diff --git "a/checkpoint-1220/trainer_state.json" "b/checkpoint-1220/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-1220/trainer_state.json" @@ -0,0 +1,9196 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.2007874015748032, + "eval_steps": 153, + "global_step": 1220, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.000984251968503937, + "grad_norm": NaN, + "learning_rate": 0.0, + "loss": 4.9603, + "step": 1 + }, + { + "epoch": 0.001968503937007874, + "grad_norm": NaN, + "learning_rate": 0.0, + "loss": 28.2529, + "step": 2 + }, + { + "epoch": 0.002952755905511811, + "grad_norm": Infinity, + "learning_rate": 0.0, + "loss": 27.6365, + "step": 3 + }, + { + "epoch": 0.003937007874015748, + "grad_norm": 36.58856201171875, + "learning_rate": 9.940357852882705e-10, + "loss": 6.1387, + "step": 4 + }, + { + "epoch": 0.004921259842519685, + "grad_norm": 34.671714782714844, + "learning_rate": 1.988071570576541e-09, + "loss": 5.5753, + "step": 5 + }, + { + "epoch": 0.005905511811023622, + "grad_norm": 35.53398132324219, + "learning_rate": 2.9821073558648116e-09, + "loss": 5.6951, + "step": 6 + }, + { + "epoch": 0.006889763779527559, + "grad_norm": Infinity, + "learning_rate": 2.9821073558648116e-09, + "loss": 6.3533, + "step": 7 + }, + { + "epoch": 0.007874015748031496, + "grad_norm": Infinity, + "learning_rate": 2.9821073558648116e-09, + "loss": 27.3848, + "step": 8 + }, + { + "epoch": 0.008858267716535433, + "grad_norm": 16.95815086364746, + "learning_rate": 3.976143141153082e-09, + "loss": 3.8501, + "step": 9 + }, + { + "epoch": 0.00984251968503937, + "grad_norm": 69.63166046142578, + "learning_rate": 4.970178926441353e-09, + "loss": 27.911, + "step": 10 + }, + { + "epoch": 0.010826771653543307, + "grad_norm": 29.516401290893555, + "learning_rate": 5.964214711729623e-09, + "loss": 4.9042, + "step": 11 + }, + { + "epoch": 0.011811023622047244, + "grad_norm": 39.30358123779297, + "learning_rate": 6.9582504970178946e-09, + "loss": 6.8003, + "step": 12 + }, + { + "epoch": 0.012795275590551181, + "grad_norm": 36.83983612060547, + "learning_rate": 7.952286282306164e-09, + "loss": 5.7317, + "step": 13 + }, + { + "epoch": 0.013779527559055118, + "grad_norm": 55.4377326965332, + "learning_rate": 8.946322067594435e-09, + "loss": 20.261, + "step": 14 + }, + { + "epoch": 0.014763779527559055, + "grad_norm": 68.62684631347656, + "learning_rate": 9.940357852882705e-09, + "loss": 27.9051, + "step": 15 + }, + { + "epoch": 0.015748031496062992, + "grad_norm": 31.27193832397461, + "learning_rate": 1.0934393638170978e-08, + "loss": 5.5959, + "step": 16 + }, + { + "epoch": 0.01673228346456693, + "grad_norm": 36.56179428100586, + "learning_rate": 1.1928429423459246e-08, + "loss": 5.8052, + "step": 17 + }, + { + "epoch": 0.017716535433070866, + "grad_norm": 23.220964431762695, + "learning_rate": 1.2922465208747517e-08, + "loss": 4.5088, + "step": 18 + }, + { + "epoch": 0.018700787401574805, + "grad_norm": 44.375823974609375, + "learning_rate": 1.3916500994035789e-08, + "loss": 7.3472, + "step": 19 + }, + { + "epoch": 0.01968503937007874, + "grad_norm": 40.480628967285156, + "learning_rate": 1.4910536779324056e-08, + "loss": 5.8668, + "step": 20 + }, + { + "epoch": 0.02066929133858268, + "grad_norm": 45.778358459472656, + "learning_rate": 1.590457256461233e-08, + "loss": 6.4083, + "step": 21 + }, + { + "epoch": 0.021653543307086614, + "grad_norm": 41.16820526123047, + "learning_rate": 1.68986083499006e-08, + "loss": 6.011, + "step": 22 + }, + { + "epoch": 0.022637795275590553, + "grad_norm": 27.49931526184082, + "learning_rate": 1.789264413518887e-08, + "loss": 5.2394, + "step": 23 + }, + { + "epoch": 0.023622047244094488, + "grad_norm": 20.837919235229492, + "learning_rate": 1.888667992047714e-08, + "loss": 4.2966, + "step": 24 + }, + { + "epoch": 0.024606299212598427, + "grad_norm": 65.7834243774414, + "learning_rate": 1.988071570576541e-08, + "loss": 26.605, + "step": 25 + }, + { + "epoch": 0.025590551181102362, + "grad_norm": 44.794960021972656, + "learning_rate": 2.087475149105368e-08, + "loss": 6.2067, + "step": 26 + }, + { + "epoch": 0.0265748031496063, + "grad_norm": 30.213058471679688, + "learning_rate": 2.1868787276341955e-08, + "loss": 6.0346, + "step": 27 + }, + { + "epoch": 0.027559055118110236, + "grad_norm": 27.49605941772461, + "learning_rate": 2.2862823061630224e-08, + "loss": 5.4676, + "step": 28 + }, + { + "epoch": 0.028543307086614175, + "grad_norm": 35.94675827026367, + "learning_rate": 2.3856858846918493e-08, + "loss": 6.4292, + "step": 29 + }, + { + "epoch": 0.02952755905511811, + "grad_norm": 65.74870300292969, + "learning_rate": 2.4850894632206765e-08, + "loss": 26.6452, + "step": 30 + }, + { + "epoch": 0.03051181102362205, + "grad_norm": 49.05170440673828, + "learning_rate": 2.5844930417495034e-08, + "loss": 18.8401, + "step": 31 + }, + { + "epoch": 0.031496062992125984, + "grad_norm": 49.38396453857422, + "learning_rate": 2.6838966202783303e-08, + "loss": 7.4531, + "step": 32 + }, + { + "epoch": 0.03248031496062992, + "grad_norm": 25.29888916015625, + "learning_rate": 2.7833001988071578e-08, + "loss": 4.8286, + "step": 33 + }, + { + "epoch": 0.03346456692913386, + "grad_norm": 28.889131546020508, + "learning_rate": 2.8827037773359847e-08, + "loss": 5.0078, + "step": 34 + }, + { + "epoch": 0.0344488188976378, + "grad_norm": 33.611812591552734, + "learning_rate": 2.982107355864811e-08, + "loss": 5.4115, + "step": 35 + }, + { + "epoch": 0.03543307086614173, + "grad_norm": 31.503385543823242, + "learning_rate": 3.081510934393639e-08, + "loss": 5.4196, + "step": 36 + }, + { + "epoch": 0.03641732283464567, + "grad_norm": 23.436307907104492, + "learning_rate": 3.180914512922466e-08, + "loss": 4.5023, + "step": 37 + }, + { + "epoch": 0.03740157480314961, + "grad_norm": 35.096893310546875, + "learning_rate": 3.280318091451293e-08, + "loss": 5.376, + "step": 38 + }, + { + "epoch": 0.038385826771653545, + "grad_norm": 25.531570434570312, + "learning_rate": 3.37972166998012e-08, + "loss": 5.2303, + "step": 39 + }, + { + "epoch": 0.03937007874015748, + "grad_norm": 29.393512725830078, + "learning_rate": 3.479125248508947e-08, + "loss": 5.6694, + "step": 40 + }, + { + "epoch": 0.040354330708661415, + "grad_norm": 26.839847564697266, + "learning_rate": 3.578528827037774e-08, + "loss": 4.7825, + "step": 41 + }, + { + "epoch": 0.04133858267716536, + "grad_norm": 21.11309814453125, + "learning_rate": 3.6779324055666005e-08, + "loss": 4.6507, + "step": 42 + }, + { + "epoch": 0.04232283464566929, + "grad_norm": 61.134098052978516, + "learning_rate": 3.777335984095428e-08, + "loss": 24.2072, + "step": 43 + }, + { + "epoch": 0.04330708661417323, + "grad_norm": 26.884740829467773, + "learning_rate": 3.8767395626242556e-08, + "loss": 4.9285, + "step": 44 + }, + { + "epoch": 0.04429133858267716, + "grad_norm": 33.500144958496094, + "learning_rate": 3.976143141153082e-08, + "loss": 6.326, + "step": 45 + }, + { + "epoch": 0.045275590551181105, + "grad_norm": 17.54262924194336, + "learning_rate": 4.0755467196819094e-08, + "loss": 4.5724, + "step": 46 + }, + { + "epoch": 0.04625984251968504, + "grad_norm": 23.30596351623535, + "learning_rate": 4.174950298210736e-08, + "loss": 4.754, + "step": 47 + }, + { + "epoch": 0.047244094488188976, + "grad_norm": 34.042816162109375, + "learning_rate": 4.274353876739563e-08, + "loss": 5.5443, + "step": 48 + }, + { + "epoch": 0.04822834645669291, + "grad_norm": 21.270071029663086, + "learning_rate": 4.373757455268391e-08, + "loss": 4.5764, + "step": 49 + }, + { + "epoch": 0.04921259842519685, + "grad_norm": 24.815349578857422, + "learning_rate": 4.4731610337972176e-08, + "loss": 5.1434, + "step": 50 + }, + { + "epoch": 0.05019685039370079, + "grad_norm": 55.756900787353516, + "learning_rate": 4.572564612326045e-08, + "loss": 22.6991, + "step": 51 + }, + { + "epoch": 0.051181102362204724, + "grad_norm": 23.544273376464844, + "learning_rate": 4.6719681908548713e-08, + "loss": 5.4277, + "step": 52 + }, + { + "epoch": 0.05216535433070866, + "grad_norm": 21.845703125, + "learning_rate": 4.7713717693836986e-08, + "loss": 5.0178, + "step": 53 + }, + { + "epoch": 0.0531496062992126, + "grad_norm": 16.331026077270508, + "learning_rate": 4.870775347912525e-08, + "loss": 4.8779, + "step": 54 + }, + { + "epoch": 0.054133858267716536, + "grad_norm": 16.72958755493164, + "learning_rate": 4.970178926441353e-08, + "loss": 4.2884, + "step": 55 + }, + { + "epoch": 0.05511811023622047, + "grad_norm": 41.22899627685547, + "learning_rate": 5.06958250497018e-08, + "loss": 16.0994, + "step": 56 + }, + { + "epoch": 0.05610236220472441, + "grad_norm": 52.578941345214844, + "learning_rate": 5.168986083499007e-08, + "loss": 21.31, + "step": 57 + }, + { + "epoch": 0.05708661417322835, + "grad_norm": 15.741512298583984, + "learning_rate": 5.268389662027834e-08, + "loss": 4.9721, + "step": 58 + }, + { + "epoch": 0.058070866141732284, + "grad_norm": 23.57728385925293, + "learning_rate": 5.3677932405566605e-08, + "loss": 5.143, + "step": 59 + }, + { + "epoch": 0.05905511811023622, + "grad_norm": 12.699495315551758, + "learning_rate": 5.467196819085488e-08, + "loss": 3.5933, + "step": 60 + }, + { + "epoch": 0.060039370078740155, + "grad_norm": 17.49776840209961, + "learning_rate": 5.5666003976143156e-08, + "loss": 5.2559, + "step": 61 + }, + { + "epoch": 0.0610236220472441, + "grad_norm": 13.251837730407715, + "learning_rate": 5.666003976143142e-08, + "loss": 