-dev_cosine_recall": 0.8439306358381503, + "eval_allNLI-dev_dot_accuracy": 0.693359375, + "eval_allNLI-dev_dot_accuracy_threshold": 698.921142578125, + "eval_allNLI-dev_dot_ap": 0.5147433660036644, + "eval_allNLI-dev_dot_f1": 0.5457943925233645, + "eval_allNLI-dev_dot_f1_threshold": 599.6630859375, + "eval_allNLI-dev_dot_precision": 0.40331491712707185, + "eval_allNLI-dev_dot_recall": 0.8439306358381503, + "eval_allNLI-dev_euclidean_accuracy": 0.693359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.29690933227539, + "eval_allNLI-dev_euclidean_ap": 0.5155161734618426, + "eval_allNLI-dev_euclidean_f1": 0.5478424015009381, + "eval_allNLI-dev_euclidean_f1_threshold": 17.960678100585938, + "eval_allNLI-dev_euclidean_precision": 0.40555555555555556, + "eval_allNLI-dev_euclidean_recall": 0.8439306358381503, + "eval_allNLI-dev_manhattan_accuracy": 0.69140625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 247.3362274169922, + "eval_allNLI-dev_manhattan_ap": 0.5162044992927799, + "eval_allNLI-dev_manhattan_f1": 0.5513307984790875, + "eval_allNLI-dev_manhattan_f1_threshold": 376.1918640136719, + "eval_allNLI-dev_manhattan_precision": 0.41076487252124644, + "eval_allNLI-dev_manhattan_recall": 0.838150289017341, + "eval_allNLI-dev_max_accuracy": 0.693359375, + "eval_allNLI-dev_max_accuracy_threshold": 698.921142578125, + "eval_allNLI-dev_max_ap": 0.5162044992927799, + "eval_allNLI-dev_max_f1": 0.5513307984790875, + "eval_allNLI-dev_max_f1_threshold": 599.6630859375, + "eval_allNLI-dev_max_precision": 0.41076487252124644, + "eval_allNLI-dev_max_recall": 0.8439306358381503, + "eval_loss": 0.9295917749404907, + "eval_runtime": 54.1496, + "eval_samples_per_second": 30.73, + "eval_sequential_score": 0.709700437823957, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.6552902390696351, + "eval_sts-test_pearson_dot": 0.6546625093787101, + "eval_sts-test_pearson_euclidean": 0.6686392487670392, + "eval_sts-test_pearson_manhattan": 0.6720928689273424, + "eval_sts-test_pearson_max": 0.6720928689273424, + "eval_sts-test_spearman_cosine": 0.6644033613533696, + "eval_sts-test_spearman_dot": 0.6645467891461292, + "eval_sts-test_spearman_euclidean": 0.6643764210940122, + "eval_sts-test_spearman_manhattan": 0.6672981861989953, + "eval_sts-test_spearman_max": 0.6672981861989953, + "step": 2295 + }, + { + "epoch": 2.2578740157480315, + "grad_norm": 5.400763988494873, + "learning_rate": 3.047824239720804e-07, + "loss": 0.128, + "step": 2296 + }, + { + "epoch": 2.258858267716535, + "grad_norm": 14.74106502532959, + "learning_rate": 3.0407835108287885e-07, + "loss": 0.7932, + "step": 2297 + }, + { + "epoch": 2.2598425196850394, + "grad_norm": 9.304885864257812, + "learning_rate": 3.03374745874372e-07, + "loss": 0.3471, + "step": 2298 + }, + { + "epoch": 2.2608267716535435, + "grad_norm": 13.81615161895752, + "learning_rate": 3.0267161001195493e-07, + "loss": 0.5931, + "step": 2299 + }, + { + "epoch": 2.2618110236220472, + "grad_norm": 12.656591415405273, + "learning_rate": 3.019689451599114e-07, + "loss": 0.3816, + "step": 2300 + }, + { + "epoch": 2.262795275590551, + "grad_norm": 8.846150398254395, + "learning_rate": 3.0126675298141056e-07, + "loss": 0.2919, + "step": 2301 + }, + { + "epoch": 2.263779527559055, + "grad_norm": 5.267794609069824, + "learning_rate": 3.005650351385029e-07, + "loss": 0.1231, + "step": 2302 + }, + { + "epoch": 2.2647637795275593, + "grad_norm": 9.467386245727539, + "learning_rate": 2.998637932921159e-07, + "loss": 0.3372, + "step": 2303 + }, + { + "epoch": 2.265748031496063, + "grad_norm": 8.390028953552246, + "learning_rate": 2.991630291020506e-07, + "loss": 0.2717, + "step": 2304 + }, + { + "epoch": 2.266732283464567, + "grad_norm": 9.141326904296875, + "learning_rate": 2.9846274422697734e-07, + "loss": 0.2641, + "step": 2305 + }, + { + "epoch": 2.267716535433071, + "grad_norm": 7.371496200561523, + "learning_rate": 2.9776294032443184e-07, + "loss": 0.1901, + "step": 2306 + }, + { + "epoch": 2.268700787401575, + "grad_norm": 9.045897483825684, + "learning_rate": 2.9706361905081164e-07, + "loss": 0.409, + "step": 2307 + }, + { + "epoch": 2.2696850393700787, + "grad_norm": 18.923290252685547, + "learning_rate": 2.9636478206137174e-07, + "loss": 0.9497, + "step": 2308 + }, + { + "epoch": 2.270669291338583, + "grad_norm": 16.945598602294922, + "learning_rate": 2.95666431010221e-07, + "loss": 0.7417, + "step": 2309 + }, + { + "epoch": 2.2716535433070866, + "grad_norm": 8.197480201721191, + "learning_rate": 2.949685675503179e-07, + "loss": 0.3104, + "step": 2310 + }, + { + "epoch": 2.2726377952755907, + "grad_norm": 8.451308250427246, + "learning_rate": 2.9427119333346706e-07, + "loss": 0.3535, + "step": 2311 + }, + { + "epoch": 2.2736220472440944, + "grad_norm": 9.730972290039062, + "learning_rate": 2.935743100103149e-07, + "loss": 0.2776, + "step": 2312 + }, + { + "epoch": 2.2746062992125986, + "grad_norm": 9.99947452545166, + "learning_rate": 2.9287791923034593e-07, + "loss": 0.4681, + "step": 2313 + }, + { + "epoch": 2.2755905511811023, + "grad_norm": 9.144072532653809, + "learning_rate": 2.921820226418793e-07, + "loss": 0.3353, + "step": 2314 + }, + { + "epoch": 2.2765748031496065, + "grad_norm": 12.268770217895508, + "learning_rate": 2.9148662189206344e-07, + "loss": 0.3965, + "step": 2315 + }, + { + "epoch": 2.27755905511811, + "grad_norm": 11.276354789733887, + "learning_rate": 2.907917186268739e-07, + "loss": 0.4975, + "step": 2316 + }, + { + "epoch": 2.2785433070866143, + "grad_norm": 12.892839431762695, + "learning_rate": 2.900973144911088e-07, + "loss": 0.6786, + "step": 2317 + }, + { + "epoch": 2.279527559055118, + "grad_norm": 6.2306342124938965, + "learning_rate": 2.894034111283843e-07, + "loss": 0.3268, + "step": 2318 + }, + { + "epoch": 2.280511811023622, + "grad_norm": 10.548524856567383, + "learning_rate": 2.8871001018113183e-07, + "loss": 0.3992, + "step": 2319 + }, + { + "epoch": 2.281496062992126, + "grad_norm": 7.72267484664917, + "learning_rate": 2.880171132905932e-07, + "loss": 0.2394, + "step": 2320 + }, + { + "epoch": 2.28248031496063, + "grad_norm": 13.049686431884766, + "learning_rate": 2.8732472209681733e-07, + "loss": 0.4223, + "step": 2321 + }, + { + "epoch": 2.283464566929134, + "grad_norm": 3.6191823482513428, + "learning_rate": 2.866328382386562e-07, + "loss": 0.096, + "step": 2322 + }, + { + "epoch": 2.284448818897638, + "grad_norm": 12.548807144165039, + "learning_rate": 2.859414633537609e-07, + "loss": 0.6566, + "step": 2323 + }, + { + "epoch": 2.2854330708661417, + "grad_norm": 16.991758346557617, + "learning_rate": 2.8525059907857786e-07, + "loss": 0.9653, + "step": 2324 + }, + { + "epoch": 2.286417322834646, + "grad_norm": 8.746471405029297, + "learning_rate": 2.8456024704834483e-07, + "loss": 0.4092, + "step": 2325 + }, + { + "epoch": 2.2874015748031495, + "grad_norm": 6.317548751831055, + "learning_rate": 2.8387040889708715e-07, + "loss": 0.1846, + "step": 2326 + }, + { + "epoch": 2.2883858267716537, + "grad_norm": 13.521329879760742, + "learning_rate": 2.8318108625761395e-07, + "loss": 0.6888, + "step": 2327 + }, + { + "epoch": 2.2893700787401574, + "grad_norm": 18.330942153930664, + "learning_rate": 2.8249228076151396e-07, + "loss": 0.7732, + "step": 2328 + }, + { + "epoch": 2.2903543307086616, + "grad_norm": 5.730623245239258, + "learning_rate": 2.81803994039152e-07, + "loss": 0.1397, + "step": 2329 + }, + { + "epoch": 2.2913385826771653, + "grad_norm": 9.373048782348633, + "learning_rate": 2.8111622771966523e-07, + "loss": 0.5713, + "step": 2330 + }, + { + "epoch": 2.2923228346456694, + "grad_norm": 8.271077156066895, + "learning_rate": 2.804289834309583e-07, + "loss": 0.3005, + "step": 2331 + }, + { + "epoch": 2.293307086614173, + "grad_norm": 10.701303482055664, + "learning_rate": 2.797422627997009e-07, + "loss": 0.5604, + "step": 2332 + }, + { + "epoch": 2.2942913385826773, + "grad_norm": 10.093025207519531, + "learning_rate": 2.7905606745132307e-07, + "loss": 0.4826, + "step": 2333 + }, + { + "epoch": 2.295275590551181, + "grad_norm": 15.213040351867676, + "learning_rate": 2.7837039901001155e-07, + "loss": 0.747, + "step": 2334 + }, + { + "epoch": 2.296259842519685, + "grad_norm": 10.684081077575684, + "learning_rate": 2.776852590987061e-07, + "loss": 0.4473, + "step": 2335 + }, + { + "epoch": 2.297244094488189, + "grad_norm": 10.946547508239746, + "learning_rate": 2.7700064933909474e-07, + "loss": 0.5294, + "step": 2336 + }, + { + "epoch": 2.298228346456693, + "grad_norm": 9.167648315429688, + "learning_rate": 2.7631657135161156e-07, + "loss": 0.439, + "step": 2337 + }, + { + "epoch": 2.2992125984251968, + "grad_norm": 7.027797222137451, + "learning_rate": 2.756330267554316e-07, + "loss": 0.2158, + "step": 2338 + }, + { + "epoch": 2.300196850393701, + "grad_norm": 9.155388832092285, + "learning_rate": 2.749500171684673e-07, + "loss": 0.4406, + "step": 2339 + }, + { + "epoch": 2.3011811023622046, + "grad_norm": 7.610925197601318, + "learning_rate": 2.7426754420736485e-07, + "loss": 0.219, + "step": 2340 + }, + { + "epoch": 2.302165354330709, + "grad_norm": 12.610198974609375, + "learning_rate": 2.735856094875003e-07, + "loss": 0.4623, + "step": 2341 + }, + { + "epoch": 2.3031496062992125, + "grad_norm": 10.615156173706055, + "learning_rate": 2.7290421462297574e-07, + "loss": 0.2466, + "step": 2342 + }, + { + "epoch": 2.3041338582677167, + "grad_norm": 12.127360343933105, + "learning_rate": 2.7222336122661535e-07, + "loss": 0.6863, + "step": 2343 + }, + { + "epoch": 2.3051181102362204, + "grad_norm": 9.351911544799805, + "learning_rate": 2.715430509099618e-07, + "loss": 0.3583, + "step": 2344 + }, + { + "epoch": 2.3061023622047245, + "grad_norm": 11.017844200134277, + "learning_rate": 2.7086328528327215e-07, + "loss": 0.5925, + "step": 2345 + }, + { + "epoch": 2.3070866141732282, + "grad_norm": 