4.0757, + "step": 62 + }, + { + "epoch": 0.06200787401574803, + "grad_norm": 11.610112190246582, + "learning_rate": 5.7654075546719694e-08, + "loss": 3.6612, + "step": 63 + }, + { + "epoch": 0.06299212598425197, + "grad_norm": 19.652385711669922, + "learning_rate": 5.864811133200796e-08, + "loss": 4.7505, + "step": 64 + }, + { + "epoch": 0.0639763779527559, + "grad_norm": 13.930652618408203, + "learning_rate": 5.964214711729623e-08, + "loss": 4.1979, + "step": 65 + }, + { + "epoch": 0.06496062992125984, + "grad_norm": 11.817291259765625, + "learning_rate": 6.06361829025845e-08, + "loss": 3.9982, + "step": 66 + }, + { + "epoch": 0.06594488188976377, + "grad_norm": 18.6019287109375, + "learning_rate": 6.163021868787278e-08, + "loss": 4.7065, + "step": 67 + }, + { + "epoch": 0.06692913385826772, + "grad_norm": 27.056259155273438, + "learning_rate": 6.262425447316104e-08, + "loss": 5.3413, + "step": 68 + }, + { + "epoch": 0.06791338582677166, + "grad_norm": 13.010573387145996, + "learning_rate": 6.361829025844931e-08, + "loss": 3.6964, + "step": 69 + }, + { + "epoch": 0.0688976377952756, + "grad_norm": 49.04872131347656, + "learning_rate": 6.461232604373759e-08, + "loss": 17.8774, + "step": 70 + }, + { + "epoch": 0.06988188976377953, + "grad_norm": 19.028602600097656, + "learning_rate": 6.560636182902586e-08, + "loss": 4.8154, + "step": 71 + }, + { + "epoch": 0.07086614173228346, + "grad_norm": 17.006460189819336, + "learning_rate": 6.660039761431412e-08, + "loss": 4.8356, + "step": 72 + }, + { + "epoch": 0.0718503937007874, + "grad_norm": 17.15074920654297, + "learning_rate": 6.75944333996024e-08, + "loss": 4.568, + "step": 73 + }, + { + "epoch": 0.07283464566929133, + "grad_norm": 14.456765174865723, + "learning_rate": 6.858846918489067e-08, + "loss": 4.0898, + "step": 74 + }, + { + "epoch": 0.07381889763779527, + "grad_norm": 9.999987602233887, + "learning_rate": 6.958250497017893e-08, + "loss": 3.4502, + "step": 75 + }, + { + "epoch": 0.07480314960629922, + "grad_norm": 13.652220726013184, + "learning_rate": 7.057654075546721e-08, + "loss": 3.7733, + "step": 76 + }, + { + "epoch": 0.07578740157480315, + "grad_norm": 17.76757049560547, + "learning_rate": 7.157057654075548e-08, + "loss": 4.5204, + "step": 77 + }, + { + "epoch": 0.07677165354330709, + "grad_norm": 11.42149829864502, + "learning_rate": 7.256461232604374e-08, + "loss": 4.2526, + "step": 78 + }, + { + "epoch": 0.07775590551181102, + "grad_norm": 16.21160125732422, + "learning_rate": 7.355864811133201e-08, + "loss": 4.4398, + "step": 79 + }, + { + "epoch": 0.07874015748031496, + "grad_norm": 12.522687911987305, + "learning_rate": 7.455268389662029e-08, + "loss": 4.0988, + "step": 80 + }, + { + "epoch": 0.0797244094488189, + "grad_norm": 12.63741683959961, + "learning_rate": 7.554671968190855e-08, + "loss": 3.9704, + "step": 81 + }, + { + "epoch": 0.08070866141732283, + "grad_norm": 11.259520530700684, + "learning_rate": 7.654075546719683e-08, + "loss": 4.3343, + "step": 82 + }, + { + "epoch": 0.08169291338582677, + "grad_norm": 14.228102684020996, + "learning_rate": 7.753479125248511e-08, + "loss": 4.2587, + "step": 83 + }, + { + "epoch": 0.08267716535433071, + "grad_norm": 45.13947677612305, + "learning_rate": 7.852882703777338e-08, + "loss": 15.0149, + "step": 84 + }, + { + "epoch": 0.08366141732283465, + "grad_norm": 45.17081069946289, + "learning_rate": 7.952286282306164e-08, + "loss": 14.6599, + "step": 85 + }, + { + "epoch": 0.08464566929133858, + "grad_norm": 15.967412948608398, + "learning_rate": 8.051689860834992e-08, + "loss": 4.0623, + "step": 86 + }, + { + "epoch": 0.08562992125984252, + "grad_norm": 10.085712432861328, + "learning_rate": 8.151093439363819e-08, + "loss": 3.7597, + "step": 87 + }, + { + "epoch": 0.08661417322834646, + "grad_norm": 13.406641960144043, + "learning_rate": 8.250497017892645e-08, + "loss": 4.3433, + "step": 88 + }, + { + "epoch": 0.08759842519685039, + "grad_norm": 9.052105903625488, + "learning_rate": 8.349900596421472e-08, + "loss": 4.0287, + "step": 89 + }, + { + "epoch": 0.08858267716535433, + "grad_norm": 12.489309310913086, + "learning_rate": 8.4493041749503e-08, + "loss": 4.6257, + "step": 90 + }, + { + "epoch": 0.08956692913385826, + "grad_norm": 40.478675842285156, + "learning_rate": 8.548707753479126e-08, + "loss": 13.4689, + "step": 91 + }, + { + "epoch": 0.09055118110236221, + "grad_norm": 14.329568862915039, + "learning_rate": 8.648111332007953e-08, + "loss": 4.6583, + "step": 92 + }, + { + "epoch": 0.09153543307086615, + "grad_norm": 10.07358455657959, + "learning_rate": 8.747514910536782e-08, + "loss": 4.2682, + "step": 93 + }, + { + "epoch": 0.09251968503937008, + "grad_norm": 12.861531257629395, + "learning_rate": 8.846918489065609e-08, + "loss": 4.468, + "step": 94 + }, + { + "epoch": 0.09350393700787402, + "grad_norm": 12.121801376342773, + "learning_rate": 8.946322067594435e-08, + "loss": 3.4333, + "step": 95 + }, + { + "epoch": 0.09448818897637795, + "grad_norm": 40.14929962158203, + "learning_rate": 9.045725646123262e-08, + "loss": 12.7654, + "step": 96 + }, + { + "epoch": 0.09547244094488189, + "grad_norm": 11.191559791564941, + "learning_rate": 9.14512922465209e-08, + "loss": 3.5577, + "step": 97 + }, + { + "epoch": 0.09645669291338582, + "grad_norm": 40.13950729370117, + "learning_rate": 9.244532803180916e-08, + "loss": 12.5875, + "step": 98 + }, + { + "epoch": 0.09744094488188976, + "grad_norm": 12.063894271850586, + "learning_rate": 9.343936381709743e-08, + "loss": 4.2206, + "step": 99 + }, + { + "epoch": 0.0984251968503937, + "grad_norm": 10.066577911376953, + "learning_rate": 9.44333996023857e-08, + "loss": 3.5981, + "step": 100 + }, + { + "epoch": 0.09940944881889764, + "grad_norm": 10.917841911315918, + "learning_rate": 9.542743538767397e-08, + "loss": 3.5575, + "step": 101 + }, + { + "epoch": 0.10039370078740158, + "grad_norm": 11.512818336486816, + "learning_rate": 9.642147117296224e-08, + "loss": 4.0271, + "step": 102 + }, + { + "epoch": 0.10137795275590551, + "grad_norm": 13.737354278564453, + "learning_rate": 9.74155069582505e-08, + "loss": 4.0803, + "step": 103 + }, + { + "epoch": 0.10236220472440945, + "grad_norm": 12.92113208770752, + "learning_rate": 9.840954274353878e-08, + "loss": 4.0886, + "step": 104 + }, + { + "epoch": 0.10334645669291338, + "grad_norm": 16.23849868774414, + "learning_rate": 9.940357852882706e-08, + "loss": 4.176, + "step": 105 + }, + { + "epoch": 0.10433070866141732, + "grad_norm": 13.244183540344238, + "learning_rate": 1.0039761431411533e-07, + "loss": 4.6653, + "step": 106 + }, + { + "epoch": 0.10531496062992125, + "grad_norm": 12.089069366455078, + "learning_rate": 1.013916500994036e-07, + "loss": 4.3076, + "step": 107 + }, + { + "epoch": 0.1062992125984252, + "grad_norm": 28.261154174804688, + "learning_rate": 1.0238568588469187e-07, + "loss": 8.7282, + "step": 108 + }, + { + "epoch": 0.10728346456692914, + "grad_norm": 10.686351776123047, + "learning_rate": 1.0337972166998014e-07, + "loss": 3.4192, + "step": 109 + }, + { + "epoch": 0.10826771653543307, + "grad_norm": 41.12674331665039, + "learning_rate": 1.043737574552684e-07, + "loss": 10.6027, + "step": 110 + }, + { + "epoch": 0.10925196850393701, + "grad_norm": 13.403799057006836, + "learning_rate": 1.0536779324055668e-07, + "loss": 4.0959, + "step": 111 + }, + { + "epoch": 0.11023622047244094, + "grad_norm": 11.321606636047363, + "learning_rate": 1.0636182902584495e-07, + "loss": 4.2785, + "step": 112 + }, + { + "epoch": 0.11122047244094488, + "grad_norm": 14.717891693115234, + "learning_rate": 1.0735586481113321e-07, + "loss": 3.9945, + "step": 113 + }, + { + "epoch": 0.11220472440944881, + "grad_norm": 45.00100326538086, + "learning_rate": 1.0834990059642149e-07, + "loss": 10.0652, + "step": 114 + }, + { + "epoch": 0.11318897637795275, + "grad_norm": 12.017743110656738, + "learning_rate": 1.0934393638170976e-07, + "loss": 3.8621, + "step": 115 + }, + { + "epoch": 0.1141732283464567, + "grad_norm": 14.086198806762695, + "learning_rate": 1.1033797216699802e-07, + "loss": 4.3975, + "step": 116 + }, + { + "epoch": 0.11515748031496063, + "grad_norm": 43.2061767578125, + "learning_rate": 1.1133200795228631e-07, + "loss": 9.7899, + "step": 117 + }, + { + "epoch": 0.11614173228346457, + "grad_norm": 12.647043228149414, + "learning_rate": 1.1232604373757458e-07, + "loss": 4.3812, + "step": 118 + }, + { + "epoch": 0.1171259842519685, + "grad_norm": 11.732013702392578, + "learning_rate": 1.1332007952286284e-07, + "loss": 3.8715, + "step": 119 + }, + { + "epoch": 0.11811023622047244, + "grad_norm": 9.888449668884277, + "learning_rate": 1.1431411530815111e-07, + "loss": 3.8327, + "step": 120 + }, + { + "epoch": 0.11909448818897637, + "grad_norm": 9.061322212219238, + "learning_rate": 1.1530815109343939e-07, + "loss": 3.5103, + "step": 121 + }, + { + "epoch": 0.12007874015748031, + "grad_norm": 38.643943786621094, + "learning_rate": 1.1630218687872765e-07, + "loss": 9.3158, + "step": 122 + }, + { + "epoch": 0.12106299212598425, + "grad_norm": 11.240921974182129, + "learning_rate": 1.1729622266401592e-07, + "loss": 3.7201, + "step": 123 + }, + { + "epoch": 0.1220472440944882, + "grad_norm": 11.231223106384277, + "learning_rate": 1.182902584493042e-07, + "loss": 3.4311, + "step": 124 + }, + { + "epoch": 0.12303149606299213, + "grad_norm": 11.026339530944824, + "learning_rate": 1.1928429423459245e-07, + "loss": 3.7946, + "step": 