12.029109954833984, + "learning_rate": 2.7018406595551434e-07, + "loss": 0.552, + "step": 2346 + }, + { + "epoch": 2.3080708661417324, + "grad_norm": 13.67851448059082, + "learning_rate": 2.695053945343633e-07, + "loss": 0.8135, + "step": 2347 + }, + { + "epoch": 2.309055118110236, + "grad_norm": 12.758377075195312, + "learning_rate": 2.6882727262619696e-07, + "loss": 0.5165, + "step": 2348 + }, + { + "epoch": 2.3100393700787403, + "grad_norm": 9.64584732055664, + "learning_rate": 2.681497018360927e-07, + "loss": 0.3632, + "step": 2349 + }, + { + "epoch": 2.311023622047244, + "grad_norm": 10.910879135131836, + "learning_rate": 2.6747268376782334e-07, + "loss": 0.4364, + "step": 2350 + }, + { + "epoch": 2.312007874015748, + "grad_norm": 7.449069023132324, + "learning_rate": 2.6679622002385354e-07, + "loss": 0.2048, + "step": 2351 + }, + { + "epoch": 2.312992125984252, + "grad_norm": 9.218055725097656, + "learning_rate": 2.6612031220533574e-07, + "loss": 0.5, + "step": 2352 + }, + { + "epoch": 2.313976377952756, + "grad_norm": 10.417744636535645, + "learning_rate": 2.654449619121067e-07, + "loss": 0.4657, + "step": 2353 + }, + { + "epoch": 2.3149606299212597, + "grad_norm": 17.5401554107666, + "learning_rate": 2.647701707426835e-07, + "loss": 0.8455, + "step": 2354 + }, + { + "epoch": 2.315944881889764, + "grad_norm": 11.498607635498047, + "learning_rate": 2.6409594029425986e-07, + "loss": 0.703, + "step": 2355 + }, + { + "epoch": 2.3169291338582676, + "grad_norm": 10.016690254211426, + "learning_rate": 2.6342227216270205e-07, + "loss": 0.4548, + "step": 2356 + }, + { + "epoch": 2.3179133858267718, + "grad_norm": 8.504423141479492, + "learning_rate": 2.6274916794254554e-07, + "loss": 0.3097, + "step": 2357 + }, + { + "epoch": 2.3188976377952755, + "grad_norm": 7.540285110473633, + "learning_rate": 2.6207662922699125e-07, + "loss": 0.2965, + "step": 2358 + }, + { + "epoch": 2.3198818897637796, + "grad_norm": 8.403088569641113, + "learning_rate": 2.6140465760790126e-07, + "loss": 0.2303, + "step": 2359 + }, + { + "epoch": 2.3208661417322833, + "grad_norm": 8.636588096618652, + "learning_rate": 2.6073325467579553e-07, + "loss": 0.2937, + "step": 2360 + }, + { + "epoch": 2.3218503937007875, + "grad_norm": 10.3641996383667, + "learning_rate": 2.60062422019848e-07, + "loss": 0.6068, + "step": 2361 + }, + { + "epoch": 2.322834645669291, + "grad_norm": 5.669926166534424, + "learning_rate": 2.5939216122788267e-07, + "loss": 0.169, + "step": 2362 + }, + { + "epoch": 2.3238188976377954, + "grad_norm": 6.79951810836792, + "learning_rate": 2.5872247388637e-07, + "loss": 0.1574, + "step": 2363 + }, + { + "epoch": 2.324803149606299, + "grad_norm": 8.884796142578125, + "learning_rate": 2.5805336158042327e-07, + "loss": 0.3442, + "step": 2364 + }, + { + "epoch": 2.3257874015748032, + "grad_norm": 6.176455974578857, + "learning_rate": 2.573848258937945e-07, + "loss": 0.2219, + "step": 2365 + }, + { + "epoch": 2.326771653543307, + "grad_norm": 5.007375240325928, + "learning_rate": 2.56716868408871e-07, + "loss": 0.1165, + "step": 2366 + }, + { + "epoch": 2.327755905511811, + "grad_norm": 12.076241493225098, + "learning_rate": 2.5604949070667156e-07, + "loss": 0.3702, + "step": 2367 + }, + { + "epoch": 2.328740157480315, + "grad_norm": 9.106301307678223, + "learning_rate": 2.553826943668423e-07, + "loss": 0.3726, + "step": 2368 + }, + { + "epoch": 2.329724409448819, + "grad_norm": 10.722580909729004, + "learning_rate": 2.5471648096765365e-07, + "loss": 0.5213, + "step": 2369 + }, + { + "epoch": 2.3307086614173227, + "grad_norm": 9.431106567382812, + "learning_rate": 2.5405085208599623e-07, + "loss": 0.3472, + "step": 2370 + }, + { + "epoch": 2.331692913385827, + "grad_norm": 7.851450443267822, + "learning_rate": 2.53385809297377e-07, + "loss": 0.3008, + "step": 2371 + }, + { + "epoch": 2.3326771653543306, + "grad_norm": 17.107131958007812, + "learning_rate": 2.527213541759156e-07, + "loss": 0.7413, + "step": 2372 + }, + { + "epoch": 2.3336614173228347, + "grad_norm": 10.417817115783691, + "learning_rate": 2.520574882943411e-07, + "loss": 0.6197, + "step": 2373 + }, + { + "epoch": 2.3346456692913384, + "grad_norm": 7.435096263885498, + "learning_rate": 2.5139421322398734e-07, + "loss": 0.2909, + "step": 2374 + }, + { + "epoch": 2.3356299212598426, + "grad_norm": 9.948948860168457, + "learning_rate": 2.5073153053479025e-07, + "loss": 0.383, + "step": 2375 + }, + { + "epoch": 2.3366141732283463, + "grad_norm": 15.071062088012695, + "learning_rate": 2.5006944179528326e-07, + "loss": 0.6285, + "step": 2376 + }, + { + "epoch": 2.3375984251968505, + "grad_norm": 15.67939281463623, + "learning_rate": 2.494079485725939e-07, + "loss": 0.5903, + "step": 2377 + }, + { + "epoch": 2.338582677165354, + "grad_norm": 5.176926612854004, + "learning_rate": 2.4874705243244044e-07, + "loss": 0.3165, + "step": 2378 + }, + { + "epoch": 2.3395669291338583, + "grad_norm": 11.611159324645996, + "learning_rate": 2.480867549391278e-07, + "loss": 0.4497, + "step": 2379 + }, + { + "epoch": 2.340551181102362, + "grad_norm": 11.762500762939453, + "learning_rate": 2.4742705765554384e-07, + "loss": 0.6278, + "step": 2380 + }, + { + "epoch": 2.341535433070866, + "grad_norm": 9.781825065612793, + "learning_rate": 2.467679621431558e-07, + "loss": 0.4037, + "step": 2381 + }, + { + "epoch": 2.34251968503937, + "grad_norm": 9.960203170776367, + "learning_rate": 2.461094699620066e-07, + "loss": 0.555, + "step": 2382 + }, + { + "epoch": 2.343503937007874, + "grad_norm": 6.709632873535156, + "learning_rate": 2.454515826707111e-07, + "loss": 0.1986, + "step": 2383 + }, + { + "epoch": 2.344488188976378, + "grad_norm": 13.113850593566895, + "learning_rate": 2.4479430182645237e-07, + "loss": 0.6334, + "step": 2384 + }, + { + "epoch": 2.345472440944882, + "grad_norm": 9.094974517822266, + "learning_rate": 2.441376289849781e-07, + "loss": 0.4468, + "step": 2385 + }, + { + "epoch": 2.3464566929133857, + "grad_norm": 10.722009658813477, + "learning_rate": 2.4348156570059685e-07, + "loss": 0.3903, + "step": 2386 + }, + { + "epoch": 2.34744094488189, + "grad_norm": 7.4452223777771, + "learning_rate": 2.428261135261743e-07, + "loss": 0.4445, + "step": 2387 + }, + { + "epoch": 2.3484251968503935, + "grad_norm": 21.786439895629883, + "learning_rate": 2.4217127401312993e-07, + "loss": 1.0605, + "step": 2388 + }, + { + "epoch": 2.3494094488188977, + "grad_norm": 10.825016975402832, + "learning_rate": 2.4151704871143276e-07, + "loss": 0.4604, + "step": 2389 + }, + { + "epoch": 2.3503937007874014, + "grad_norm": 10.936182975769043, + "learning_rate": 2.4086343916959827e-07, + "loss": 0.5446, + "step": 2390 + }, + { + "epoch": 2.3513779527559056, + "grad_norm": 4.6355462074279785, + "learning_rate": 2.402104469346844e-07, + "loss": 0.127, + "step": 2391 + }, + { + "epoch": 2.3523622047244093, + "grad_norm": 8.731232643127441, + "learning_rate": 2.3955807355228767e-07, + "loss": 0.457, + "step": 2392 + }, + { + "epoch": 2.3533464566929134, + "grad_norm": 7.753507614135742, + "learning_rate": 2.3890632056654027e-07, + "loss": 0.2166, + "step": 2393 + }, + { + "epoch": 2.354330708661417, + "grad_norm": 10.839706420898438, + "learning_rate": 2.3825518952010577e-07, + "loss": 0.4146, + "step": 2394 + }, + { + "epoch": 2.3553149606299213, + "grad_norm": 12.403355598449707, + "learning_rate": 2.3760468195417546e-07, + "loss": 0.6237, + "step": 2395 + }, + { + "epoch": 2.356299212598425, + "grad_norm": 13.206016540527344, + "learning_rate": 2.3695479940846513e-07, + "loss": 0.6888, + "step": 2396 + }, + { + "epoch": 2.357283464566929, + "grad_norm": 14.071619987487793, + "learning_rate": 2.363055434212112e-07, + "loss": 0.9276, + "step": 2397 + }, + { + "epoch": 2.358267716535433, + "grad_norm": 10.460234642028809, + "learning_rate": 2.3565691552916664e-07, + "loss": 0.4188, + "step": 2398 + }, + { + "epoch": 2.359251968503937, + "grad_norm": 13.995847702026367, + "learning_rate": 2.3500891726759822e-07, + "loss": 0.7293, + "step": 2399 + }, + { + "epoch": 2.3602362204724407, + "grad_norm": 7.614719867706299, + "learning_rate": 2.3436155017028224e-07, + "loss": 0.3724, + "step": 2400 + }, + { + "epoch": 2.361220472440945, + "grad_norm": 8.675552368164062, + "learning_rate": 2.3371481576950107e-07, + "loss": 0.3509, + "step": 2401 + }, + { + "epoch": 2.362204724409449, + "grad_norm": 11.069519996643066, + "learning_rate": 2.3306871559603962e-07, + "loss": 0.4966, + "step": 2402 + }, + { + "epoch": 2.3631889763779528, + "grad_norm": 11.743781089782715, + "learning_rate": 2.324232511791814e-07, + "loss": 0.5803, + "step": 2403 + }, + { + "epoch": 2.3641732283464565, + "grad_norm": 12.193204879760742, + "learning_rate": 2.3177842404670548e-07, + "loss": 0.6173, + "step": 2404 + }, + { + "epoch": 2.3651574803149606, + "grad_norm": 14.182714462280273, + "learning_rate": 2.3113423572488192e-07, + "loss": 0.3544, + "step": 2405 + }, + { + "epoch": 2.366141732283465, + "grad_norm": 9.212885856628418, + "learning_rate": 2.304906877384692e-07, + "loss": 0.3204, + "step": 2406 + }, + { + "epoch": 2.3671259842519685, + "grad_norm": 16.641233444213867, + "learning_rate": 2.298477816107103e-07, + "loss": 0.7335, + "step": 2407 + }, + { + "epoch": 2.3681102362204722, + "grad_norm": 6.9422383308410645, + "learning_rate": 2.292055188633285e-07, + "loss": 0.2462, + "step": 2408 + }, + { + "epoch": 2.3690944881889764, + "grad_norm": 7.835033893585205, + "learning_rate": 2.2856390101652452e-07, + "loss": 0.2921, + "step": 2409 + }, + { + "epoch": 2.3700787401574805, + "grad_norm": 11.217126846313477, + "learning_rate": 2.2792292958897266e-07, + "loss": 0.5388, + "step": 2410 + }, + { + "epoch": 2.3710629921259843, + "grad_norm": 10.122322082519531, + "learning_rate": 2.2728260609781702e-07, + "loss": 0.5551, + "step": 2411 + }, + { + "epoch": 