125 + }, + { + "epoch": 0.12401574803149606, + "grad_norm": 11.620814323425293, + "learning_rate": 1.2027833001988073e-07, + "loss": 4.0456, + "step": 126 + }, + { + "epoch": 0.125, + "grad_norm": 9.652909278869629, + "learning_rate": 1.21272365805169e-07, + "loss": 3.482, + "step": 127 + }, + { + "epoch": 0.12598425196850394, + "grad_norm": 9.82579231262207, + "learning_rate": 1.222664015904573e-07, + "loss": 3.1901, + "step": 128 + }, + { + "epoch": 0.12696850393700787, + "grad_norm": 10.219281196594238, + "learning_rate": 1.2326043737574557e-07, + "loss": 3.414, + "step": 129 + }, + { + "epoch": 0.1279527559055118, + "grad_norm": 9.734150886535645, + "learning_rate": 1.2425447316103382e-07, + "loss": 3.4967, + "step": 130 + }, + { + "epoch": 0.12893700787401574, + "grad_norm": 10.505714416503906, + "learning_rate": 1.2524850894632207e-07, + "loss": 3.6594, + "step": 131 + }, + { + "epoch": 0.12992125984251968, + "grad_norm": 39.98779296875, + "learning_rate": 1.2624254473161035e-07, + "loss": 8.066, + "step": 132 + }, + { + "epoch": 0.1309055118110236, + "grad_norm": 9.36937427520752, + "learning_rate": 1.2723658051689863e-07, + "loss": 3.7872, + "step": 133 + }, + { + "epoch": 0.13188976377952755, + "grad_norm": 10.953369140625, + "learning_rate": 1.282306163021869e-07, + "loss": 4.0023, + "step": 134 + }, + { + "epoch": 0.1328740157480315, + "grad_norm": 11.828601837158203, + "learning_rate": 1.2922465208747519e-07, + "loss": 3.7728, + "step": 135 + }, + { + "epoch": 0.13385826771653545, + "grad_norm": 9.795777320861816, + "learning_rate": 1.3021868787276344e-07, + "loss": 3.1893, + "step": 136 + }, + { + "epoch": 0.13484251968503938, + "grad_norm": 10.846212387084961, + "learning_rate": 1.3121272365805172e-07, + "loss": 3.3635, + "step": 137 + }, + { + "epoch": 0.13582677165354332, + "grad_norm": 13.001614570617676, + "learning_rate": 1.3220675944333997e-07, + "loss": 4.0195, + "step": 138 + }, + { + "epoch": 0.13681102362204725, + "grad_norm": 11.749286651611328, + "learning_rate": 1.3320079522862825e-07, + "loss": 4.1097, + "step": 139 + }, + { + "epoch": 0.1377952755905512, + "grad_norm": 12.08881664276123, + "learning_rate": 1.3419483101391653e-07, + "loss": 3.7903, + "step": 140 + }, + { + "epoch": 0.13877952755905512, + "grad_norm": 10.898722648620605, + "learning_rate": 1.351888667992048e-07, + "loss": 3.5748, + "step": 141 + }, + { + "epoch": 0.13976377952755906, + "grad_norm": 11.456235885620117, + "learning_rate": 1.3618290258449306e-07, + "loss": 3.8104, + "step": 142 + }, + { + "epoch": 0.140748031496063, + "grad_norm": 47.697425842285156, + "learning_rate": 1.3717693836978134e-07, + "loss": 8.0411, + "step": 143 + }, + { + "epoch": 0.14173228346456693, + "grad_norm": 10.220965385437012, + "learning_rate": 1.381709741550696e-07, + "loss": 3.4819, + "step": 144 + }, + { + "epoch": 0.14271653543307086, + "grad_norm": 13.412939071655273, + "learning_rate": 1.3916500994035787e-07, + "loss": 3.452, + "step": 145 + }, + { + "epoch": 0.1437007874015748, + "grad_norm": 11.866227149963379, + "learning_rate": 1.4015904572564615e-07, + "loss": 3.5861, + "step": 146 + }, + { + "epoch": 0.14468503937007873, + "grad_norm": 10.724785804748535, + "learning_rate": 1.4115308151093443e-07, + "loss": 3.4324, + "step": 147 + }, + { + "epoch": 0.14566929133858267, + "grad_norm": 11.023091316223145, + "learning_rate": 1.421471172962227e-07, + "loss": 3.521, + "step": 148 + }, + { + "epoch": 0.1466535433070866, + "grad_norm": 12.216788291931152, + "learning_rate": 1.4314115308151096e-07, + "loss": 3.8868, + "step": 149 + }, + { + "epoch": 0.14763779527559054, + "grad_norm": 71.6353530883789, + "learning_rate": 1.4413518886679924e-07, + "loss": 8.1191, + "step": 150 + }, + { + "epoch": 0.1486220472440945, + "grad_norm": 10.8847074508667, + "learning_rate": 1.451292246520875e-07, + "loss": 3.6447, + "step": 151 + }, + { + "epoch": 0.14960629921259844, + "grad_norm": 11.547443389892578, + "learning_rate": 1.4612326043737577e-07, + "loss": 2.9436, + "step": 152 + }, + { + "epoch": 0.15059055118110237, + "grad_norm": 42.24600601196289, + "learning_rate": 1.4711729622266402e-07, + "loss": 8.1535, + "step": 153 + }, + { + "epoch": 0.15059055118110237, + "eval_Qnli-dev_cosine_accuracy": 0.591796875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.9479926824569702, + "eval_Qnli-dev_cosine_ap": 0.5658036772817674, + "eval_Qnli-dev_cosine_f1": 0.6291834002677376, + "eval_Qnli-dev_cosine_f1_threshold": 0.7761930823326111, + "eval_Qnli-dev_cosine_precision": 0.4598825831702544, + "eval_Qnli-dev_cosine_recall": 0.9957627118644068, + "eval_Qnli-dev_dot_accuracy": 0.59375, + "eval_Qnli-dev_dot_accuracy_threshold": 724.091064453125, + "eval_Qnli-dev_dot_ap": 0.5657459555147606, + "eval_Qnli-dev_dot_f1": 0.6291834002677376, + "eval_Qnli-dev_dot_f1_threshold": 596.2498779296875, + "eval_Qnli-dev_dot_precision": 0.4598825831702544, + "eval_Qnli-dev_dot_recall": 0.9957627118644068, + "eval_Qnli-dev_euclidean_accuracy": 0.591796875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 8.938886642456055, + "eval_Qnli-dev_euclidean_ap": 0.5658036772817674, + "eval_Qnli-dev_euclidean_f1": 0.6291834002677376, + "eval_Qnli-dev_euclidean_f1_threshold": 18.542938232421875, + "eval_Qnli-dev_euclidean_precision": 0.4598825831702544, + "eval_Qnli-dev_euclidean_recall": 0.9957627118644068, + "eval_Qnli-dev_manhattan_accuracy": 0.6171875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 202.07958984375, + "eval_Qnli-dev_manhattan_ap": 0.5891966424964378, + "eval_Qnli-dev_manhattan_f1": 0.6291834002677376, + "eval_Qnli-dev_manhattan_f1_threshold": 307.9236145019531, + "eval_Qnli-dev_manhattan_precision": 0.4598825831702544, + "eval_Qnli-dev_manhattan_recall": 0.9957627118644068, + "eval_Qnli-dev_max_accuracy": 0.6171875, + "eval_Qnli-dev_max_accuracy_threshold": 724.091064453125, + "eval_Qnli-dev_max_ap": 0.5891966424964378, + "eval_Qnli-dev_max_f1": 0.6291834002677376, + "eval_Qnli-dev_max_f1_threshold": 596.2498779296875, + "eval_Qnli-dev_max_precision": 0.4598825831702544, + "eval_Qnli-dev_max_recall": 0.9957627118644068, + "eval_allNLI-dev_cosine_accuracy": 0.666015625, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9797871112823486, + "eval_allNLI-dev_cosine_ap": 0.4008449937025217, + "eval_allNLI-dev_cosine_f1": 0.504258943781942, + "eval_allNLI-dev_cosine_f1_threshold": 0.8929213285446167, + "eval_allNLI-dev_cosine_precision": 0.357487922705314, + "eval_allNLI-dev_cosine_recall": 0.8554913294797688, + "eval_allNLI-dev_dot_accuracy": 0.666015625, + "eval_allNLI-dev_dot_accuracy_threshold": 752.6634521484375, + "eval_allNLI-dev_dot_ap": 0.40071344979441287, + "eval_allNLI-dev_dot_f1": 0.504258943781942, + "eval_allNLI-dev_dot_f1_threshold": 685.9220581054688, + "eval_allNLI-dev_dot_precision": 0.357487922705314, + "eval_allNLI-dev_dot_recall": 0.8554913294797688, + "eval_allNLI-dev_euclidean_accuracy": 0.666015625, + "eval_allNLI-dev_euclidean_accuracy_threshold": 5.572628974914551, + "eval_allNLI-dev_euclidean_ap": 0.40083962142052487, + "eval_allNLI-dev_euclidean_f1": 0.504258943781942, + "eval_allNLI-dev_euclidean_f1_threshold": 12.826179504394531, + "eval_allNLI-dev_euclidean_precision": 0.357487922705314, + "eval_allNLI-dev_euclidean_recall": 0.8554913294797688, + "eval_allNLI-dev_manhattan_accuracy": 0.66796875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 144.52613830566406, + "eval_allNLI-dev_manhattan_ap": 0.4008700157620745, + "eval_allNLI-dev_manhattan_f1": 0.5075987841945289, + "eval_allNLI-dev_manhattan_f1_threshold": 267.046875, + "eval_allNLI-dev_manhattan_precision": 0.3443298969072165, + "eval_allNLI-dev_manhattan_recall": 0.9653179190751445, + "eval_allNLI-dev_max_accuracy": 0.66796875, + "eval_allNLI-dev_max_accuracy_threshold": 752.6634521484375, + "eval_allNLI-dev_max_ap": 0.4008700157620745, + "eval_allNLI-dev_max_f1": 0.5075987841945289, + "eval_allNLI-dev_max_f1_threshold": 685.9220581054688, + "eval_allNLI-dev_max_precision": 0.357487922705314, + "eval_allNLI-dev_max_recall": 0.9653179190751445, + "eval_loss": 2.203232526779175, + "eval_runtime": 50.3824, + "eval_samples_per_second": 33.027, + "eval_sequential_score": 0.5891966424964378, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.1561600438268545, + "eval_sts-test_pearson_dot": 0.15588248423807516, + "eval_sts-test_pearson_euclidean": 0.1908690981304929, + "eval_sts-test_pearson_manhattan": 0.2216924674035587, + "eval_sts-test_pearson_max": 0.2216924674035587, + "eval_sts-test_spearman_cosine": 0.22356441354815124, + "eval_sts-test_spearman_dot": 0.22337189362164545, + "eval_sts-test_spearman_euclidean": 0.22363767136304896, + "eval_sts-test_spearman_manhattan": 0.24997065610359018, + "eval_sts-test_spearman_max": 0.24997065610359018, + "step": 153 + }, + { + "epoch": 0.1515748031496063, + "grad_norm": 13.303340911865234, + "learning_rate": 1.4811133200795232e-07, + "loss": 3.9619, + "step": 154 + }, + { + "epoch": 0.15255905511811024, + "grad_norm": 10.895444869995117, + "learning_rate": 1.4910536779324058e-07, + "loss": 3.1301, + "step": 155 + }, + { + "epoch": 0.15354330708661418, + "grad_norm": 9.847289085388184, + "learning_rate": 1.5009940357852886e-07, + "loss": 3.0478, + "step": 156 + }, + { + "epoch": 0.1545275590551181, + "grad_norm": 9.858047485351562, + "learning_rate": 1.510934393638171e-07, + "loss": 3.2986, + "step": 157 + }, + { + "epoch": 0.15551181102362205, + "grad_norm": 11.554423332214355, + "learning_rate": 1.5208747514910539e-07, + "loss": 3.2847, + "step": 158 + }, + { + "epoch": 0.15649606299212598, + "grad_norm": 11.145759582519531, + "learning_rate": 1.5308151093439367e-07, + "loss": 3.6599, + "step": 159 + }, + { + "epoch": 0.15748031496062992, + "grad_norm": 9.598219871520996, + "learning_rate": 1.5407554671968192e-07, + "loss": 3.2238, + "step": 160 + }, + { + "epoch": 0.15846456692913385, + "grad_norm": 11.533960342407227, + "learning_rate": 1.5506958250497022e-07, + "loss": 2.8897, + "step": 161 + }, + { + "epoch": 0.1594488188976378, + "grad_norm": 12.17159652709961, + "learning_rate": 1.5606361829025848e-07, + "loss": 3.9443, + "step": 162 + }, + { + "epoch": 0.16043307086614172, + "grad_norm": 10.684307098388672, + "learning_rate": 1.5705765407554675e-07, + "loss": 3.3733, + "step": 163 + }, + { + "epoch": 0.16141732283464566, + "grad_norm": 12.093358039855957, + "learning_rate": 1.58051689860835e-07, + "loss": 3.7444, + "step": 164 + }, + { + "epoch": 0.1624015748031496, + "grad_norm": 12.204547882080078, + "learning_rate": 1.5904572564612329e-07, + "loss": 3.4813, + "step": 165 + }, + { + "epoch": 0.16338582677165353, + "grad_norm": 11.32477855682373, + "learning_rate": 1.6003976143141154e-07, + "loss": 2.6865, + "step": 166 + }, + { + "epoch": 0.1643700787401575, + "grad_norm": 10.9214506149292, + "learning_rate": 1.6103379721669984e-07, + "loss": 2.7587, + "step": 167 + }, + { + "epoch": 0.16535433070866143, + "grad_norm": 10.949960708618164, + "learning_rate": 1.620278330019881e-07, + "loss": 3.3628, + "step": 168 + }, + { + "epoch": 0.16633858267716536, + "grad_norm": 12.229423522949219, + "learning_rate": 1.6302186878727637e-07, + "loss": 3.0035, + "step": 169 + }, + { + "epoch": 0.1673228346456693, + "grad_norm": 77.52136993408203, + "learning_rate": 1.6401590457256465e-07, + "loss": 10.1591, + "step": 170 + }, + { + "epoch": 0.16830708661417323, + "grad_norm": 13.680435180664062, + "learning_rate": 1.650099403578529e-07, + "loss": 3.5366, + "step": 171 + }, + { + "epoch": 0.16929133858267717, + "grad_norm": 51.19025421142578, + "learning_rate": 1.6600397614314118e-07, + "loss": 8.4047, + "step": 172 + }, + { + "epoch": 0.1702755905511811, + "grad_norm": 12.889047622680664, + "learning_rate": 1.6699801192842944e-07, + "loss": 3.8643, + "step": 173 + }, + { + "epoch": 0.17125984251968504, + "grad_norm": 12.575840950012207, + "learning_rate": 1.6799204771371774e-07, + "loss": 3.3529, + "step": 174 + }, + { + "epoch": 0.17224409448818898, + "grad_norm": 11.404970169067383, + "learning_rate": 1.68986083499006e-07, + "loss": 3.7143, + "step": 175 + }, + { + "epoch": 0.1732283464566929, + "grad_norm": 11.78778076171875, + "learning_rate": 1.6998011928429427e-07, + "loss": 3.3323, + "step": 176 + }, + { + "epoch": 0.17421259842519685, + "grad_norm": 11.07532787322998, + "learning_rate": 1.7097415506958253e-07, + "loss": 3.1206, + "step": 177 + }, + { + "epoch": 0.17519685039370078, + "grad_norm": 13.040700912475586, + "learning_rate": 1.719681908548708e-07, + "loss": 3.1348, + "step": 178 + }, + { + "epoch": 0.17618110236220472, + "grad_norm": 49.289466857910156, + "learning_rate": 1.7296222664015906e-07, + "loss": 7.6011, + "step": 179 + }, + { + "epoch": 0.17716535433070865, + "grad_norm": 14.549922943115234, + "learning_rate": 1.7395626242544734e-07, + "loss": 3.7025, + "step": 180 + }, + { + "epoch": 0.1781496062992126, + "grad_norm": 104.96332550048828, + "learning_rate": 1.7495029821073564e-07, + "loss": 10.5662, + "step": 181 + }, + { + "epoch": 0.17913385826771652, + "grad_norm": 61.00530242919922, + "learning_rate": 1.759443339960239e-07, + "loss": 8.966, + "step": 182 + }, + { + "epoch": 0.18011811023622049, + "grad_norm": 68.94571685791016, + "learning_rate": 1.7693836978131217e-07, + "loss": 9.426, + "step": 183 + }, + { + "epoch": 0.18110236220472442, + "grad_norm": 11.119109153747559, + "learning_rate": 1.7793240556660042e-07, + "loss": 3.0025, + "step": 184 + }, + { + "epoch": 0.18208661417322836, + "grad_norm": 34.247840881347656, + "learning_rate": 1.789264413518887e-07, + "loss": 7.0984, + "step": 185 + }, + { + "epoch": 0.1830708661417323, + "grad_norm": 47.956634521484375, + "learning_rate": 1.7992047713717695e-07, + "loss": 7.3808, + "step": 186 + }, + { + "epoch": 0.18405511811023623, + "grad_norm": 11.798696517944336, + "learning_rate": 1.8091451292246523e-07, + "loss": 2.8657, + "step": 187 + }, + { + "epoch": 0.18503937007874016, + "grad_norm": 38.751102447509766, + "learning_rate": 1.819085487077535e-07, + "loss": 6.5636, + "step": 188 + }, + { + "epoch": 0.1860236220472441, + "grad_norm": 14.265003204345703, + "learning_rate": 1.829025844930418e-07, + "loss": 3.4702, + "step": 189 + }, + { + "epoch": 0.18700787401574803, + "grad_norm": 35.365360260009766, + "learning_rate": 1.8389662027833004e-07, + "loss": 5.9302, + "step": 190 + }, + { + "epoch": 0.18799212598425197, + "grad_norm": 10.978341102600098, + "learning_rate": 1.8489065606361832e-07, + "loss": 3.2406, + "step": 191 + }, + { + "epoch": 0.1889763779527559, + "grad_norm": 12.03227710723877, + "learning_rate": 1.8588469184890657e-07, + "loss": 3.4459, + "step": 192 + }, + { + "epoch": 0.18996062992125984, + "grad_norm": 21.640823364257812, + "learning_rate": 1.8687872763419485e-07, + "loss": 5.269, + "step": 193 + }, + { + "epoch": 0.19094488188976377, + "grad_norm": 21.88094139099121, + "learning_rate": 1.8787276341948313e-07, + "loss": 4.8605, + "step": 194 + }, + { + "epoch": 0.1919291338582677, + "grad_norm": 10.063691139221191, + "learning_rate": 1.888667992047714e-07, + "loss": 2.9891, + "step": 195 + }, + { + "epoch": 0.19291338582677164, + "grad_norm": 10.375823974609375, + "learning_rate": 1.898608349900597e-07, + "loss": 3.6681, + "step": 196 + }, + { + "epoch": 0.19389763779527558, + "grad_norm": 10.918342590332031, + "learning_rate": 1.9085487077534794e-07, + "loss": 3.1589, + "step": 197 + }, + { + "epoch": 0.19488188976377951, + "grad_norm": 12.367669105529785, + "learning_rate": 1.9184890656063622e-07, + "loss": 3.1835, + "step": 198 + }, + { + "epoch": 0.19586614173228348, + "grad_norm": 12.073822021484375, + "learning_rate": 1.9284294234592447e-07, + "loss": 3.7561, + "step": 199 + }, + { + "epoch": 0.1968503937007874, + "grad_norm": 12.768759727478027, + "learning_rate": 1.9383697813121275e-07, + "loss": 4.0891, + "step": 200 + }, + { + "epoch": 0.19783464566929135, + "grad_norm": 12.075078964233398, + "learning_rate": 1.94831013916501e-07, + "loss": 3.563, + "step": 201 + }, + { + "epoch": 0.19881889763779528, + "grad_norm": 13.486553192138672, + "learning_rate": 1.958250497017893e-07, + "loss": 3.7433, + "step": 202 + }, + { + "epoch": 0.19980314960629922, + "grad_norm": 12.1553955078125, + "learning_rate": 1.9681908548707756e-07, + "loss": 3.3813, + "step": 203 + }, + { + "epoch": 0.20078740157480315, + "grad_norm": 23.534427642822266, + "learning_rate": 1.9781312127236584e-07, + "loss": 5.2311, + "step": 204 + }, + { + "epoch": 0.2017716535433071, + "grad_norm": 11.403944969177246, + "learning_rate": 1.9880715705765412e-07, + "loss": 3.3494, + "step": 205 + }, + { + "epoch": 0.20275590551181102, + "grad_norm": 10.039087295532227, + "learning_rate": 1.9980119284294237e-07, + "loss": 3.3533, + "step": 206 + }, + { + "epoch": 0.20374015748031496, + "grad_norm": 11.510744094848633, + "learning_rate": 2.0079522862823065e-07, + "loss": 3.688, + "step": 207 + }, + { + "epoch": 0.2047244094488189, + "grad_norm": 10.009934425354004, + "learning_rate": 2.017892644135189e-07, + "loss": 3.5342, + "step": 208 + }, + { + "epoch": 0.20570866141732283, + "grad_norm": 21.598899841308594, + "learning_rate": 2.027833001988072e-07, + "loss": 4.9381, + "step": 209 + }, + { + "epoch": 0.20669291338582677, + "grad_norm": 10.349525451660156, + "learning_rate": 2.0377733598409546e-07, + "loss": 3.1839, + "step": 210 + }, + { + "epoch": 0.2076771653543307, + "grad_norm": 10.374922752380371, + "learning_rate": 2.0477137176938374e-07, + "loss": 3.0465, + "step": 211 + }, + { + "epoch": 0.20866141732283464, + "grad_norm": 12.201885223388672, + "learning_rate": 2.05765407554672e-07, + "loss": 3.1232, + "step": 212 + }, + { + "epoch": 0.20964566929133857, + "grad_norm": 21.732847213745117, + "learning_rate": 2.0675944333996027e-07, + "loss": 4.6297, + "step": 213 + }, + { + "epoch": 0.2106299212598425, + "grad_norm": 9.907038688659668, + "learning_rate": 2.0775347912524852e-07, + "loss": 2.9834, + "step": 214 + }, + { + "epoch": 0.21161417322834647, + "grad_norm": 17.139137268066406, + "learning_rate": 2.087475149105368e-07, + "loss": 4.2231, + "step": 215 + }, + { + "epoch": 0.2125984251968504, + "grad_norm": 12.830426216125488, + "learning_rate": 2.097415506958251e-07, + "loss": 3.1458, + "step": 216 + }, + { + "epoch": 0.21358267716535434, + "grad_norm": 10.250080108642578, + "learning_rate": 2.1073558648111336e-07, + "loss": 3.2525, + "step": 217 + }, + { + "epoch": 0.21456692913385828, + "grad_norm": 11.067953109741211, + "learning_rate": 2.1172962226640164e-07, + "loss": 3.5971, + "step": 218 + }, + { + "epoch": 0.2155511811023622, + "grad_norm": 10.0042724609375, + "learning_rate": 2.127236580516899e-07, + "loss": 3.5616, + "step": 219 + }, + { + "epoch": 0.21653543307086615, + "grad_norm": 10.706488609313965, + "learning_rate": 2.1371769383697817e-07, + "loss": 3.2378, + "step": 220 + }, + { + "epoch": 0.21751968503937008, + "grad_norm": 12.0921630859375, + "learning_rate": 2.1471172962226642e-07, + "loss": 2.9075, + "step": 221 + }, + { + "epoch": 0.21850393700787402, + "grad_norm": 11.22988510131836, + "learning_rate": 2.1570576540755473e-07, + "loss": 3.0391, + "step": 222 + }, + { + "epoch": 0.21948818897637795, + "grad_norm": 11.993961334228516, + "learning_rate": 2.1669980119284298e-07, + "loss": 3.5573, + "step": 223 + }, + { + "epoch": 0.2204724409448819, + "grad_norm": 11.548995971679688, + "learning_rate": 2.1769383697813126e-07, + "loss": 3.2092, + "step": 224 + }, + { + "epoch": 0.22145669291338582, + "grad_norm": 9.982552528381348, + "learning_rate": 2.186878727634195e-07, + "loss": 3.2646, + "step": 225 + }, + { + "epoch": 0.22244094488188976, + "grad_norm": 9.191473007202148, + "learning_rate": 2.196819085487078e-07, + "loss": 3.0886, + "step": 226 + }, + { + "epoch": 0.2234251968503937, + "grad_norm": 14.892210006713867, + "learning_rate": 2.2067594433399604e-07, + "loss": 3.5241, + "step": 227 + }, + { + "epoch": 0.22440944881889763, + "grad_norm": 12.9964599609375, + "learning_rate": 2.2166998011928432e-07, + "loss": 3.0111, + "step": 228 + }, + { + "epoch": 0.22539370078740156, + "grad_norm": 14.988781929016113, + "learning_rate": 2.2266401590457263e-07, + "loss": 3.707, + "step": 229 + }, + { + "epoch": 0.2263779527559055, + "grad_norm": 35.95463180541992, + "learning_rate": 2.2365805168986088e-07, + "loss": 5.3822, + "step": 230 + }, + { + "epoch": 0.22736220472440946, + "grad_norm": 14.787928581237793, + "learning_rate": 2.2465208747514916e-07, + "loss": 3.2646, + "step": 231 + }, + { + "epoch": 0.2283464566929134, + "grad_norm": 10.399752616882324, + "learning_rate": 2.256461232604374e-07, + "loss": 2.7021, + "step": 232 + }, + { + "epoch": 0.22933070866141733, + "grad_norm": 12.82553482055664, + "learning_rate": 2.266401590457257e-07, + "loss": 3.5131, + "step": 233 + }, + { + "epoch": 0.23031496062992127, + "grad_norm": 13.701425552368164, + "learning_rate": 2.2763419483101394e-07, + "loss": 3.103, + "step": 234 + }, + { + "epoch": 0.2312992125984252, + "grad_norm": 11.487199783325195, + "learning_rate": 2.2862823061630222e-07, + "loss": 2.9535, + "step": 235 + }, + { + "epoch": 0.23228346456692914, + "grad_norm": 11.957090377807617, + "learning_rate": 2.296222664015905e-07, + "loss": 2.9631, + "step": 236 + }, + { + "epoch": 0.23326771653543307, + "grad_norm": 11.277992248535156, + "learning_rate": 2.3061630218687878e-07, + "loss": 2.8068, + "step": 237 + }, + { + "epoch": 0.234251968503937, + "grad_norm": 14.584067344665527, + "learning_rate": 2.3161033797216703e-07, + "loss": 3.4251, + "step": 238 + }, + { + "epoch": 0.23523622047244094, + "grad_norm": 12.476365089416504, + "learning_rate": 2.326043737574553e-07, + "loss": 2.8495, + "step": 239 + }, + { + "epoch": 0.23622047244094488, + "grad_norm": 10.975320816040039, + "learning_rate": 2.335984095427436e-07, + "loss": 2.9972, + "step": 240 + }, + { + "epoch": 0.2372047244094488, + "grad_norm": 15.740063667297363, + "learning_rate": 2.3459244532803184e-07, + "loss": 3.3509, + "step": 241 + }, + { + "epoch": 0.23818897637795275, + "grad_norm": 11.939135551452637, + "learning_rate": 2.3558648111332012e-07, + "loss": 2.9234, + "step": 242 + }, + { + "epoch": 0.23917322834645668, + "grad_norm": 11.509920120239258, + "learning_rate": 2.365805168986084e-07, + "loss": 2.4086, + "step": 243 + }, + { + "epoch": 0.24015748031496062, + "grad_norm": 12.131606101989746, + "learning_rate": 2.3757455268389668e-07, + "loss": 3.1282, + "step": 244 + }, + { + "epoch": 0.24114173228346455, + "grad_norm": 10.977266311645508, + "learning_rate": 2.385685884691849e-07, + "loss": 2.3352, + "step": 245 + }, + { + "epoch": 0.2421259842519685, + "grad_norm": 13.836033821105957, + "learning_rate": 2.395626242544732e-07, + "loss": 2.4706, + "step": 246 + }, + { + "epoch": 0.24311023622047245, + "grad_norm": 14.244443893432617, + "learning_rate": 2.4055666003976146e-07, + "loss": 3.5449, + "step": 247 + }, + { + "epoch": 0.2440944881889764, + "grad_norm": 13.271723747253418, + "learning_rate": 2.4155069582504976e-07, + "loss": 2.8963, + "step": 248 + }, + { + "epoch": 0.24507874015748032, + "grad_norm": 11.74501895904541, + "learning_rate": 2.42544731610338e-07, + "loss": 2.773, + "step": 249 + }, + { + "epoch": 0.24606299212598426, + "grad_norm": 11.362481117248535, + "learning_rate": 2.4353876739562627e-07, + "loss": 2.355, + "step": 250 + }, + { + "epoch": 0.2470472440944882, + "grad_norm": 14.688889503479004, + "learning_rate": 2.445328031809146e-07, + "loss": 2.656, + "step": 251 + }, + { + "epoch": 0.24803149606299213, + "grad_norm": 10.654073715209961, + "learning_rate": 2.4552683896620283e-07, + "loss": 2.6221, + "step": 252 + }, + { + "epoch": 0.24901574803149606, + "grad_norm": 54.549888610839844, + "learning_rate": 2.4652087475149113e-07, + "loss": 8.6739, + "step": 253 + }, + { + "epoch": 0.25, + "grad_norm": 96.43404388427734, + "learning_rate": 2.475149105367794e-07, + "loss": 10.8242, + "step": 254 + }, + { + "epoch": 0.25098425196850394, + "grad_norm": 12.798384666442871, + "learning_rate": 2.4850894632206764e-07, + "loss": 2.3408, + "step": 255 + }, + { + "epoch": 0.25196850393700787, + "grad_norm": 11.494848251342773, + "learning_rate": 2.495029821073559e-07, + "loss": 2.1221, + "step": 256 + }, + { + "epoch": 0.2529527559055118, + "grad_norm": 18.67085075378418, + "learning_rate": 2.5049701789264414e-07, + "loss": 3.295, + "step": 257 + }, + { + "epoch": 0.25393700787401574, + "grad_norm": 12.794750213623047, + "learning_rate": 2.5149105367793245e-07, + "loss": 2.5896, + "step": 258 + }, + { + "epoch": 0.2549212598425197, + "grad_norm": 13.11160659790039, + "learning_rate": 2.524850894632207e-07, + "loss": 2.1215, + "step": 259 + }, + { + "epoch": 0.2559055118110236, + "grad_norm": 82.75247192382812, + "learning_rate": 2.53479125248509e-07, + "loss": 9.4851, + "step": 260 + }, + { + "epoch": 0.25688976377952755, + "grad_norm": 10.959670066833496, + "learning_rate": 2.5447316103379726e-07, + "loss": 2.1982, + "step": 261 + }, + { + "epoch": 0.2578740157480315, + "grad_norm": 35.51136016845703, + "learning_rate": 2.554671968190855e-07, + "loss": 3.0568, + "step": 262 + }, + { + "epoch": 0.2588582677165354, + "grad_norm": 11.970784187316895, + "learning_rate": 2.564612326043738e-07, + "loss": 2.6269, + "step": 263 + }, + { + "epoch": 0.25984251968503935, + "grad_norm": 13.664376258850098, + "learning_rate": 2.5745526838966207e-07, + "loss": 2.4792, + "step": 264 + }, + { + "epoch": 0.2608267716535433, + "grad_norm": 10.64925479888916, + "learning_rate": 2.5844930417495037e-07, + "loss": 1.9445, + "step": 265 + }, + { + "epoch": 0.2618110236220472, + "grad_norm": 13.321276664733887, + "learning_rate": 2.5944333996023857e-07, + "loss": 2.4061, + "step": 266 + }, + { + "epoch": 0.26279527559055116, + "grad_norm": 77.70325469970703, + "learning_rate": 2.604373757455269e-07, + "loss": 8.3116, + "step": 267 + }, + { + "epoch": 0.2637795275590551, + "grad_norm": 53.12438201904297, + "learning_rate": 2.614314115308152e-07, + "loss": 8.0804, + "step": 268 + }, + { + "epoch": 0.26476377952755903, + "grad_norm": 10.435575485229492, + "learning_rate": 2.6242544731610343e-07, + "loss": 2.1674, + "step": 269 + }, + { + "epoch": 0.265748031496063, + "grad_norm": 51.96613311767578, + "learning_rate": 2.634194831013917e-07, + "loss": 7.1975, + "step": 270 + }, + { + "epoch": 0.26673228346456695, + "grad_norm": 46.066497802734375, + "learning_rate": 2.6441351888667994e-07, + "loss": 5.9104, + "step": 271 + }, + { + "epoch": 0.2677165354330709, + "grad_norm": 11.553542137145996, + "learning_rate": 2.6540755467196824e-07, + "loss": 2.498, + "step": 272 + }, + { + "epoch": 0.2687007874015748, + "grad_norm": 12.689590454101562, + "learning_rate": 2.664015904572565e-07, + "loss": 2.5249, + "step": 273 + }, + { + "epoch": 0.26968503937007876, + "grad_norm": 11.272449493408203, + "learning_rate": 2.6739562624254475e-07, + "loss": 2.7152, + "step": 274 + }, + { + "epoch": 0.2706692913385827, + "grad_norm": 11.380216598510742, + "learning_rate": 2.6838966202783305e-07, + "loss": 2.7904, + "step": 275 + }, + { + "epoch": 0.27165354330708663, + "grad_norm": 13.75843334197998, + "learning_rate": 2.693836978131213e-07, + "loss": 2.7745, + "step": 276 + }, + { + "epoch": 0.27263779527559057, + "grad_norm": 13.37204647064209, + "learning_rate": 2.703777335984096e-07, + "loss": 2.9741, + "step": 277 + }, + { + "epoch": 0.2736220472440945, + "grad_norm": 11.115955352783203, + "learning_rate": 2.7137176938369786e-07, + "loss": 1.8215, + "step": 278 + }, + { + "epoch": 0.27460629921259844, + "grad_norm": 26.816680908203125, + "learning_rate": 2.723658051689861e-07, + "loss": 4.6844, + "step": 279 + }, + { + "epoch": 0.2755905511811024, + "grad_norm": 12.469667434692383, + "learning_rate": 2.7335984095427437e-07, + "loss": 2.8613, + "step": 280 + }, + { + "epoch": 0.2765748031496063, + "grad_norm": 12.972196578979492, + "learning_rate": 2.743538767395627e-07, + "loss": 2.7147, + "step": 281 + }, + { + "epoch": 0.27755905511811024, + "grad_norm": 12.858671188354492, + "learning_rate": 2.75347912524851e-07, + "loss": 2.814, + "step": 282 + }, + { + "epoch": 0.2785433070866142, + "grad_norm": 10.301759719848633, + "learning_rate": 2.763419483101392e-07, + "loss": 2.3569, + "step": 283 + }, + { + "epoch": 