2.372047244094488, + "grad_norm": 11.416410446166992, + "learning_rate": 2.2664293205866817e-07, + "loss": 0.4502, + "step": 2412 + }, + { + "epoch": 2.373031496062992, + "grad_norm": 18.35036849975586, + "learning_rate": 2.2600390898559944e-07, + "loss": 0.9301, + "step": 2413 + }, + { + "epoch": 2.3740157480314963, + "grad_norm": 10.62098503112793, + "learning_rate": 2.2536553839114338e-07, + "loss": 0.4878, + "step": 2414 + }, + { + "epoch": 2.375, + "grad_norm": 9.163315773010254, + "learning_rate": 2.2472782178628807e-07, + "loss": 0.3984, + "step": 2415 + }, + { + "epoch": 2.3759842519685037, + "grad_norm": 7.131898880004883, + "learning_rate": 2.2409076068047373e-07, + "loss": 0.2384, + "step": 2416 + }, + { + "epoch": 2.376968503937008, + "grad_norm": 8.367667198181152, + "learning_rate": 2.2345435658158908e-07, + "loss": 0.3365, + "step": 2417 + }, + { + "epoch": 2.377952755905512, + "grad_norm": 11.66543960571289, + "learning_rate": 2.2281861099596783e-07, + "loss": 0.5808, + "step": 2418 + }, + { + "epoch": 2.3789370078740157, + "grad_norm": 10.385133743286133, + "learning_rate": 2.2218352542838447e-07, + "loss": 0.2567, + "step": 2419 + }, + { + "epoch": 2.3799212598425195, + "grad_norm": 7.88507080078125, + "learning_rate": 2.215491013820519e-07, + "loss": 0.2747, + "step": 2420 + }, + { + "epoch": 2.3809055118110236, + "grad_norm": 6.82606315612793, + "learning_rate": 2.2091534035861703e-07, + "loss": 0.2653, + "step": 2421 + }, + { + "epoch": 2.3818897637795278, + "grad_norm": 13.949604034423828, + "learning_rate": 2.2028224385815726e-07, + "loss": 0.7345, + "step": 2422 + }, + { + "epoch": 2.3828740157480315, + "grad_norm": 8.991510391235352, + "learning_rate": 2.1964981337917744e-07, + "loss": 0.3903, + "step": 2423 + }, + { + "epoch": 2.383858267716535, + "grad_norm": 17.700889587402344, + "learning_rate": 2.1901805041860567e-07, + "loss": 0.9409, + "step": 2424 + }, + { + "epoch": 2.3848425196850394, + "grad_norm": 19.267620086669922, + "learning_rate": 2.1838695647179024e-07, + "loss": 0.7465, + "step": 2425 + }, + { + "epoch": 2.3858267716535435, + "grad_norm": 15.618047714233398, + "learning_rate": 2.1775653303249587e-07, + "loss": 0.7577, + "step": 2426 + }, + { + "epoch": 2.3868110236220472, + "grad_norm": 5.0363450050354, + "learning_rate": 2.1712678159290023e-07, + "loss": 0.1028, + "step": 2427 + }, + { + "epoch": 2.387795275590551, + "grad_norm": 9.236956596374512, + "learning_rate": 2.164977036435904e-07, + "loss": 0.4326, + "step": 2428 + }, + { + "epoch": 2.388779527559055, + "grad_norm": 3.6530535221099854, + "learning_rate": 2.1586930067355932e-07, + "loss": 0.0553, + "step": 2429 + }, + { + "epoch": 2.3897637795275593, + "grad_norm": 6.015213966369629, + "learning_rate": 2.152415741702023e-07, + "loss": 0.1769, + "step": 2430 + }, + { + "epoch": 2.390748031496063, + "grad_norm": 9.619933128356934, + "learning_rate": 2.146145256193136e-07, + "loss": 0.2792, + "step": 2431 + }, + { + "epoch": 2.391732283464567, + "grad_norm": 11.973121643066406, + "learning_rate": 2.1398815650508263e-07, + "loss": 0.5102, + "step": 2432 + }, + { + "epoch": 2.392716535433071, + "grad_norm": 9.242534637451172, + "learning_rate": 2.1336246831009075e-07, + "loss": 0.2668, + "step": 2433 + }, + { + "epoch": 2.393700787401575, + "grad_norm": 11.018208503723145, + "learning_rate": 2.1273746251530752e-07, + "loss": 0.5544, + "step": 2434 + }, + { + "epoch": 2.3946850393700787, + "grad_norm": 8.128729820251465, + "learning_rate": 2.1211314060008755e-07, + "loss": 0.2821, + "step": 2435 + }, + { + "epoch": 2.395669291338583, + "grad_norm": 11.07681941986084, + "learning_rate": 2.1148950404216617e-07, + "loss": 0.3712, + "step": 2436 + }, + { + "epoch": 2.3966535433070866, + "grad_norm": 7.771712303161621, + "learning_rate": 2.10866554317657e-07, + "loss": 0.2709, + "step": 2437 + }, + { + "epoch": 2.3976377952755907, + "grad_norm": 9.18880558013916, + "learning_rate": 2.1024429290104796e-07, + "loss": 0.3861, + "step": 2438 + }, + { + "epoch": 2.3986220472440944, + "grad_norm": 8.970224380493164, + "learning_rate": 2.0962272126519756e-07, + "loss": 0.2303, + "step": 2439 + }, + { + "epoch": 2.3996062992125986, + "grad_norm": 13.390796661376953, + "learning_rate": 2.0900184088133173e-07, + "loss": 0.7863, + "step": 2440 + }, + { + "epoch": 2.4005905511811023, + "grad_norm": 9.731471061706543, + "learning_rate": 2.0838165321904045e-07, + "loss": 0.2954, + "step": 2441 + }, + { + "epoch": 2.4015748031496065, + "grad_norm": 8.255941390991211, + "learning_rate": 2.0776215974627359e-07, + "loss": 0.3938, + "step": 2442 + }, + { + "epoch": 2.40255905511811, + "grad_norm": 10.414751052856445, + "learning_rate": 2.0714336192933824e-07, + "loss": 0.5809, + "step": 2443 + }, + { + "epoch": 2.4035433070866143, + "grad_norm": 9.516477584838867, + "learning_rate": 2.0652526123289488e-07, + "loss": 0.3192, + "step": 2444 + }, + { + "epoch": 2.404527559055118, + "grad_norm": 8.152141571044922, + "learning_rate": 2.0590785911995406e-07, + "loss": 0.3214, + "step": 2445 + }, + { + "epoch": 2.405511811023622, + "grad_norm": 12.904752731323242, + "learning_rate": 2.052911570518725e-07, + "loss": 0.6149, + "step": 2446 + }, + { + "epoch": 2.406496062992126, + "grad_norm": 11.767104148864746, + "learning_rate": 2.046751564883503e-07, + "loss": 0.6265, + "step": 2447 + }, + { + "epoch": 2.40748031496063, + "grad_norm": 15.724270820617676, + "learning_rate": 2.040598588874268e-07, + "loss": 0.5605, + "step": 2448 + }, + { + "epoch": 2.40748031496063, + "eval_Qnli-dev_cosine_accuracy": 0.685546875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.7995734214782715, + "eval_Qnli-dev_cosine_ap": 0.708134684015663, + "eval_Qnli-dev_cosine_f1": 0.6960784313725491, + "eval_Qnli-dev_cosine_f1_threshold": 0.7281224131584167, + "eval_Qnli-dev_cosine_precision": 0.5664893617021277, + "eval_Qnli-dev_cosine_recall": 0.902542372881356, + "eval_Qnli-dev_dot_accuracy": 0.681640625, + "eval_Qnli-dev_dot_accuracy_threshold": 609.6793212890625, + "eval_Qnli-dev_dot_ap": 0.7047383707764411, + "eval_Qnli-dev_dot_f1": 0.6929392446633826, + "eval_Qnli-dev_dot_f1_threshold": 555.753173828125, + "eval_Qnli-dev_dot_precision": 0.5656836461126006, + "eval_Qnli-dev_dot_recall": 0.8940677966101694, + "eval_Qnli-dev_euclidean_accuracy": 0.685546875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.47189712524414, + "eval_Qnli-dev_euclidean_ap": 0.7083737231857451, + "eval_Qnli-dev_euclidean_f1": 0.6950819672131149, + "eval_Qnli-dev_euclidean_f1_threshold": 20.271848678588867, + "eval_Qnli-dev_euclidean_precision": 0.5668449197860963, + "eval_Qnli-dev_euclidean_recall": 0.8983050847457628, + "eval_Qnli-dev_manhattan_accuracy": 0.693359375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 365.6907958984375, + "eval_Qnli-dev_manhattan_ap": 0.7170160615271656, + "eval_Qnli-dev_manhattan_f1": 0.7000000000000001, + "eval_Qnli-dev_manhattan_f1_threshold": 435.0411376953125, + "eval_Qnli-dev_manhattan_precision": 0.5651041666666666, + "eval_Qnli-dev_manhattan_recall": 0.9194915254237288, + "eval_Qnli-dev_max_accuracy": 0.693359375, + "eval_Qnli-dev_max_accuracy_threshold": 609.6793212890625, + "eval_Qnli-dev_max_ap": 0.7170160615271656, + "eval_Qnli-dev_max_f1": 0.7000000000000001, + "eval_Qnli-dev_max_f1_threshold": 555.753173828125, + "eval_Qnli-dev_max_precision": 0.5668449197860963, + "eval_Qnli-dev_max_recall": 0.9194915254237288, + "eval_allNLI-dev_cosine_accuracy": 0.697265625, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.918925404548645, + "eval_allNLI-dev_cosine_ap": 0.520334459841962, + "eval_allNLI-dev_cosine_f1": 0.557919621749409, + "eval_allNLI-dev_cosine_f1_threshold": 0.8274284601211548, + "eval_allNLI-dev_cosine_precision": 0.472, + "eval_allNLI-dev_cosine_recall": 0.6820809248554913, + "eval_allNLI-dev_dot_accuracy": 0.697265625, + "eval_allNLI-dev_dot_accuracy_threshold": 701.4763793945312, + "eval_allNLI-dev_dot_ap": 0.5215384128413414, + "eval_allNLI-dev_dot_f1": 0.557919621749409, + "eval_allNLI-dev_dot_f1_threshold": 631.42578125, + "eval_allNLI-dev_dot_precision": 0.472, + "eval_allNLI-dev_dot_recall": 0.6820809248554913, + "eval_allNLI-dev_euclidean_accuracy": 0.697265625, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.115690231323242, + "eval_allNLI-dev_euclidean_ap": 0.5206070999404473, + "eval_allNLI-dev_euclidean_f1": 0.557919621749409, + "eval_allNLI-dev_euclidean_f1_threshold": 16.241830825805664, + "eval_allNLI-dev_euclidean_precision": 0.472, + "eval_allNLI-dev_euclidean_recall": 0.6820809248554913, + "eval_allNLI-dev_manhattan_accuracy": 0.69140625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 240.7887725830078, + "eval_allNLI-dev_manhattan_ap": 0.5210401368161633, + "eval_allNLI-dev_manhattan_f1": 0.5545454545454546, + "eval_allNLI-dev_manhattan_f1_threshold": 348.866455078125, + "eval_allNLI-dev_manhattan_precision": 0.45692883895131087, + "eval_allNLI-dev_manhattan_recall": 0.7052023121387283, + "eval_allNLI-dev_max_accuracy": 0.697265625, + "eval_allNLI-dev_max_accuracy_threshold": 701.4763793945312, + "eval_allNLI-dev_max_ap": 0.5215384128413414, + "eval_allNLI-dev_max_f1": 0.557919621749409, + "eval_allNLI-dev_max_f1_threshold": 631.42578125, + "eval_allNLI-dev_max_precision": 0.472, + "eval_allNLI-dev_max_recall": 0.7052023121387283, + "eval_loss": 0.9221314787864685, + "eval_runtime": 54.3282, + "eval_samples_per_second": 30.629, + "eval_sequential_score": 0.7170160615271656, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.6584277178847996, + "eval_sts-test_pearson_dot": 0.657756195496527, + "eval_sts-test_pearson_euclidean": 0.6707158629188267, + "eval_sts-test_pearson_manhattan": 0.6745148890301079, + "eval_sts-test_pearson_max": 0.6745148890301079, + "eval_sts-test_spearman_cosine": 0.6639480172687122, + "eval_sts-test_spearman_dot": 0.6642357678794589, + "eval_sts-test_spearman_euclidean": 0.6641763382043312, + "eval_sts-test_spearman_manhattan": 0.6678039202039794, + "eval_sts-test_spearman_max": 0.6678039202039794, + "step": 2448 + }, + { + "epoch": 2.408464566929134, + "grad_norm": 15.768280982971191, + "learning_rate": 2.0344526570547774e-07, + "loss": 0.6515, + "step": 2449 + }, + { + "epoch": 2.409448818897638, + "grad_norm": 7.4104814529418945, + "learning_rate": 2.0283137839721143e-07, + "loss": 0.258, + "step": 2450 + }, + { + "epoch": 2.4104330708661417, + "grad_norm": 10.254469871520996, + "learning_rate": 2.022181984156654e-07, + "loss": 0.5888, + "step": 2451 + }, + { + "epoch": 2.411417322834646, + "grad_norm": 9.213778495788574, + "learning_rate": 2.0160572721220312e-07, + "loss": 0.3751, + "step": 2452 + }, + { + "epoch": 2.4124015748031495, + "grad_norm": 6.009047985076904, + "learning_rate": 2.0099396623651012e-07, + "loss": 0.1577, + "step": 2453 + }, + { + "epoch": 2.4133858267716537, + "grad_norm": 18.305646896362305, + "learning_rate": 2.0038291693659113e-07, + "loss": 0.8158, + "step": 2454 + }, + { + "epoch": 2.4143700787401574, + "grad_norm": 14.949423789978027, + "learning_rate": 1.997725807587663e-07, + "loss": 0.4467, + "step": 2455 + }, + { + "epoch": 2.4153543307086616, + "grad_norm": 8.558846473693848, + "learning_rate": 1.991629591476679e-07, + "loss": 0.3115, + "step": 2456 + }, + { + "epoch": 2.4163385826771653, + "grad_norm": 9.066378593444824, + "learning_rate": 1.9855405354623664e-07, + "loss": 0.3204, + "step": 2457 + }, + { + "epoch": 2.4173228346456694, + "grad_norm": 10.729354858398438, + "learning_rate": 1.9794586539571887e-07, + "loss": 0.4332, + "step": 2458 + }, + { + "epoch": 2.418307086614173, + "grad_norm": 13.04471206665039, + "learning_rate": 1.9733839613566244e-07, + "loss": 0.7696, + "step": 2459 + }, + { + "epoch": 2.4192913385826773, + "grad_norm": 7.856568813323975, + "learning_rate": 1.9673164720391387e-07, + "loss": 0.2334, + "step": 2460 + }, + { + "epoch": 2.420275590551181, + "grad_norm": 7.954514503479004, + "learning_rate": 1.9612562003661423e-07, + "loss": 0.346, + "step": 2461 + }, + { + "epoch": 2.421259842519685, + "grad_norm": 6.173386096954346, + "learning_rate": 1.9552031606819685e-07, + "loss": 0.1947, + "step": 2462 + }, + { + "epoch": 2.422244094488189, + "grad_norm": 9.832446098327637, + "learning_rate": 1.9491573673138287e-07, + "loss": 0.4588, + "step": 2463 + }, + { + "epoch": 2.423228346456693, + "grad_norm": 9.947352409362793, + "learning_rate": 1.9431188345717844e-07, + "loss": 0.3671, + "step": 2464 + }, + { + "epoch": 2.4242125984251968, + "grad_norm": 11.656517028808594, + "learning_rate": 1.9370875767487104e-07, + "loss": 0.3214, + "step": 2465 + }, + { + "epoch": 2.425196850393701, + "grad_norm": 14.011076927185059, + "learning_rate": 1.9310636081202636e-07, + "loss": 0.5728, + "step": 2466 + }, + { + "epoch": 2.4261811023622046, + "grad_norm": 12.770512580871582, + "learning_rate": 1.925046942944847e-07, + "loss": 0.7554, + "step": 2467 + }, + { + "epoch": 2.427165354330709, + "grad_norm": 9.351787567138672, + "learning_rate": 1.919037595463576e-07, + "loss": 0.3476, + "step": 2468 + }, + { + "epoch": 2.4281496062992125, + "grad_norm": 9.923174858093262, + "learning_rate": 1.9130355799002467e-07, + "loss": 0.4924, + "step": 2469 + }, + { + "epoch": 2.4291338582677167, + "grad_norm": 12.662055969238281, + "learning_rate": 1.9070409104613e-07, + "loss": 0.6821, + "step": 2470 + }, + { + "epoch": 2.4301181102362204, + "grad_norm": 12.125091552734375, + "learning_rate": 1.90105360133579e-07, + "loss": 0.5663, + "step": 2471 + }, + { + "epoch": 2.4311023622047245, + "grad_norm": 7.01002311706543, + "learning_rate": 1.89507366669535e-07, + "loss": 0.2414, + "step": 2472 + }, + { + "epoch": 2.4320866141732282, + "grad_norm": 7.311014175415039, + "learning_rate": 1.8891011206941535e-07, + "loss": 0.1729, + "step": 2473 + }, + { + "epoch": 2.4330708661417324, + "grad_norm": 13.780094146728516, + "learning_rate": 1.883135977468891e-07, + "loss": 0.8102, + "step": 2474 + }, + { + "epoch": 2.434055118110236, + "grad_norm": 11.154767036437988, + "learning_rate": 1.8771782511387285e-07, + "loss": 0.5511, + "step": 2475 + }, + { + "epoch": 2.4350393700787403, + "grad_norm": 13.085041999816895, + "learning_rate": 1.8712279558052777e-07, + "loss": 0.5231, + "step": 2476 + }, + { + "epoch": 2.436023622047244, + "grad_norm": 2.9605321884155273, + "learning_rate": 1.8652851055525605e-07, + "loss": 0.0735, + "step": 2477 + }, + { + "epoch": 2.437007874015748, + "grad_norm": 9.01956844329834, + "learning_rate": 1.859349714446977e-07, + "loss": 0.344, + "step": 2478 + }, + { + "epoch": 2.437992125984252, + "grad_norm": 10.247904777526855, + "learning_rate": 1.853421796537273e-07, + "loss": 0.5571, + "step": 2479 + }, + { + "epoch": 2.438976377952756, + "grad_norm": 9.704607009887695, + "learning_rate": 1.8475013658545054e-07, + "loss": 0.3844, + "step": 2480 + }, + { + "epoch": 2.4399606299212597, + "grad_norm": 7.756630897521973, + "learning_rate": 1.841588436412006e-07, + "loss": 0.2894, + "step": 2481 + }, + { + "epoch": 2.440944881889764, + "grad_norm": 10.263578414916992, + "learning_rate": 1.8356830222053546e-07, + "loss": 0.2938, + "step": 2482 + }, + { + "epoch": 2.4419291338582676, + "grad_norm": 7.927118301391602, + "learning_rate": 1.8297851372123444e-07, + "loss": 0.2579, + "step": 2483 + }, + { + "epoch": 2.4429133858267718, + "grad_norm": 12.571815490722656, + "learning_rate": 1.8238947953929436e-07, + "loss": 0.6039, + "step": 2484 + }, + { + "epoch": 2.4438976377952755, + "grad_norm": 10.60606575012207, + "learning_rate": 1.8180120106892685e-07, + "loss": 0.4512, + "step": 2485 + }, + { + "epoch": 2.4448818897637796, + "grad_norm": 9.76247787475586, + "learning_rate": 1.8121367970255475e-07, + "loss": 0.2862, + "step": 2486 + }, + { + "epoch": 2.4458661417322833, + "grad_norm": 11.170296669006348, + "learning_rate": 1.806269168308089e-07, + "loss": 0.3739, + "step": 2487 + }, + { + "epoch": 2.4468503937007875, + "grad_norm": 9.177827835083008, + "learning_rate": 1.800409138425248e-07, + "loss": 0.328, + "step": 2488 + }, + { + "epoch": 2.447834645669291, + "grad_norm": 11.069591522216797, + "learning_rate": 1.794556721247393e-07, + "loss": 0.344, + "step": 2489 + }, + { + "epoch": 2.4488188976377954, + "grad_norm": 13.294142723083496, + "learning_rate": 1.7887119306268743e-07, + "loss": 0.7, + "step": 2490 + }, + { + "epoch": 2.449803149606299, + "grad_norm": 11.407720565795898, + "learning_rate": 1.7828747803979914e-07, + "loss": 0.6593, + "step": 2491 + }, + { + "epoch": 2.4507874015748032, + "grad_norm": 10.573090553283691, + "learning_rate": 1.7770452843769567e-07, + "loss": 0.3137, + "step": 2492 + }, + { + "epoch": 2.451771653543307, + "grad_norm": 8.411659240722656, + "learning_rate": 1.771223456361868e-07, + "loss": 0.2765, + "step": 2493 + }, + { + "epoch": 2.452755905511811, + "grad_norm": 8.66166877746582, + "learning_rate": 1.765409310132673e-07, + "loss": 0.5268, + "step": 2494 + }, + { + "epoch": 2.453740157480315, + "grad_norm": 10.782097816467285, + "learning_rate": 1.7596028594511364e-07, + "loss": 0.5169, + "step": 2495 + }, + { + "epoch": 2.454724409448819, + "grad_norm": 11.39598274230957, + "learning_rate": 1.753804118060808e-07, + "loss": 0.3784, + "step": 2496 + }, + { + "epoch": 2.4557086614173227, + "grad_norm": 8.448685646057129, + "learning_rate": 1.7480130996869905e-07, + "loss": 0.1408, + "step": 2497 + }, + { + "epoch": 2.456692913385827, + "grad_norm": 6.417379379272461, + "learning_rate": 1.7422298180367063e-07, + "loss": 0.2154, + "step": 2498 + }, + { + "epoch": 2.4576771653543306, + "grad_norm": 10.22805118560791, + "learning_rate": 1.736454286798665e-07, + "loss": 0.4219, + "step": 2499 + }, + { + "epoch": 2.4586614173228347, + "grad_norm": 9.10873031616211, + "learning_rate": 1.730686519643233e-07, + "loss": 0.3657, + "step": 2500 + }, + { + "epoch": 2.4596456692913384, + "grad_norm": 8.159028053283691, + "learning_rate": 1.7249265302223998e-07, + "loss": 0.2133, + "step": 2501 + }, + { + "epoch": 2.4606299212598426, + "grad_norm": 13.752241134643555, + "learning_rate": 1.7191743321697403e-07, + "loss": 0.9923, + "step": 2502 + }, + { + "epoch": 2.4616141732283463, + "grad_norm": 5.2988786697387695, + "learning_rate": 1.713429939100393e-07, + "loss": 0.1073, + "step": 2503 + }, + { + "epoch": 2.4625984251968505, + "grad_norm": 9.656152725219727, + "learning_rate": 1.7076933646110222e-07, + "loss": 0.2467, + "step": 2504 + }, + { + "epoch": 2.463582677165354, + "grad_norm": 12.059698104858398, + "learning_rate": 1.7019646222797834e-07, + "loss": 0.609, + "step": 2505 + }, + { + "epoch": 2.4645669291338583, + "grad_norm": 9.394807815551758, + "learning_rate": 1.696243725666296e-07, + "loss": 0.2827, + "step": 2506 + }, + { + "epoch": 2.465551181102362, + "grad_norm": 7.839410781860352, + "learning_rate": 1.6905306883116077e-07, + "loss": 0.3833, + "step": 2507 + }, + { + "epoch": 2.466535433070866, + "grad_norm": 8.78386116027832, + "learning_rate": 1.684825523738165e-07, + "loss": 0.3798, + "step": 2508 + }, + { + "epoch": 2.46751968503937, + "grad_norm": 16.044845581054688, + "learning_rate": 1.67912824544978e-07, + "loss": 0.5413, + "step": 2509 + }, + { + "epoch": 2.468503937007874, + "grad_norm": 6.673086643218994, + "learning_rate": 1.6734388669315944e-07, + "loss": 0.2506, + "step": 2510 + }, + { + "epoch": 2.469488188976378, + "grad_norm": 5.23464298248291, + "learning_rate": 1.667757401650057e-07, + "loss": 0.102, + "step": 2511 + }, + { + "epoch": 2.470472440944882, + "grad_norm": 8.43914794921875, + "learning_rate": 1.6620838630528833e-07, + "loss": 0.223, + "step": 2512 + }, + { + "epoch": 2.4714566929133857, + "grad_norm": 