0.2795275590551181, + "grad_norm": 12.948614120483398, + "learning_rate": 2.773359840954275e-07, + "loss": 2.672, + "step": 284 + }, + { + "epoch": 0.28051181102362205, + "grad_norm": 16.839580535888672, + "learning_rate": 2.7833001988071574e-07, + "loss": 3.2052, + "step": 285 + }, + { + "epoch": 0.281496062992126, + "grad_norm": 12.991905212402344, + "learning_rate": 2.7932405566600404e-07, + "loss": 2.8056, + "step": 286 + }, + { + "epoch": 0.2824803149606299, + "grad_norm": 12.984047889709473, + "learning_rate": 2.803180914512923e-07, + "loss": 2.6268, + "step": 287 + }, + { + "epoch": 0.28346456692913385, + "grad_norm": 12.771299362182617, + "learning_rate": 2.8131212723658055e-07, + "loss": 2.5641, + "step": 288 + }, + { + "epoch": 0.2844488188976378, + "grad_norm": 12.763790130615234, + "learning_rate": 2.8230616302186885e-07, + "loss": 2.4475, + "step": 289 + }, + { + "epoch": 0.2854330708661417, + "grad_norm": 12.817102432250977, + "learning_rate": 2.833001988071571e-07, + "loss": 2.7377, + "step": 290 + }, + { + "epoch": 0.28641732283464566, + "grad_norm": 11.62403678894043, + "learning_rate": 2.842942345924454e-07, + "loss": 2.3831, + "step": 291 + }, + { + "epoch": 0.2874015748031496, + "grad_norm": 88.97967529296875, + "learning_rate": 2.852882703777336e-07, + "loss": 8.8069, + "step": 292 + }, + { + "epoch": 0.28838582677165353, + "grad_norm": 12.380749702453613, + "learning_rate": 2.862823061630219e-07, + "loss": 2.186, + "step": 293 + }, + { + "epoch": 0.28937007874015747, + "grad_norm": 12.181745529174805, + "learning_rate": 2.8727634194831017e-07, + "loss": 2.3389, + "step": 294 + }, + { + "epoch": 0.2903543307086614, + "grad_norm": 11.23538875579834, + "learning_rate": 2.8827037773359847e-07, + "loss": 1.9744, + "step": 295 + }, + { + "epoch": 0.29133858267716534, + "grad_norm": 13.454959869384766, + "learning_rate": 2.892644135188867e-07, + "loss": 2.4491, + "step": 296 + }, + { + "epoch": 0.29232283464566927, + "grad_norm": 12.226387977600098, + "learning_rate": 2.90258449304175e-07, + "loss": 2.5668, + "step": 297 + }, + { + "epoch": 0.2933070866141732, + "grad_norm": 13.08324146270752, + "learning_rate": 2.912524850894633e-07, + "loss": 2.1939, + "step": 298 + }, + { + "epoch": 0.29429133858267714, + "grad_norm": 12.1226224899292, + "learning_rate": 2.9224652087475153e-07, + "loss": 2.2832, + "step": 299 + }, + { + "epoch": 0.2952755905511811, + "grad_norm": 12.738725662231445, + "learning_rate": 2.9324055666003984e-07, + "loss": 2.7508, + "step": 300 + }, + { + "epoch": 0.296259842519685, + "grad_norm": 13.919729232788086, + "learning_rate": 2.9423459244532804e-07, + "loss": 2.5206, + "step": 301 + }, + { + "epoch": 0.297244094488189, + "grad_norm": 13.623347282409668, + "learning_rate": 2.9522862823061634e-07, + "loss": 2.3522, + "step": 302 + }, + { + "epoch": 0.29822834645669294, + "grad_norm": 15.347495079040527, + "learning_rate": 2.9622266401590465e-07, + "loss": 2.7186, + "step": 303 + }, + { + "epoch": 0.2992125984251969, + "grad_norm": 13.009486198425293, + "learning_rate": 2.972166998011929e-07, + "loss": 2.1369, + "step": 304 + }, + { + "epoch": 0.3001968503937008, + "grad_norm": 77.14462280273438, + "learning_rate": 2.9821073558648115e-07, + "loss": 9.7972, + "step": 305 + }, + { + "epoch": 0.30118110236220474, + "grad_norm": 11.583063125610352, + "learning_rate": 2.992047713717694e-07, + "loss": 1.9378, + "step": 306 + }, + { + "epoch": 0.30118110236220474, + "eval_Qnli-dev_cosine_accuracy": 0.611328125, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8805997967720032, + "eval_Qnli-dev_cosine_ap": 0.6040598080761177, + "eval_Qnli-dev_cosine_f1": 0.6359945872801083, + "eval_Qnli-dev_cosine_f1_threshold": 0.7411552667617798, + "eval_Qnli-dev_cosine_precision": 0.4671968190854871, + "eval_Qnli-dev_cosine_recall": 0.9957627118644068, + "eval_Qnli-dev_dot_accuracy": 0.611328125, + "eval_Qnli-dev_dot_accuracy_threshold": 676.0317993164062, + "eval_Qnli-dev_dot_ap": 0.6037363342775177, + "eval_Qnli-dev_dot_f1": 0.6359945872801083, + "eval_Qnli-dev_dot_f1_threshold": 568.9611206054688, + "eval_Qnli-dev_dot_precision": 0.4671968190854871, + "eval_Qnli-dev_dot_recall": 0.9957627118644068, + "eval_Qnli-dev_euclidean_accuracy": 0.611328125, + "eval_Qnli-dev_euclidean_accuracy_threshold": 13.539779663085938, + "eval_Qnli-dev_euclidean_ap": 0.6040065895497431, + "eval_Qnli-dev_euclidean_f1": 0.6359945872801083, + "eval_Qnli-dev_euclidean_f1_threshold": 19.93484115600586, + "eval_Qnli-dev_euclidean_precision": 0.4671968190854871, + "eval_Qnli-dev_euclidean_recall": 0.9957627118644068, + "eval_Qnli-dev_manhattan_accuracy": 0.619140625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 255.49671936035156, + "eval_Qnli-dev_manhattan_ap": 0.6159146231463524, + "eval_Qnli-dev_manhattan_f1": 0.6444121915820029, + "eval_Qnli-dev_manhattan_f1_threshold": 334.76043701171875, + "eval_Qnli-dev_manhattan_precision": 0.4900662251655629, + "eval_Qnli-dev_manhattan_recall": 0.940677966101695, + "eval_Qnli-dev_max_accuracy": 0.619140625, + "eval_Qnli-dev_max_accuracy_threshold": 676.0317993164062, + "eval_Qnli-dev_max_ap": 0.6159146231463524, + "eval_Qnli-dev_max_f1": 0.6444121915820029, + "eval_Qnli-dev_max_f1_threshold": 568.9611206054688, + "eval_Qnli-dev_max_precision": 0.4900662251655629, + "eval_Qnli-dev_max_recall": 0.9957627118644068, + "eval_allNLI-dev_cosine_accuracy": 0.673828125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9558601379394531, + "eval_allNLI-dev_cosine_ap": 0.427219753620832, + "eval_allNLI-dev_cosine_f1": 0.5138211382113822, + "eval_allNLI-dev_cosine_f1_threshold": 0.8555588722229004, + "eval_allNLI-dev_cosine_precision": 0.3574660633484163, + "eval_allNLI-dev_cosine_recall": 0.9132947976878613, + "eval_allNLI-dev_dot_accuracy": 0.673828125, + "eval_allNLI-dev_dot_accuracy_threshold": 733.9632568359375, + "eval_allNLI-dev_dot_ap": 0.4268777253440225, + "eval_allNLI-dev_dot_f1": 0.5138211382113822, + "eval_allNLI-dev_dot_f1_threshold": 656.796875, + "eval_allNLI-dev_dot_precision": 0.3574660633484163, + "eval_allNLI-dev_dot_recall": 0.9132947976878613, + "eval_allNLI-dev_euclidean_accuracy": 0.673828125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 8.233429908752441, + "eval_allNLI-dev_euclidean_ap": 0.42721636257991424, + "eval_allNLI-dev_euclidean_f1": 0.5138211382113822, + "eval_allNLI-dev_euclidean_f1_threshold": 14.891927719116211, + "eval_allNLI-dev_euclidean_precision": 0.3574660633484163, + "eval_allNLI-dev_euclidean_recall": 0.9132947976878613, + "eval_allNLI-dev_manhattan_accuracy": 0.669921875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 148.85415649414062, + "eval_allNLI-dev_manhattan_ap": 0.4231015555464324, + "eval_allNLI-dev_manhattan_f1": 0.5162337662337663, + "eval_allNLI-dev_manhattan_f1_threshold": 291.796142578125, + "eval_allNLI-dev_manhattan_precision": 0.35891647855530473, + "eval_allNLI-dev_manhattan_recall": 0.9190751445086706, + "eval_allNLI-dev_max_accuracy": 0.673828125, + "eval_allNLI-dev_max_accuracy_threshold": 733.9632568359375, + "eval_allNLI-dev_max_ap": 0.427219753620832, + "eval_allNLI-dev_max_f1": 0.5162337662337663, + "eval_allNLI-dev_max_f1_threshold": 656.796875, + "eval_allNLI-dev_max_precision": 0.35891647855530473, + "eval_allNLI-dev_max_recall": 0.9190751445086706, + "eval_loss": 1.5785757303237915, + "eval_runtime": 50.3493, + "eval_samples_per_second": 33.049, + "eval_sequential_score": 0.6159146231463524, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.23228063496639384, + "eval_sts-test_pearson_dot": 0.2315348787904144, + "eval_sts-test_pearson_euclidean": 0.2640388209260748, + "eval_sts-test_pearson_manhattan": 0.2689319416635326, + "eval_sts-test_pearson_max": 0.2689319416635326, + "eval_sts-test_spearman_cosine": 0.2924440270022472, + "eval_sts-test_spearman_dot": 0.29185572693314543, + "eval_sts-test_spearman_euclidean": 0.29247741144249895, + "eval_sts-test_spearman_manhattan": 0.29675903303168044, + "eval_sts-test_spearman_max": 0.29675903303168044, + "step": 306 + }, + { + "epoch": 0.3021653543307087, + "grad_norm": 16.417388916015625, + "learning_rate": 3.001988071570577e-07, + "loss": 2.5365, + "step": 307 + }, + { + "epoch": 0.3031496062992126, + "grad_norm": 13.291057586669922, + "learning_rate": 3.01192842942346e-07, + "loss": 2.0346, + "step": 308 + }, + { + "epoch": 0.30413385826771655, + "grad_norm": 12.769084930419922, + "learning_rate": 3.021868787276342e-07, + "loss": 2.0721, + "step": 309 + }, + { + "epoch": 0.3051181102362205, + "grad_norm": 14.768454551696777, + "learning_rate": 3.031809145129225e-07, + "loss": 2.6966, + "step": 310 + }, + { + "epoch": 0.3061023622047244, + "grad_norm": 13.93484115600586, + "learning_rate": 3.0417495029821077e-07, + "loss": 2.6757, + "step": 311 + }, + { + "epoch": 0.30708661417322836, + "grad_norm": 79.92414093017578, + "learning_rate": 3.051689860834991e-07, + "loss": 10.6395, + "step": 312 + }, + { + "epoch": 0.3080708661417323, + "grad_norm": 15.812857627868652, + "learning_rate": 3.0616302186878733e-07, + "loss": 2.8671, + "step": 313 + }, + { + "epoch": 0.3090551181102362, + "grad_norm": 15.051225662231445, + "learning_rate": 3.071570576540756e-07, + "loss": 2.0144, + "step": 314 + }, + { + "epoch": 0.31003937007874016, + "grad_norm": 66.15606689453125, + "learning_rate": 3.0815109343936384e-07, + "loss": 9.9338, + "step": 315 + }, + { + "epoch": 0.3110236220472441, + "grad_norm": 