14.900171279907227, + "learning_rate": 1.6564182645690283e-07, + "loss": 0.5648, + "step": 2513 + }, + { + "epoch": 2.47244094488189, + "grad_norm": 12.017776489257812, + "learning_rate": 1.650760619608653e-07, + "loss": 0.6887, + "step": 2514 + }, + { + "epoch": 2.4734251968503935, + "grad_norm": 12.223630905151367, + "learning_rate": 1.645110941563092e-07, + "loss": 0.5032, + "step": 2515 + }, + { + "epoch": 2.4744094488188977, + "grad_norm": 14.946356773376465, + "learning_rate": 1.639469243804822e-07, + "loss": 0.6455, + "step": 2516 + }, + { + "epoch": 2.4753937007874014, + "grad_norm": 6.877208232879639, + "learning_rate": 1.6338355396874332e-07, + "loss": 0.1955, + "step": 2517 + }, + { + "epoch": 2.4763779527559056, + "grad_norm": 14.754609107971191, + "learning_rate": 1.6282098425455936e-07, + "loss": 0.9092, + "step": 2518 + }, + { + "epoch": 2.4773622047244093, + "grad_norm": 7.898383140563965, + "learning_rate": 1.622592165695019e-07, + "loss": 0.2389, + "step": 2519 + }, + { + "epoch": 2.4783464566929134, + "grad_norm": 10.59056568145752, + "learning_rate": 1.616982522432443e-07, + "loss": 0.5426, + "step": 2520 + }, + { + "epoch": 2.479330708661417, + "grad_norm": 9.697310447692871, + "learning_rate": 1.6113809260355824e-07, + "loss": 0.6541, + "step": 2521 + }, + { + "epoch": 2.4803149606299213, + "grad_norm": 12.979537010192871, + "learning_rate": 1.6057873897631103e-07, + "loss": 0.5123, + "step": 2522 + }, + { + "epoch": 2.481299212598425, + "grad_norm": 11.808232307434082, + "learning_rate": 1.600201926854616e-07, + "loss": 0.5188, + "step": 2523 + }, + { + "epoch": 2.482283464566929, + "grad_norm": 9.810900688171387, + "learning_rate": 1.5946245505305863e-07, + "loss": 0.3183, + "step": 2524 + }, + { + "epoch": 2.483267716535433, + "grad_norm": 10.592032432556152, + "learning_rate": 1.5890552739923647e-07, + "loss": 0.2874, + "step": 2525 + }, + { + "epoch": 2.484251968503937, + "grad_norm": 11.300968170166016, + "learning_rate": 1.5834941104221217e-07, + "loss": 0.6203, + "step": 2526 + }, + { + "epoch": 2.4852362204724407, + "grad_norm": 9.015886306762695, + "learning_rate": 1.577941072982827e-07, + "loss": 0.3392, + "step": 2527 + }, + { + "epoch": 2.486220472440945, + "grad_norm": 5.7811689376831055, + "learning_rate": 1.5723961748182163e-07, + "loss": 0.1664, + "step": 2528 + }, + { + "epoch": 2.487204724409449, + "grad_norm": 11.204570770263672, + "learning_rate": 1.5668594290527568e-07, + "loss": 0.5371, + "step": 2529 + }, + { + "epoch": 2.4881889763779528, + "grad_norm": 16.38753890991211, + "learning_rate": 1.5613308487916242e-07, + "loss": 0.7918, + "step": 2530 + }, + { + "epoch": 2.4891732283464565, + "grad_norm": 6.875202178955078, + "learning_rate": 1.5558104471206626e-07, + "loss": 0.2345, + "step": 2531 + }, + { + "epoch": 2.4901574803149606, + "grad_norm": 8.88460636138916, + "learning_rate": 1.5502982371063602e-07, + "loss": 0.2692, + "step": 2532 + }, + { + "epoch": 2.491141732283465, + "grad_norm": 12.828963279724121, + "learning_rate": 1.5447942317958161e-07, + "loss": 0.4839, + "step": 2533 + }, + { + "epoch": 2.4921259842519685, + "grad_norm": 9.021425247192383, + "learning_rate": 1.539298444216708e-07, + "loss": 0.3797, + "step": 2534 + }, + { + "epoch": 2.4931102362204722, + "grad_norm": 7.733816623687744, + "learning_rate": 1.5338108873772638e-07, + "loss": 0.5387, + "step": 2535 + }, + { + "epoch": 2.4940944881889764, + "grad_norm": 12.049376487731934, + "learning_rate": 1.5283315742662288e-07, + "loss": 0.3812, + "step": 2536 + }, + { + "epoch": 2.4950787401574805, + "grad_norm": 7.660175323486328, + "learning_rate": 1.5228605178528368e-07, + "loss": 0.2541, + "step": 2537 + }, + { + "epoch": 2.4960629921259843, + "grad_norm": 6.382974624633789, + "learning_rate": 1.5173977310867775e-07, + "loss": 0.1971, + "step": 2538 + }, + { + "epoch": 2.497047244094488, + "grad_norm": 12.477048873901367, + "learning_rate": 1.5119432268981675e-07, + "loss": 0.5732, + "step": 2539 + }, + { + "epoch": 2.498031496062992, + "grad_norm": 15.173735618591309, + "learning_rate": 1.5064970181975204e-07, + "loss": 0.6801, + "step": 2540 + }, + { + "epoch": 2.4990157480314963, + "grad_norm": 13.453434944152832, + "learning_rate": 1.5010591178757092e-07, + "loss": 0.7655, + "step": 2541 + }, + { + "epoch": 2.5, + "grad_norm": 7.837796211242676, + "learning_rate": 1.4956295388039484e-07, + "loss": 0.1766, + "step": 2542 + }, + { + "epoch": 2.5009842519685037, + "grad_norm": 14.00476360321045, + "learning_rate": 1.4902082938337527e-07, + "loss": 0.3666, + "step": 2543 + }, + { + "epoch": 2.501968503937008, + "grad_norm": 9.707633972167969, + "learning_rate": 1.4847953957969112e-07, + "loss": 0.2594, + "step": 2544 + }, + { + "epoch": 2.502952755905512, + "grad_norm": 10.913497924804688, + "learning_rate": 1.479390857505457e-07, + "loss": 0.5756, + "step": 2545 + }, + { + "epoch": 2.5039370078740157, + "grad_norm": 5.065262794494629, + "learning_rate": 1.4739946917516362e-07, + "loss": 0.1573, + "step": 2546 + }, + { + "epoch": 2.5049212598425195, + "grad_norm": 12.768359184265137, + "learning_rate": 1.4686069113078744e-07, + "loss": 0.7746, + "step": 2547 + }, + { + "epoch": 2.5059055118110236, + "grad_norm": 25.458555221557617, + "learning_rate": 1.4632275289267535e-07, + "loss": 1.303, + "step": 2548 + }, + { + "epoch": 2.5068897637795278, + "grad_norm": 7.535705089569092, + "learning_rate": 1.457856557340976e-07, + "loss": 0.2377, + "step": 2549 + }, + { + "epoch": 2.5078740157480315, + "grad_norm": 6.729396820068359, + "learning_rate": 1.4524940092633376e-07, + "loss": 0.2366, + "step": 2550 + }, + { + "epoch": 2.508858267716535, + "grad_norm": 6.5007429122924805, + "learning_rate": 1.447139897386694e-07, + "loss": 0.1585, + "step": 2551 + }, + { + "epoch": 2.5098425196850394, + "grad_norm": 13.314377784729004, + "learning_rate": 1.4417942343839348e-07, + "loss": 0.8265, + "step": 2552 + }, + { + "epoch": 2.5108267716535435, + "grad_norm": 9.079033851623535, + "learning_rate": 1.4364570329079504e-07, + "loss": 0.3276, + "step": 2553 + }, + { + "epoch": 2.5118110236220472, + "grad_norm": 10.1290864944458, + "learning_rate": 1.4311283055916042e-07, + "loss": 0.437, + "step": 2554 + }, + { + "epoch": 2.512795275590551, + "grad_norm": 11.950562477111816, + "learning_rate": 1.4258080650477008e-07, + "loss": 0.5382, + "step": 2555 + }, + { + "epoch": 2.513779527559055, + "grad_norm": 14.388399124145508, + "learning_rate": 1.4204963238689582e-07, + "loss": 0.3736, + "step": 2556 + }, + { + "epoch": 2.5147637795275593, + "grad_norm": 6.537032127380371, + "learning_rate": 1.4151930946279746e-07, + "loss": 0.1608, + "step": 2557 + }, + { + "epoch": 2.515748031496063, + "grad_norm": 10.22529125213623, + "learning_rate": 1.4098983898772037e-07, + "loss": 0.4933, + "step": 2558 + }, + { + "epoch": 2.5167322834645667, + "grad_norm": 8.023038864135742, + "learning_rate": 1.404612222148921e-07, + "loss": 0.3567, + "step": 2559 + }, + { + "epoch": 2.517716535433071, + "grad_norm": 10.37587833404541, + "learning_rate": 1.3993346039551943e-07, + "loss": 0.5998, + "step": 2560 + }, + { + "epoch": 2.518700787401575, + "grad_norm": 5.052724361419678, + "learning_rate": 1.3940655477878572e-07, + "loss": 0.1126, + "step": 2561 + }, + { + "epoch": 2.5196850393700787, + "grad_norm": 9.678995132446289, + "learning_rate": 1.3888050661184755e-07, + "loss": 0.4245, + "step": 2562 + }, + { + "epoch": 2.5206692913385824, + "grad_norm": 8.200692176818848, + "learning_rate": 1.3835531713983213e-07, + "loss": 0.2628, + "step": 2563 + }, + { + "epoch": 2.5216535433070866, + "grad_norm": 8.967484474182129, + "learning_rate": 1.3783098760583417e-07, + "loss": 0.3958, + "step": 2564 + }, + { + "epoch": 2.5226377952755907, + "grad_norm": 8.796217918395996, + "learning_rate": 1.3730751925091263e-07, + "loss": 0.2905, + "step": 2565 + }, + { + "epoch": 2.5236220472440944, + "grad_norm": 6.851153373718262, + "learning_rate": 1.3678491331408854e-07, + "loss": 0.2747, + "step": 2566 + }, + { + "epoch": 2.524606299212598, + "grad_norm": 7.8298258781433105, + "learning_rate": 1.3626317103234146e-07, + "loss": 0.2156, + "step": 2567 + }, + { + "epoch": 2.5255905511811023, + "grad_norm": 9.772421836853027, + "learning_rate": 1.3574229364060663e-07, + "loss": 0.4541, + "step": 2568 + }, + { + "epoch": 2.5265748031496065, + "grad_norm": 8.752959251403809, + "learning_rate": 1.3522228237177238e-07, + "loss": 0.4792, + "step": 2569 + }, + { + "epoch": 2.52755905511811, + "grad_norm": 8.197463989257812, + "learning_rate": 1.347031384566767e-07, + "loss": 0.2525, + "step": 2570 + }, + { + "epoch": 2.528543307086614, + "grad_norm": 6.956662178039551, + "learning_rate": 1.3418486312410484e-07, + "loss": 0.1934, + "step": 2571 + }, + { + "epoch": 2.529527559055118, + "grad_norm": 5.082679271697998, + "learning_rate": 1.3366745760078604e-07, + "loss": 0.1325, + "step": 2572 + }, + { + "epoch": 2.530511811023622, + "grad_norm": 11.066061019897461, + "learning_rate": 1.3315092311139072e-07, + "loss": 0.3603, + "step": 2573 + }, + { + "epoch": 2.531496062992126, + "grad_norm": 7.229071617126465, + "learning_rate": 1.3263526087852766e-07, + "loss": 0.3196, + "step": 2574 + }, + { + "epoch": 2.53248031496063, + "grad_norm": 10.58609390258789, + "learning_rate": 1.321204721227411e-07, + "loss": 0.4006, + "step": 2575 + }, + { + "epoch": 2.533464566929134, + "grad_norm": 11.721096992492676, + "learning_rate": 1.316065580625077e-07, + "loss": 0.6041, + "step": 2576 + }, + { + "epoch": 2.534448818897638, + "grad_norm": 11.265833854675293, + "learning_rate": 1.3109351991423402e-07, + "loss": 0.6093, + "step": 2577 + }, + { + "epoch": 2.5354330708661417, + "grad_norm": 19.454679489135742, + "learning_rate": 1.3058135889225286e-07, + "loss": 1.0304, + "step": 2578 + }, + { + "epoch": 2.536417322834646, + "grad_norm": 12.229068756103516, + "learning_rate": 1.3007007620882145e-07, + "loss": 0.591, + "step": 2579 + }, + { + "epoch": 2.5374015748031495, + "grad_norm": 12.774563789367676, + "learning_rate": 1.2955967307411777e-07, + "loss": 0.5078, + "step": 2580 + }, + { + "epoch": 2.5383858267716537, + "grad_norm": 10.360858917236328, + "learning_rate": 1.290501506962381e-07, + "loss": 0.4428, + "step": 2581 + }, + { + "epoch": 2.5393700787401574, + "grad_norm": 14.671048164367676, + "learning_rate": 1.2854151028119375e-07, + "loss": 0.5756, + "step": 2582 + }, + { + "epoch": 2.5403543307086616, + "grad_norm": 8.14897632598877, + "learning_rate": 1.2803375303290884e-07, + "loss": 0.2725, + "step": 2583 + }, + { + "epoch": 2.5413385826771653, + "grad_norm": 14.281302452087402, + "learning_rate": 1.2752688015321684e-07, + "loss": 0.6336, + "step": 2584 + }, + { + "epoch": 2.5423228346456694, + "grad_norm": 9.67027759552002, + "learning_rate": 1.2702089284185817e-07, + "loss": 0.3588, + "step": 2585 + }, + { + "epoch": 2.543307086614173, + "grad_norm": 14.037250518798828, + "learning_rate": 1.265157922964767e-07, + "loss": 0.4981, + "step": 2586 + }, + { + "epoch": 2.5442913385826773, + "grad_norm": 6.424685955047607, + "learning_rate": 1.2601157971261788e-07, + "loss": 0.1576, + "step": 2587 + }, + { + "epoch": 2.545275590551181, + "grad_norm": 6.243099212646484, + "learning_rate": 1.255082562837252e-07, + "loss": 0.2362, + "step": 2588 + }, + { + "epoch": 2.546259842519685, + "grad_norm": 8.598806381225586, + "learning_rate": 1.250058232011375e-07, + "loss": 0.3347, + "step": 2589 + }, + { + "epoch": 2.547244094488189, + "grad_norm": 8.309163093566895, + "learning_rate": 1.2450428165408641e-07, + "loss": 0.3335, + "step": 2590 + }, + { + "epoch": 2.548228346456693, + "grad_norm": 9.962733268737793, + "learning_rate": 1.2400363282969322e-07, + "loss": 0.4493, + "step": 2591 + }, + { + "epoch": 2.5492125984251968, + "grad_norm": 8.566224098205566, + "learning_rate": 1.2350387791296617e-07, + "loss": 0.2456, + "step": 2592 + }, + { + "epoch": 2.550196850393701, + "grad_norm": 10.342599868774414, + "learning_rate": 1.2300501808679776e-07, + "loss": 0.3954, + "step": 2593 + }, + { + "epoch": 2.5511811023622046, + "grad_norm": 5.752167701721191, + "learning_rate": 1.2250705453196177e-07, + "loss": 0.1741, + "step": 2594 + }, + { + "epoch": 2.552165354330709, + "grad_norm": 8.814495086669922, + "learning_rate": 1.2200998842711062e-07, + "loss": 0.3343, + "step": 2595 + }, + { + "epoch": 2.5531496062992125, + "grad_norm": 7.947423934936523, + "learning_rate": 1.2151382094877245e-07, + "loss": 0.1977, + "step": 2596 + }, + { + "epoch": 2.5541338582677167, + "grad_norm": 16.38239288330078, + "learning_rate": 1.2101855327134849e-07, + "loss": 0.8805, + "step": 2597 + }, + { + "epoch": 2.5551181102362204, + "grad_norm": 12.04304313659668, + "learning_rate": 1.2052418656711007e-07, + "loss": 0.5718, + "step": 2598 + }, + { + "epoch": 2.5561023622047245, + "grad_norm": 7.0162787437438965, + "learning_rate": 1.200307220061961e-07, + "loss": 0.1922, + "step": 2599 + }, + { + "epoch": 2.5570866141732282, + "grad_norm": 8.008231163024902, + "learning_rate": 1.1953816075661008e-07, + "loss": 0.307, + "step": 2600 + }, + { + "epoch": 2.5580708661417324, + "grad_norm": 7.720792293548584, + "learning_rate": 1.1904650398421742e-07, + "loss": 0.1893, + "step": 2601 + }, + { + "epoch": 2.5580708661417324, + "eval_Qnli-dev_cosine_accuracy": 0.6953125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8067620396614075, + "eval_Qnli-dev_cosine_ap": 0.7043241142333805, + "eval_Qnli-dev_cosine_f1": 0.6953748006379585, + "eval_Qnli-dev_cosine_f1_threshold": 0.7095954418182373, + "eval_Qnli-dev_cosine_precision": 0.5575447570332481, + "eval_Qnli-dev_cosine_recall": 0.923728813559322, + "eval_Qnli-dev_dot_accuracy": 0.6875, + "eval_Qnli-dev_dot_accuracy_threshold": 612.6456298828125, + "eval_Qnli-dev_dot_ap": 0.7015226184100491, + "eval_Qnli-dev_dot_f1": 0.6943999999999998, + "eval_Qnli-dev_dot_f1_threshold": 541.830322265625, + "eval_Qnli-dev_dot_precision": 0.5578406169665809, + "eval_Qnli-dev_dot_recall": 0.9194915254237288, + "eval_Qnli-dev_euclidean_accuracy": 0.701171875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 17.201496124267578, + "eval_Qnli-dev_euclidean_ap": 0.7050333505348279, + "eval_Qnli-dev_euclidean_f1": 0.6953748006379585, + "eval_Qnli-dev_euclidean_f1_threshold": 20.963668823242188, + "eval_Qnli-dev_euclidean_precision": 0.5575447570332481, + "eval_Qnli-dev_euclidean_recall": 0.923728813559322, + "eval_Qnli-dev_manhattan_accuracy": 0.6953125, + "eval_Qnli-dev_manhattan_accuracy_threshold": 364.14117431640625, + "eval_Qnli-dev_manhattan_ap": 0.7137495271732022, + "eval_Qnli-dev_manhattan_f1": 0.694136291600634, + "eval_Qnli-dev_manhattan_f1_threshold": 444.42218017578125, + "eval_Qnli-dev_manhattan_precision": 0.5544303797468354, + "eval_Qnli-dev_manhattan_recall": 0.9279661016949152, + "eval_Qnli-dev_max_accuracy": 0.701171875, + "eval_Qnli-dev_max_accuracy_threshold": 612.6456298828125, + "eval_Qnli-dev_max_ap": 0.7137495271732022, + "eval_Qnli-dev_max_f1": 0.6953748006379585, + "eval_Qnli-dev_max_f1_threshold": 541.830322265625, + "eval_Qnli-dev_max_precision": 0.5578406169665809, + "eval_Qnli-dev_max_recall": 0.9279661016949152, + "eval_allNLI-dev_cosine_accuracy": 0.697265625, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.912941038608551, + "eval_allNLI-dev_cosine_ap": 0.5170912462717843, + "eval_allNLI-dev_cosine_f1": 0.558891454965358, + "eval_allNLI-dev_cosine_f1_threshold": 0.826492190361023, + "eval_allNLI-dev_cosine_precision": 0.4653846153846154, + "eval_allNLI-dev_cosine_recall": 0.6994219653179191, + "eval_allNLI-dev_dot_accuracy": 0.6953125, + "eval_allNLI-dev_dot_accuracy_threshold": 699.905517578125, + "eval_allNLI-dev_dot_ap": 0.5173897942840021, + "eval_allNLI-dev_dot_f1": 0.5568445475638051, + "eval_allNLI-dev_dot_f1_threshold": 630.53271484375, + "eval_allNLI-dev_dot_precision": 0.46511627906976744, + "eval_allNLI-dev_dot_recall": 0.6936416184971098, + "eval_allNLI-dev_euclidean_accuracy": 0.697265625, + "eval_allNLI-dev_euclidean_accuracy_threshold": 11.538026809692383, + "eval_allNLI-dev_euclidean_ap": 0.5171319156824372, + "eval_allNLI-dev_euclidean_f1": 0.5576036866359447, + "eval_allNLI-dev_euclidean_f1_threshold": 16.292882919311523, + "eval_allNLI-dev_euclidean_precision": 0.46360153256704983, + "eval_allNLI-dev_euclidean_recall": 0.6994219653179191, + "eval_allNLI-dev_manhattan_accuracy": 0.689453125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 248.03265380859375, + "eval_allNLI-dev_manhattan_ap": 0.5172164569460806, + "eval_allNLI-dev_manhattan_f1": 0.5596330275229358, + "eval_allNLI-dev_manhattan_f1_threshold": 344.65313720703125, + "eval_allNLI-dev_manhattan_precision": 0.46387832699619774, + "eval_allNLI-dev_manhattan_recall": 0.7052023121387283, + "eval_allNLI-dev_max_accuracy": 0.697265625, + "eval_allNLI-dev_max_accuracy_threshold": 699.905517578125, + "eval_allNLI-dev_max_ap": 0.5173897942840021, + "eval_allNLI-dev_max_f1": 0.5596330275229358, + "eval_allNLI-dev_max_f1_threshold": 630.53271484375, + "eval_allNLI-dev_max_precision": 0.4653846153846154, + "eval_allNLI-dev_max_recall": 0.7052023121387283, + "eval_loss": 0.901435136795044, + "eval_runtime": 54.1536, + "eval_samples_per_second": 30.727, + "eval_sequential_score": 0.7137495271732022, + "eval_steps_per_second": 0.129, + "eval_sts-test_pearson_cosine": 0.6591049782717581, + "eval_sts-test_pearson_dot": 0.6585371125233661, + "eval_sts-test_pearson_euclidean": 0.6710226989754035, + "eval_sts-test_pearson_manhattan": 0.674239366272805, + "eval_sts-test_pearson_max": 0.674239366272805, + "eval_sts-test_spearman_cosine": 0.6655085832559531, + "eval_sts-test_spearman_dot": 0.6647672338597431, + "eval_sts-test_spearman_euclidean": 0.6657110379624879, + "eval_sts-test_spearman_manhattan": 0.6690548074623918, + "eval_sts-test_spearman_max": 0.6690548074623918, + "step": 2601 + }, + { + "epoch": 2.559055118110236, + "grad_norm": 5.273887634277344, + "learning_rate": 1.185557528527428e-07, + "loss": 0.1929, + "step": 2602 + }, + { + "epoch": 2.5600393700787403, + "grad_norm": 10.100863456726074, + "learning_rate": 1.1806590852376707e-07, + "loss": 0.5182, + "step": 2603 + }, + { + "epoch": 2.561023622047244, + "grad_norm": 4.793201923370361, + "learning_rate": 1.1757697215672498e-07, + "loss": 0.1183, + "step": 2604 + }, + { + "epoch": 2.562007874015748, + "grad_norm": 15.92398452758789, + "learning_rate": 1.17088944908902e-07, + "loss": 0.9434, + "step": 2605 + }, + { + "epoch": 2.562992125984252, + "grad_norm": 9.905328750610352, + "learning_rate": 1.1660182793543199e-07, + "loss": 0.4401, + "step": 2606 + }, + { + "epoch": 2.563976377952756, + "grad_norm": 8.067452430725098, + "learning_rate": 1.161156223892938e-07, + "loss": 0.4079, + "step": 2607 + }, + { + "epoch": 2.5649606299212597, + "grad_norm": 4.889436721801758, + "learning_rate": 1.1563032942130937e-07, + "loss": 0.2161, + "step": 2608 + }, + { + "epoch": 2.565944881889764, + "grad_norm": 4.1334686279296875, + "learning_rate": 1.1514595018014061e-07, + "loss": 0.0959, + "step": 2609 + }, + { + "epoch": 2.5669291338582676, + "grad_norm": 9.387733459472656, + "learning_rate": 1.1466248581228649e-07, + "loss": 0.2877, + "step": 2610 + }, + { + "epoch": 2.5679133858267718, + "grad_norm": 5.032553672790527, + "learning_rate": 1.1417993746208068e-07, + "loss": 0.1447, + "step": 2611 + }, + { + "epoch": 2.5688976377952755, + "grad_norm": 12.307943344116211, + "learning_rate": 1.136983062716886e-07, + "loss": 0.5285, + "step": 2612 + }, + { + "epoch": 2.5698818897637796, + "grad_norm": 15.693968772888184, + "learning_rate": 1.1321759338110483e-07, + "loss": 0.4332, + "step": 2613 + }, + { + "epoch": 