13.650707244873047, + "learning_rate": 3.0914512922465214e-07, + "loss": 2.6167, + "step": 316 + }, + { + "epoch": 0.31200787401574803, + "grad_norm": 14.176420211791992, + "learning_rate": 3.1013916500994045e-07, + "loss": 2.1342, + "step": 317 + }, + { + "epoch": 0.31299212598425197, + "grad_norm": 66.7360610961914, + "learning_rate": 3.1113320079522865e-07, + "loss": 9.0369, + "step": 318 + }, + { + "epoch": 0.3139763779527559, + "grad_norm": 11.438474655151367, + "learning_rate": 3.1212723658051695e-07, + "loss": 2.0182, + "step": 319 + }, + { + "epoch": 0.31496062992125984, + "grad_norm": 12.611298561096191, + "learning_rate": 3.131212723658052e-07, + "loss": 2.2189, + "step": 320 + }, + { + "epoch": 0.3159448818897638, + "grad_norm": 13.037810325622559, + "learning_rate": 3.141153081510935e-07, + "loss": 1.9667, + "step": 321 + }, + { + "epoch": 0.3169291338582677, + "grad_norm": 13.01930046081543, + "learning_rate": 3.151093439363818e-07, + "loss": 2.3371, + "step": 322 + }, + { + "epoch": 0.31791338582677164, + "grad_norm": 41.48295974731445, + "learning_rate": 3.1610337972167e-07, + "loss": 6.9866, + "step": 323 + }, + { + "epoch": 0.3188976377952756, + "grad_norm": 10.921730995178223, + "learning_rate": 3.170974155069583e-07, + "loss": 1.6119, + "step": 324 + }, + { + "epoch": 0.3198818897637795, + "grad_norm": 13.068329811096191, + "learning_rate": 3.1809145129224657e-07, + "loss": 1.8615, + "step": 325 + }, + { + "epoch": 0.32086614173228345, + "grad_norm": 11.956509590148926, + "learning_rate": 3.190854870775349e-07, + "loss": 2.1708, + "step": 326 + }, + { + "epoch": 0.3218503937007874, + "grad_norm": 11.630243301391602, + "learning_rate": 3.200795228628231e-07, + "loss": 2.0174, + "step": 327 + }, + { + "epoch": 0.3228346456692913, + "grad_norm": 55.17266082763672, + "learning_rate": 3.210735586481114e-07, + "loss": 6.7891, + "step": 328 + }, + { + "epoch": 0.32381889763779526, + "grad_norm": 12.17352294921875, + "learning_rate": 3.220675944333997e-07, + "loss": 2.155, + "step": 329 + }, + { + "epoch": 0.3248031496062992, + "grad_norm": 12.89798641204834, + "learning_rate": 3.2306163021868794e-07, + "loss": 2.4636, + "step": 330 + }, + { + "epoch": 0.3257874015748031, + "grad_norm": 12.991835594177246, + "learning_rate": 3.240556660039762e-07, + "loss": 1.9844, + "step": 331 + }, + { + "epoch": 0.32677165354330706, + "grad_norm": 14.8333101272583, + "learning_rate": 3.2504970178926444e-07, + "loss": 1.9035, + "step": 332 + }, + { + "epoch": 0.327755905511811, + "grad_norm": 10.349666595458984, + "learning_rate": 3.2604373757455275e-07, + "loss": 2.0729, + "step": 333 + }, + { + "epoch": 0.328740157480315, + "grad_norm": 12.129368782043457, + "learning_rate": 3.27037773359841e-07, + "loss": 1.5715, + "step": 334 + }, + { + "epoch": 0.3297244094488189, + "grad_norm": 15.0595703125, + "learning_rate": 3.280318091451293e-07, + "loss": 2.7211, + "step": 335 + }, + { + "epoch": 0.33070866141732286, + "grad_norm": 12.674631118774414, + "learning_rate": 3.2902584493041756e-07, + "loss": 2.0351, + "step": 336 + }, + { + "epoch": 0.3316929133858268, + "grad_norm": 12.62523078918457, + "learning_rate": 3.300198807157058e-07, + "loss": 2.4049, + "step": 337 + }, + { + "epoch": 0.33267716535433073, + "grad_norm": 13.49946117401123, + "learning_rate": 3.310139165009941e-07, + "loss": 2.3939, + "step": 338 + }, + { + "epoch": 0.33366141732283466, + "grad_norm": 11.902497291564941, + "learning_rate": 3.3200795228628237e-07, + "loss": 1.7353, + "step": 339 + }, + { + "epoch": 0.3346456692913386, + "grad_norm": 13.154413223266602, + "learning_rate": 3.330019880715706e-07, + "loss": 1.8393, + "step": 340 + }, + { + "epoch": 0.33562992125984253, + "grad_norm": 14.124127388000488, + "learning_rate": 3.3399602385685887e-07, + "loss": 2.2874, + "step": 341 + }, + { + "epoch": 0.33661417322834647, + "grad_norm": 12.059540748596191, + "learning_rate": 3.349900596421472e-07, + "loss": 1.8566, + "step": 342 + }, + { + "epoch": 0.3375984251968504, + "grad_norm": 13.709589004516602, + "learning_rate": 3.359840954274355e-07, + "loss": 2.2676, + "step": 343 + }, + { + "epoch": 0.33858267716535434, + "grad_norm": 12.062017440795898, + "learning_rate": 3.369781312127237e-07, + "loss": 1.7895, + "step": 344 + }, + { + "epoch": 0.3395669291338583, + "grad_norm": 13.544132232666016, + "learning_rate": 3.37972166998012e-07, + "loss": 2.2506, + "step": 345 + }, + { + "epoch": 0.3405511811023622, + "grad_norm": 11.094194412231445, + "learning_rate": 3.3896620278330024e-07, + "loss": 1.5613, + "step": 346 + }, + { + "epoch": 0.34153543307086615, + "grad_norm": 14.6980619430542, + "learning_rate": 3.3996023856858855e-07, + "loss": 2.3531, + "step": 347 + }, + { + "epoch": 0.3425196850393701, + "grad_norm": 13.778131484985352, + "learning_rate": 3.409542743538768e-07, + "loss": 1.99, + "step": 348 + }, + { + "epoch": 0.343503937007874, + "grad_norm": 80.99553680419922, + "learning_rate": 3.4194831013916505e-07, + "loss": 12.0831, + "step": 349 + }, + { + "epoch": 0.34448818897637795, + "grad_norm": 13.358895301818848, + "learning_rate": 3.4294234592445336e-07, + "loss": 2.0959, + "step": 350 + }, + { + "epoch": 0.3454724409448819, + "grad_norm": 13.7051362991333, + "learning_rate": 3.439363817097416e-07, + "loss": 2.0641, + "step": 351 + }, + { + "epoch": 0.3464566929133858, + "grad_norm": 12.88119125366211, + "learning_rate": 3.449304174950299e-07, + "loss": 1.9197, + "step": 352 + }, + { + "epoch": 0.34744094488188976, + "grad_norm": 12.675789833068848, + "learning_rate": 3.459244532803181e-07, + "loss": 1.9382, + "step": 353 + }, + { + "epoch": 0.3484251968503937, + "grad_norm": 13.377501487731934, + "learning_rate": 3.469184890656064e-07, + "loss": 2.3819, + "step": 354 + }, + { + "epoch": 0.3494094488188976, + "grad_norm": 10.680826187133789, + "learning_rate": 3.4791252485089467e-07, + "loss": 1.6053, + "step": 355 + }, + { + "epoch": 0.35039370078740156, + "grad_norm": 14.856592178344727, + "learning_rate": 3.48906560636183e-07, + "loss": 2.4719, + "step": 356 + }, + { + "epoch": 0.3513779527559055, + "grad_norm": 12.028703689575195, + "learning_rate": 3.499005964214713e-07, + "loss": 1.5602, + "step": 357 + }, + { + "epoch": 0.35236220472440943, + "grad_norm": 16.001672744750977, + "learning_rate": 3.508946322067595e-07, + "loss": 2.1675, + "step": 358 + }, + { + "epoch": 0.35334645669291337, + "grad_norm": 68.59622192382812, + "learning_rate": 3.518886679920478e-07, + "loss": 11.5856, + "step": 359 + }, + { + "epoch": 0.3543307086614173, + "grad_norm": 88.02190399169922, + "learning_rate": 3.5288270377733604e-07, + "loss": 9.3718, + "step": 360 + }, + { + "epoch": 0.35531496062992124, + "grad_norm": 13.347969055175781, + "learning_rate": 3.5387673956262434e-07, + "loss": 1.8952, + "step": 361 + }, + { + "epoch": 0.3562992125984252, + "grad_norm": 13.802129745483398, + "learning_rate": 3.5487077534791254e-07, + "loss": 1.701, + "step": 362 + }, + { + "epoch": 0.3572834645669291, + "grad_norm": 10.31098747253418, + "learning_rate": 3.5586481113320085e-07, + "loss": 1.46, + "step": 363 + }, + { + "epoch": 0.35826771653543305, + "grad_norm": 12.081972122192383, + "learning_rate": 3.5685884691848915e-07, + "loss": 1.7913, + "step": 364 + }, + { + "epoch": 0.359251968503937, + "grad_norm": 51.74912643432617, + "learning_rate": 3.578528827037774e-07, + "loss": 9.1152, + "step": 365 + }, + { + "epoch": 0.36023622047244097, + "grad_norm": 64.62492370605469, + "learning_rate": 3.5884691848906566e-07, + "loss": 9.2681, + "step": 366 + }, + { + "epoch": 0.3612204724409449, + "grad_norm": 13.49717903137207, + "learning_rate": 3.598409542743539e-07, + "loss": 2.2932, + "step": 367 + }, + { + "epoch": 0.36220472440944884, + "grad_norm": 10.775108337402344, + "learning_rate": 3.608349900596422e-07, + "loss": 1.7176, + "step": 368 + }, + { + "epoch": 0.3631889763779528, + "grad_norm": 12.510822296142578, + "learning_rate": 3.6182902584493047e-07, + "loss": 2.2559, + "step": 369 + }, + { + "epoch": 0.3641732283464567, + "grad_norm": 11.558751106262207, + "learning_rate": 3.6282306163021877e-07, + "loss": 1.9846, + "step": 370 + }, + { + "epoch": 0.36515748031496065, + "grad_norm": 12.446829795837402, + "learning_rate": 3.63817097415507e-07, + "loss": 1.8022, + "step": 371 + }, + { + "epoch": 0.3661417322834646, + "grad_norm": 59.3734130859375, + "learning_rate": 3.648111332007953e-07, + "loss": 8.1128, + "step": 372 + }, + { + "epoch": 0.3671259842519685, + "grad_norm": 52.91653060913086, + "learning_rate": 3.658051689860836e-07, + "loss": 6.929, + "step": 373 + }, + { + "epoch": 0.36811023622047245, + "grad_norm": 11.077016830444336, + "learning_rate": 3.6679920477137184e-07, + "loss": 1.9038, + "step": 374 + }, + { + "epoch": 0.3690944881889764, + "grad_norm": 11.368824005126953, + "learning_rate": 3.677932405566601e-07, + "loss": 1.3899, + "step": 375 + }, + { + "epoch": 0.3700787401574803, + "grad_norm": 9.926651000976562, + "learning_rate": 3.6878727634194834e-07, + "loss": 1.5677, + "step": 376 + }, + { + "epoch": 0.37106299212598426, + "grad_norm": 34.99372482299805, + "learning_rate": 3.6978131212723665e-07, + "loss": 5.2357, + "step": 377 + }, + { + "epoch": 0.3720472440944882, + "grad_norm": 10.987953186035156, + "learning_rate": 3.7077534791252495e-07, + "loss": 2.2304, + "step": 378 + }, + { + "epoch": 0.37303149606299213, + "grad_norm": 