2.5708661417322833, + "grad_norm": 7.844369411468506, + "learning_rate": 1.1273779992815057e-07, + "loss": 0.1841, + "step": 2614 + }, + { + "epoch": 2.5718503937007875, + "grad_norm": 10.694293022155762, + "learning_rate": 1.1225892704847027e-07, + "loss": 0.4788, + "step": 2615 + }, + { + "epoch": 2.572834645669291, + "grad_norm": 11.026900291442871, + "learning_rate": 1.1178097587552988e-07, + "loss": 0.4086, + "step": 2616 + }, + { + "epoch": 2.5738188976377954, + "grad_norm": 5.624032020568848, + "learning_rate": 1.113039475406136e-07, + "loss": 0.1467, + "step": 2617 + }, + { + "epoch": 2.574803149606299, + "grad_norm": 8.110433578491211, + "learning_rate": 1.1082784317282129e-07, + "loss": 0.389, + "step": 2618 + }, + { + "epoch": 2.5757874015748032, + "grad_norm": 8.149201393127441, + "learning_rate": 1.1035266389906588e-07, + "loss": 0.2413, + "step": 2619 + }, + { + "epoch": 2.576771653543307, + "grad_norm": 10.457308769226074, + "learning_rate": 1.0987841084407059e-07, + "loss": 0.3814, + "step": 2620 + }, + { + "epoch": 2.577755905511811, + "grad_norm": 19.59735107421875, + "learning_rate": 1.0940508513036641e-07, + "loss": 1.1875, + "step": 2621 + }, + { + "epoch": 2.578740157480315, + "grad_norm": 11.51667594909668, + "learning_rate": 1.0893268787828929e-07, + "loss": 0.5347, + "step": 2622 + }, + { + "epoch": 2.579724409448819, + "grad_norm": 7.69096040725708, + "learning_rate": 1.0846122020597759e-07, + "loss": 0.3168, + "step": 2623 + }, + { + "epoch": 2.5807086614173227, + "grad_norm": 9.124616622924805, + "learning_rate": 1.0799068322936935e-07, + "loss": 0.3111, + "step": 2624 + }, + { + "epoch": 2.581692913385827, + "grad_norm": 7.774921894073486, + "learning_rate": 1.0752107806219986e-07, + "loss": 0.2107, + "step": 2625 + }, + { + "epoch": 2.5826771653543306, + "grad_norm": 11.955632209777832, + "learning_rate": 1.0705240581599881e-07, + "loss": 0.4728, + "step": 2626 + }, + { + "epoch": 2.5836614173228347, + "grad_norm": 8.494672775268555, + "learning_rate": 1.0658466760008749e-07, + "loss": 0.4143, + "step": 2627 + }, + { + "epoch": 2.5846456692913384, + "grad_norm": 15.70314884185791, + "learning_rate": 1.0611786452157668e-07, + "loss": 0.8418, + "step": 2628 + }, + { + "epoch": 2.5856299212598426, + "grad_norm": 4.958100318908691, + "learning_rate": 1.0565199768536355e-07, + "loss": 0.1177, + "step": 2629 + }, + { + "epoch": 2.5866141732283463, + "grad_norm": 11.516602516174316, + "learning_rate": 1.0518706819412948e-07, + "loss": 0.5185, + "step": 2630 + }, + { + "epoch": 2.5875984251968505, + "grad_norm": 16.29201889038086, + "learning_rate": 1.0472307714833692e-07, + "loss": 0.5703, + "step": 2631 + }, + { + "epoch": 2.588582677165354, + "grad_norm": 6.621307373046875, + "learning_rate": 1.042600256462273e-07, + "loss": 0.1895, + "step": 2632 + }, + { + "epoch": 2.5895669291338583, + "grad_norm": 8.731075286865234, + "learning_rate": 1.0379791478381805e-07, + "loss": 0.3434, + "step": 2633 + }, + { + "epoch": 2.590551181102362, + "grad_norm": 13.496260643005371, + "learning_rate": 1.0333674565490014e-07, + "loss": 0.6885, + "step": 2634 + }, + { + "epoch": 2.591535433070866, + "grad_norm": 12.25928020477295, + "learning_rate": 1.0287651935103572e-07, + "loss": 0.5406, + "step": 2635 + }, + { + "epoch": 2.59251968503937, + "grad_norm": 7.45269250869751, + "learning_rate": 1.024172369615551e-07, + "loss": 0.3012, + "step": 2636 + }, + { + "epoch": 2.593503937007874, + "grad_norm": 17.18060874938965, + "learning_rate": 1.0195889957355454e-07, + "loss": 0.9385, + "step": 2637 + }, + { + "epoch": 2.594488188976378, + "grad_norm": 9.831514358520508, + "learning_rate": 1.0150150827189337e-07, + "loss": 0.3235, + "step": 2638 + }, + { + "epoch": 2.595472440944882, + "grad_norm": 7.493289947509766, + "learning_rate": 1.0104506413919173e-07, + "loss": 0.2036, + "step": 2639 + }, + { + "epoch": 2.596456692913386, + "grad_norm": 9.814960479736328, + "learning_rate": 1.0058956825582792e-07, + "loss": 0.3403, + "step": 2640 + }, + { + "epoch": 2.59744094488189, + "grad_norm": 8.145853042602539, + "learning_rate": 1.0013502169993552e-07, + "loss": 0.3688, + "step": 2641 + }, + { + "epoch": 2.5984251968503935, + "grad_norm": 5.9386773109436035, + "learning_rate": 9.96814255474014e-08, + "loss": 0.3998, + "step": 2642 + }, + { + "epoch": 2.5994094488188977, + "grad_norm": 10.198760986328125, + "learning_rate": 9.922878087186273e-08, + "loss": 0.3149, + "step": 2643 + }, + { + "epoch": 2.600393700787402, + "grad_norm": 8.756124496459961, + "learning_rate": 9.877708874470455e-08, + "loss": 0.2704, + "step": 2644 + }, + { + "epoch": 2.6013779527559056, + "grad_norm": 6.515359401702881, + "learning_rate": 9.832635023505755e-08, + "loss": 0.158, + "step": 2645 + }, + { + "epoch": 2.6023622047244093, + "grad_norm": 8.050736427307129, + "learning_rate": 9.787656640979466e-08, + "loss": 0.2982, + "step": 2646 + }, + { + "epoch": 2.6033464566929134, + "grad_norm": 7.078582286834717, + "learning_rate": 9.74277383335297e-08, + "loss": 0.2121, + "step": 2647 + }, + { + "epoch": 2.6043307086614176, + "grad_norm": 12.721890449523926, + "learning_rate": 9.697986706861403e-08, + "loss": 0.426, + "step": 2648 + }, + { + "epoch": 2.6053149606299213, + "grad_norm": 5.437317371368408, + "learning_rate": 9.653295367513442e-08, + "loss": 0.1346, + "step": 2649 + }, + { + "epoch": 2.606299212598425, + "grad_norm": 10.403167724609375, + "learning_rate": 9.608699921091029e-08, + "loss": 0.5152, + "step": 2650 + }, + { + "epoch": 2.607283464566929, + "grad_norm": 15.584598541259766, + "learning_rate": 9.564200473149156e-08, + "loss": 0.705, + "step": 2651 + }, + { + "epoch": 2.6082677165354333, + "grad_norm": 10.282527923583984, + "learning_rate": 9.519797129015536e-08, + "loss": 0.3815, + "step": 2652 + }, + { + "epoch": 2.609251968503937, + "grad_norm": 8.903995513916016, + "learning_rate": 9.475489993790467e-08, + "loss": 0.3878, + "step": 2653 + }, + { + "epoch": 2.6102362204724407, + "grad_norm": 11.657297134399414, + "learning_rate": 9.431279172346506e-08, + "loss": 0.2256, + "step": 2654 + }, + { + "epoch": 2.611220472440945, + "grad_norm": 13.781116485595703, + "learning_rate": 9.387164769328237e-08, + "loss": 0.6192, + "step": 2655 + }, + { + "epoch": 2.612204724409449, + "grad_norm": 16.146852493286133, + "learning_rate": 9.343146889152028e-08, + "loss": 0.6693, + "step": 2656 + }, + { + "epoch": 2.6131889763779528, + "grad_norm": 10.342763900756836, + "learning_rate": 9.299225636005791e-08, + "loss": 0.2787, + "step": 2657 + }, + { + "epoch": 2.6141732283464565, + "grad_norm": 6.541711807250977, + "learning_rate": 9.25540111384872e-08, + "loss": 0.2554, + "step": 2658 + }, + { + "epoch": 2.6151574803149606, + "grad_norm": 6.099989414215088, + "learning_rate": 9.211673426411052e-08, + "loss": 0.1382, + "step": 2659 + }, + { + "epoch": 2.616141732283465, + "grad_norm": 9.739631652832031, + "learning_rate": 9.168042677193818e-08, + "loss": 0.4049, + "step": 2660 + }, + { + "epoch": 2.6171259842519685, + "grad_norm": 8.16472339630127, + "learning_rate": 9.124508969468606e-08, + "loss": 0.4021, + "step": 2661 + }, + { + "epoch": 2.6181102362204722, + "grad_norm": 7.355652809143066, + "learning_rate": 9.08107240627732e-08, + "loss": 0.2101, + "step": 2662 + }, + { + "epoch": 2.6190944881889764, + "grad_norm": 14.677267074584961, + "learning_rate": 9.037733090431912e-08, + "loss": 0.4881, + "step": 2663 + }, + { + "epoch": 2.6200787401574805, + "grad_norm": 8.316322326660156, + "learning_rate": 8.994491124514166e-08, + "loss": 0.2375, + "step": 2664 + }, + { + "epoch": 2.6210629921259843, + "grad_norm": 18.017789840698242, + "learning_rate": 8.951346610875441e-08, + "loss": 0.7727, + "step": 2665 + }, + { + "epoch": 2.622047244094488, + "grad_norm": 10.476648330688477, + "learning_rate": 8.908299651636436e-08, + "loss": 0.489, + "step": 2666 + }, + { + "epoch": 2.623031496062992, + "grad_norm": 14.659133911132812, + "learning_rate": 8.865350348686934e-08, + "loss": 0.7648, + "step": 2667 + }, + { + "epoch": 2.6240157480314963, + "grad_norm": 11.297569274902344, + "learning_rate": 8.822498803685595e-08, + "loss": 0.3395, + "step": 2668 + }, + { + "epoch": 2.625, + "grad_norm": 2.457162380218506, + "learning_rate": 8.779745118059643e-08, + "loss": 0.0534, + "step": 2669 + }, + { + "epoch": 2.6259842519685037, + "grad_norm": 7.145143032073975, + "learning_rate": 8.737089393004722e-08, + "loss": 0.3496, + "step": 2670 + }, + { + "epoch": 2.626968503937008, + "grad_norm": 12.202618598937988, + "learning_rate": 8.694531729484584e-08, + "loss": 0.7193, + "step": 2671 + }, + { + "epoch": 2.627952755905512, + "grad_norm": 8.030990600585938, + "learning_rate": 8.65207222823089e-08, + "loss": 0.2831, + "step": 2672 + }, + { + "epoch": 2.6289370078740157, + "grad_norm": 8.702947616577148, + "learning_rate": 8.609710989742948e-08, + "loss": 0.4369, + "step": 2673 + }, + { + "epoch": 2.6299212598425195, + "grad_norm": 11.548181533813477, + "learning_rate": 8.567448114287475e-08, + "loss": 0.4547, + "step": 2674 + }, + { + "epoch": 2.6309055118110236, + "grad_norm": 8.138830184936523, + "learning_rate": 8.525283701898387e-08, + "loss": 0.1804, + "step": 2675 + }, + { + "epoch": 2.6318897637795278, + "grad_norm": 8.969474792480469, + "learning_rate": 8.483217852376527e-08, + "loss": 0.3648, + "step": 2676 + }, + { + "epoch": 2.6328740157480315, + "grad_norm": 7.652669429779053, + "learning_rate": 8.441250665289457e-08, + "loss": 0.255, + "step": 2677 + }, + { + "epoch": 2.633858267716535, + "grad_norm": 6.664198875427246, + "learning_rate": 8.399382239971196e-08, + "loss": 0.2309, + "step": 2678 + }, + { + "epoch": 2.6348425196850394, + "grad_norm": 11.025222778320312, + "learning_rate": 8.357612675522011e-08, + "loss": 0.7033, + "step": 2679 + }, + { + "epoch": 2.6358267716535435, + "grad_norm": 17.735673904418945, + "learning_rate": 8.315942070808176e-08, + "loss": 0.8678, + "step": 2680 + }, + { + "epoch": 2.6368110236220472, + "grad_norm": 12.674147605895996, + "learning_rate": 8.274370524461712e-08, + "loss": 0.5771, + "step": 2681 + }, + { + "epoch": 2.637795275590551, + "grad_norm": 8.278904914855957, + "learning_rate": 8.232898134880213e-08, + "loss": 0.3597, + "step": 2682 + }, + { + "epoch": 2.638779527559055, + "grad_norm": 3.23533034324646, + "learning_rate": 8.19152500022651e-08, + "loss": 0.0554, + "step": 2683 + }, + { + "epoch": 2.6397637795275593, + "grad_norm": 8.985905647277832, + "learning_rate": 8.150251218428557e-08, + "loss": 0.2783, + "step": 2684 + }, + { + "epoch": 2.640748031496063, + "grad_norm": 6.580881595611572, + "learning_rate": 8.109076887179138e-08, + "loss": 0.2926, + "step": 2685 + }, + { + "epoch": 2.6417322834645667, + "grad_norm": 6.928036689758301, + "learning_rate": 8.068002103935624e-08, + "loss": 0.2097, + "step": 2686 + }, + { + "epoch": 2.642716535433071, + "grad_norm": 17.22882843017578, + "learning_rate": 8.027026965919774e-08, + "loss": 1.0672, + "step": 2687 + }, + { + "epoch": 2.643700787401575, + "grad_norm": 6.01613712310791, + "learning_rate": 7.986151570117498e-08, + "loss": 0.1549, + "step": 2688 + }, + { + "epoch": 2.6446850393700787, + "grad_norm": 2.506317377090454, + "learning_rate": 7.945376013278619e-08, + "loss": 0.0484, + "step": 2689 + }, + { + "epoch": 2.6456692913385824, + "grad_norm": 3.891319513320923, + "learning_rate": 7.904700391916614e-08, + "loss": 0.0606, + "step": 2690 + }, + { + "epoch": 2.6466535433070866, + "grad_norm": 7.721035957336426, + "learning_rate": 7.864124802308461e-08, + "loss": 0.2239, + "step": 2691 + }, + { + "epoch": 2.6476377952755907, + "grad_norm": 10.72614574432373, + "learning_rate": 7.823649340494358e-08, + "loss": 0.3587, + "step": 2692 + }, + { + "epoch": 2.6486220472440944, + "grad_norm": 7.484883785247803, + "learning_rate": 7.783274102277492e-08, + "loss": 0.1615, + "step": 2693 + }, + { + "epoch": 2.649606299212598, + "grad_norm": 13.634208679199219, + "learning_rate": 7.742999183223844e-08, + "loss": 0.9166, + "step": 2694 + }, + { + "epoch": 2.6505905511811023, + "grad_norm": 13.837207794189453, + "learning_rate": 7.702824678661923e-08, + "loss": 0.9832, + "step": 2695 + }, + { + "epoch": 2.6515748031496065, + "grad_norm": 13.681209564208984, + "learning_rate": 7.662750683682593e-08, + "loss": 0.788, + "step": 2696 + }, + { + "epoch": 2.65255905511811, + "grad_norm": 10.74668025970459, + "learning_rate": 7.622777293138787e-08, + "loss": 0.4497, + "step": 2697 + }, + { + "epoch": 2.653543307086614, + "grad_norm": 8.555436134338379, + "learning_rate": 7.582904601645333e-08, + "loss": 0.2578, + "step": 2698 + }, + { + "epoch": 2.654527559055118, + "grad_norm": 7.200206279754639, + "learning_rate": 7.543132703578697e-08, + "loss": 0.2575, + "step": 2699 + }, + { + "epoch": 2.655511811023622, + "grad_norm": 7.164278984069824, + "learning_rate": 7.503461693076783e-08, + "loss": 0.2621, + "step": 2700 + }, + { + "epoch": 2.656496062992126, + "grad_norm": 7.4781036376953125, + "learning_rate": 7.463891664038691e-08, + "loss": 0.2647, + "step": 2701 + }, + { + "epoch": 2.65748031496063, + "grad_norm": 9.039846420288086, + "learning_rate": 7.424422710124503e-08, + "loss": 0.3083, + "step": 2702 + }, + { + "epoch": 2.658464566929134, + "grad_norm": 11.146456718444824, + "learning_rate": 7.385054924755081e-08, + "loss": 0.326, + "step": 2703 + }, + { + "epoch": 2.659448818897638, + "grad_norm": 7.017238140106201, + "learning_rate": 7.345788401111793e-08, + "loss": 0.3751, + "step": 2704 + }, + { + "epoch": 2.6604330708661417, + "grad_norm": 15.564183235168457, + "learning_rate": 7.306623232136358e-08, + "loss": 1.0587, + "step": 2705 + }, + { + "epoch": 2.661417322834646, + "grad_norm": 5.695704460144043, + "learning_rate": 7.26755951053058e-08, + "loss": 0.1722, + "step": 2706 + }, + { + "epoch": 2.6624015748031495, + "grad_norm": 8.566353797912598, + "learning_rate": 7.228597328756136e-08, + "loss": 0.2922, + "step": 2707 + }, + { + "epoch": 2.6633858267716537, + "grad_norm": 11.269908905029297, + "learning_rate": 7.189736779034375e-08, + "loss": 0.409, + "step": 2708 + }, + { + "epoch": 2.6643700787401574, + "grad_norm": 11.922059059143066, + "learning_rate": 7.150977953346082e-08, + "loss": 0.2721, + "step": 2709 + }, + { + "epoch": 2.6653543307086616, + "grad_norm": 9.26468563079834, + "learning_rate": 7.112320943431287e-08, + "loss": 0.2416, + "step": 2710 + }, + { + "epoch": 2.6663385826771653, + "grad_norm": 9.859065055847168, + "learning_rate": 7.07376584078898e-08, + "loss": 0.3148, + "step": 2711 + }, + { + "epoch": 2.6673228346456694, + "grad_norm": 10.863545417785645, + "learning_rate": 7.035312736676993e-08, + "loss": 0.3821, + "step": 2712 + }, + { + "epoch": 2.668307086614173, + "grad_norm": 13.722128868103027, + "learning_rate": 6.996961722111701e-08, + "loss": 0.6524, + "step": 2713 + }, + { + "epoch": 2.6692913385826773, + "grad_norm": 12.660308837890625, + "learning_rate": 6.958712887867856e-08, + "loss": 0.5353, + "step": 2714 + }, + { + "epoch": 2.670275590551181, + "grad_norm": 15.21914291381836, + "learning_rate": 6.920566324478348e-08, + "loss": 0.4884, + "step": 2715 + }, + { + "epoch": 2.671259842519685, + "grad_norm": 6.661035060882568, + "learning_rate": 6.882522122234002e-08, + "loss": 0.1514, + "step": 2716 + }, + { + "epoch": 2.672244094488189, + "grad_norm": 8.941335678100586, + "learning_rate": 6.844580371183347e-08, + "loss": 0.3488, + "step": 2717 + }, + { + "epoch": 2.673228346456693, + "grad_norm": 8.471234321594238, + "learning_rate": 6.806741161132428e-08, + "loss": 0.2628, + "step": 2718 + }, + { + "epoch": 2.6742125984251968, + "grad_norm": 13.892240524291992, + "learning_rate": 6.769004581644596e-08, + "loss": 0.8696, + "step": 2719 + }, + { + "epoch": 2.675196850393701, + "grad_norm": 7.2624921798706055, + "learning_rate": 6.73137072204023e-08, + "loss": 0.199, + "step": 2720 + }, + { + "epoch": 2.6761811023622046, + "grad_norm": 15.945473670959473, + "learning_rate": 6.693839671396629e-08, + "loss": 0.8202, + "step": 2721 + }, + { + "epoch": 2.677165354330709, + "grad_norm": 12.784812927246094, + "learning_rate": 6.656411518547725e-08, + "loss": 0.905, + "step": 2722 + }, + { + "epoch": 2.6781496062992125, + "grad_norm": 10.727723121643066, + "learning_rate": 6.619086352083902e-08, + "loss": 0.4115, + "step": 2723 + }, + { + "epoch": 2.6791338582677167, + "grad_norm": 8.653194427490234, + "learning_rate": 6.581864260351775e-08, + "loss": 0.2155, + "step": 2724 + }, + { + "epoch": 2.6801181102362204, + "grad_norm": 10.35145378112793, + "learning_rate": 6.544745331454003e-08, + "loss": 0.408, + "step": 2725 + }, + { + "epoch": 2.6811023622047245, + "grad_norm": 4.746224880218506, + "learning_rate": 6.50772965324904e-08, + "loss": 0.104, + "step": 2726 + }, + { + "epoch": 2.6820866141732282, + "grad_norm": 9.364527702331543, + "learning_rate": 6.470817313350971e-08, + "loss": 0.5343, + "step": 2727 + }, + { + "epoch": 2.6830708661417324, + "grad_norm": 8.605332374572754, + "learning_rate": 6.434008399129279e-08, + "loss": 0.258, + "step": 2728 + }, + { + "epoch": 2.684055118110236, + "grad_norm": 4.468958854675293, + "learning_rate": 6.397302997708641e-08, + "loss": 0.133, + "step": 2729 + }, + { + "epoch": 2.6850393700787403, + "grad_norm": 7.568987846374512, + "learning_rate": 6.360701195968723e-08, + "loss": 0.24, + "step": 2730 + }, + { + "epoch": 2.686023622047244, + "grad_norm": 6.589425086975098, + "learning_rate": 6.324203080543998e-08, + "loss": 0.2113, + "step": 2731 + }, + { + "epoch": 2.687007874015748, + "grad_norm": 11.292271614074707, + "learning_rate": 6.287808737823473e-08, + "loss": 0.5365, + "step": 2732 + }, + { + "epoch": 2.687992125984252, + "grad_norm": 2.5217807292938232, + "learning_rate": 6.251518253950574e-08, + "loss": 0.0353, + "step": 2733 + }, + { + "epoch": 2.688976377952756, + "grad_norm": 6.140711307525635, + "learning_rate": 6.215331714822883e-08, + "loss": 0.145, + "step": 2734 + }, + { + "epoch": 2.6899606299212597, + "grad_norm": 9.195682525634766, + "learning_rate": 6.179249206091951e-08, + "loss": 0.3852, + "step": 2735 + }, + { + "epoch": 2.690944881889764, + "grad_norm": 11.93599796295166, + "learning_rate": 6.143270813163099e-08, + "loss": 0.5938, + "step": 2736 + }, + { + "epoch": 2.6919291338582676, + "grad_norm": 10.107711791992188, + "learning_rate": 6.107396621195201e-08, + "loss": 0.4196, + "step": 2737 + }, + { + "epoch": 2.6929133858267718, + "grad_norm": 14.341265678405762, + "learning_rate": 6.071626715100503e-08, + "loss": 0.7104, + "step": 2738 + }, + { + "epoch": 2.6938976377952755, + "grad_norm": 8.13415241241455, + "learning_rate": 6.035961179544409e-08, + "loss": 0.2068, + "step": 2739 + }, + { + "epoch": 2.6948818897637796, + "grad_norm": 7.959664821624756, + "learning_rate": 6.000400098945283e-08, + "loss": 0.2763, + "step": 2740 + }, + { + "epoch": 2.6958661417322833, + "grad_norm": 11.592156410217285, + "learning_rate": 5.96494355747425e-08, + "loss": 0.4907, + "step": 2741 + }, + { + "epoch": 2.6968503937007875, + "grad_norm": 16.181486129760742, + "learning_rate": 5.9295916390550034e-08, + "loss": 0.5446, + "step": 2742 + }, + { + "epoch": 2.697834645669291, + "grad_norm": 14.021303176879883, + "learning_rate": 5.89434442736358e-08, + "loss": 0.694, + "step": 2743 + }, + { + "epoch": 2.6988188976377954, + "grad_norm": 8.919441223144531, + "learning_rate": 5.859202005828208e-08, + "loss": 0.3986, + "step": 2744 + }, + { + "epoch": 2.699803149606299, + "grad_norm": 14.602743148803711, + "learning_rate": 5.824164457629061e-08, + "loss": 0.6666, + "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 +}