10.966182708740234, + "learning_rate": 3.7176938369781315e-07, + "loss": 2.1727, + "step": 379 + }, + { + "epoch": 0.37401574803149606, + "grad_norm": 11.722915649414062, + "learning_rate": 3.7276341948310145e-07, + "loss": 2.2941, + "step": 380 + }, + { + "epoch": 0.375, + "grad_norm": 12.356597900390625, + "learning_rate": 3.737574552683897e-07, + "loss": 2.2257, + "step": 381 + }, + { + "epoch": 0.37598425196850394, + "grad_norm": 12.089561462402344, + "learning_rate": 3.74751491053678e-07, + "loss": 1.7489, + "step": 382 + }, + { + "epoch": 0.37696850393700787, + "grad_norm": 10.173055648803711, + "learning_rate": 3.7574552683896626e-07, + "loss": 1.5027, + "step": 383 + }, + { + "epoch": 0.3779527559055118, + "grad_norm": 11.658833503723145, + "learning_rate": 3.767395626242545e-07, + "loss": 1.6917, + "step": 384 + }, + { + "epoch": 0.37893700787401574, + "grad_norm": 41.831546783447266, + "learning_rate": 3.777335984095428e-07, + "loss": 5.7867, + "step": 385 + }, + { + "epoch": 0.3799212598425197, + "grad_norm": 10.72303581237793, + "learning_rate": 3.787276341948311e-07, + "loss": 1.6871, + "step": 386 + }, + { + "epoch": 0.3809055118110236, + "grad_norm": 10.949180603027344, + "learning_rate": 3.797216699801194e-07, + "loss": 1.5652, + "step": 387 + }, + { + "epoch": 0.38188976377952755, + "grad_norm": 12.642234802246094, + "learning_rate": 3.807157057654076e-07, + "loss": 2.1691, + "step": 388 + }, + { + "epoch": 0.3828740157480315, + "grad_norm": 10.835790634155273, + "learning_rate": 3.817097415506959e-07, + "loss": 1.869, + "step": 389 + }, + { + "epoch": 0.3838582677165354, + "grad_norm": 11.778974533081055, + "learning_rate": 3.8270377733598414e-07, + "loss": 2.1934, + "step": 390 + }, + { + "epoch": 0.38484251968503935, + "grad_norm": 65.27779388427734, + "learning_rate": 3.8369781312127244e-07, + "loss": 7.0152, + "step": 391 + }, + { + "epoch": 0.3858267716535433, + "grad_norm": 12.2304048538208, + "learning_rate": 3.8469184890656075e-07, + "loss": 2.0454, + "step": 392 + }, + { + "epoch": 0.3868110236220472, + "grad_norm": 12.670406341552734, + "learning_rate": 3.8568588469184895e-07, + "loss": 1.8098, + "step": 393 + }, + { + "epoch": 0.38779527559055116, + "grad_norm": 44.1097526550293, + "learning_rate": 3.8667992047713725e-07, + "loss": 5.7529, + "step": 394 + }, + { + "epoch": 0.3887795275590551, + "grad_norm": 12.897631645202637, + "learning_rate": 3.876739562624255e-07, + "loss": 1.3949, + "step": 395 + }, + { + "epoch": 0.38976377952755903, + "grad_norm": 11.003929138183594, + "learning_rate": 3.886679920477138e-07, + "loss": 1.5962, + "step": 396 + }, + { + "epoch": 0.390748031496063, + "grad_norm": 41.88210678100586, + "learning_rate": 3.89662027833002e-07, + "loss": 6.1436, + "step": 397 + }, + { + "epoch": 0.39173228346456695, + "grad_norm": 35.17912292480469, + "learning_rate": 3.906560636182903e-07, + "loss": 5.2979, + "step": 398 + }, + { + "epoch": 0.3927165354330709, + "grad_norm": 10.979583740234375, + "learning_rate": 3.916500994035786e-07, + "loss": 1.2422, + "step": 399 + }, + { + "epoch": 0.3937007874015748, + "grad_norm": 11.90557861328125, + "learning_rate": 3.9264413518886687e-07, + "loss": 2.1152, + "step": 400 + }, + { + "epoch": 0.39468503937007876, + "grad_norm": 11.298015594482422, + "learning_rate": 3.936381709741551e-07, + "loss": 1.6679, + "step": 401 + }, + { + "epoch": 0.3956692913385827, + "grad_norm": 30.55457305908203, + "learning_rate": 3.946322067594434e-07, + "loss": 4.2978, + "step": 402 + }, + { + "epoch": 0.39665354330708663, + "grad_norm": 11.460474967956543, + "learning_rate": 3.956262425447317e-07, + "loss": 1.624, + "step": 403 + }, + { + "epoch": 0.39763779527559057, + "grad_norm": 11.218096733093262, + "learning_rate": 3.9662027833001993e-07, + "loss": 2.0267, + "step": 404 + }, + { + "epoch": 0.3986220472440945, + "grad_norm": 10.295607566833496, + "learning_rate": 3.9761431411530824e-07, + "loss": 1.3975, + "step": 405 + }, + { + "epoch": 0.39960629921259844, + "grad_norm": 11.429953575134277, + "learning_rate": 3.986083499005965e-07, + "loss": 1.905, + "step": 406 + }, + { + "epoch": 0.4005905511811024, + "grad_norm": 38.397621154785156, + "learning_rate": 3.9960238568588474e-07, + "loss": 5.4419, + "step": 407 + }, + { + "epoch": 0.4015748031496063, + "grad_norm": 13.075526237487793, + "learning_rate": 4.0059642147117305e-07, + "loss": 2.0008, + "step": 408 + }, + { + "epoch": 0.40255905511811024, + "grad_norm": 11.999564170837402, + "learning_rate": 4.015904572564613e-07, + "loss": 1.8387, + "step": 409 + }, + { + "epoch": 0.4035433070866142, + "grad_norm": 12.46625804901123, + "learning_rate": 4.0258449304174955e-07, + "loss": 2.2391, + "step": 410 + }, + { + "epoch": 0.4045275590551181, + "grad_norm": 13.231268882751465, + "learning_rate": 4.035785288270378e-07, + "loss": 1.7153, + "step": 411 + }, + { + "epoch": 0.40551181102362205, + "grad_norm": 11.404708862304688, + "learning_rate": 4.045725646123261e-07, + "loss": 2.1533, + "step": 412 + }, + { + "epoch": 0.406496062992126, + "grad_norm": 11.410626411437988, + "learning_rate": 4.055666003976144e-07, + "loss": 1.788, + "step": 413 + }, + { + "epoch": 0.4074803149606299, + "grad_norm": 27.646928787231445, + "learning_rate": 4.065606361829026e-07, + "loss": 3.482, + "step": 414 + }, + { + "epoch": 0.40846456692913385, + "grad_norm": 12.279366493225098, + "learning_rate": 4.075546719681909e-07, + "loss": 1.8376, + "step": 415 + }, + { + "epoch": 0.4094488188976378, + "grad_norm": 46.18857955932617, + "learning_rate": 4.085487077534792e-07, + "loss": 4.8811, + "step": 416 + }, + { + "epoch": 0.4104330708661417, + "grad_norm": 12.051934242248535, + "learning_rate": 4.095427435387675e-07, + "loss": 1.9421, + "step": 417 + }, + { + "epoch": 0.41141732283464566, + "grad_norm": 11.133675575256348, + "learning_rate": 4.1053677932405573e-07, + "loss": 1.4796, + "step": 418 + }, + { + "epoch": 0.4124015748031496, + "grad_norm": 10.847027778625488, + "learning_rate": 4.11530815109344e-07, + "loss": 1.6209, + "step": 419 + }, + { + "epoch": 0.41338582677165353, + "grad_norm": 12.144013404846191, + "learning_rate": 4.125248508946323e-07, + "loss": 1.8734, + "step": 420 + }, + { + "epoch": 0.41437007874015747, + "grad_norm": 13.027422904968262, + "learning_rate": 4.1351888667992054e-07, + "loss": 1.9444, + "step": 421 + }, + { + "epoch": 0.4153543307086614, + "grad_norm": 13.315413475036621, + "learning_rate": 4.1451292246520885e-07, + "loss": 1.9581, + "step": 422 + }, + { + "epoch": 0.41633858267716534, + "grad_norm": 11.68213176727295, + "learning_rate": 4.1550695825049705e-07, + "loss": 1.5175, + "step": 423 + }, + { + "epoch": 0.41732283464566927, + "grad_norm": 11.046775817871094, + "learning_rate": 4.1650099403578535e-07, + "loss": 1.2831, + "step": 424 + }, + { + "epoch": 0.4183070866141732, + "grad_norm": 9.949254035949707, + "learning_rate": 4.174950298210736e-07, + "loss": 1.1355, + "step": 425 + }, + { + "epoch": 0.41929133858267714, + "grad_norm": 13.619227409362793, + "learning_rate": 4.184890656063619e-07, + "loss": 1.864, + "step": 426 + }, + { + "epoch": 0.4202755905511811, + "grad_norm": 49.670379638671875, + "learning_rate": 4.194831013916502e-07, + "loss": 5.1574, + "step": 427 + }, + { + "epoch": 0.421259842519685, + "grad_norm": 41.61232376098633, + "learning_rate": 4.204771371769384e-07, + "loss": 5.323, + "step": 428 + }, + { + "epoch": 0.422244094488189, + "grad_norm": 13.148953437805176, + "learning_rate": 4.214711729622267e-07, + "loss": 1.385, + "step": 429 + }, + { + "epoch": 0.42322834645669294, + "grad_norm": 10.928688049316406, + "learning_rate": 4.2246520874751497e-07, + "loss": 1.1691, + "step": 430 + }, + { + "epoch": 0.4242125984251969, + "grad_norm": 13.726749420166016, + "learning_rate": 4.234592445328033e-07, + "loss": 1.8994, + "step": 431 + }, + { + "epoch": 0.4251968503937008, + "grad_norm": 55.364986419677734, + "learning_rate": 4.244532803180915e-07, + "loss": 5.4254, + "step": 432 + }, + { + "epoch": 0.42618110236220474, + "grad_norm": 12.936726570129395, + "learning_rate": 4.254473161033798e-07, + "loss": 1.9113, + "step": 433 + }, + { + "epoch": 0.4271653543307087, + "grad_norm": 14.45608139038086, + "learning_rate": 4.264413518886681e-07, + "loss": 2.1108, + "step": 434 + }, + { + "epoch": 0.4281496062992126, + "grad_norm": 12.692010879516602, + "learning_rate": 4.2743538767395634e-07, + "loss": 1.7012, + "step": 435 + }, + { + "epoch": 0.42913385826771655, + "grad_norm": 12.209145545959473, + "learning_rate": 4.284294234592446e-07, + "loss": 1.5722, + "step": 436 + }, + { + "epoch": 0.4301181102362205, + "grad_norm": 11.499476432800293, + "learning_rate": 4.2942345924453284e-07, + "loss": 1.5967, + "step": 437 + }, + { + "epoch": 0.4311023622047244, + "grad_norm": 48.46207046508789, + "learning_rate": 4.3041749502982115e-07, + "loss": 5.609, + "step": 438 + }, + { + "epoch": 0.43208661417322836, + "grad_norm": 11.709005355834961, + "learning_rate": 4.3141153081510945e-07, + "loss": 1.4444, + "step": 439 + }, + { + "epoch": 0.4330708661417323, + "grad_norm": 45.81867218017578, + "learning_rate": 4.324055666003977e-07, + "loss": 5.3153, + "step": 440 + }, + { + "epoch": 0.4340551181102362, + "grad_norm": 41.934791564941406, + "learning_rate": 4.3339960238568596e-07, + "loss": 5.0934, + "step": 441 + }, + { + "epoch": 0.43503937007874016, + "grad_norm": 10.987709045410156, + "learning_rate": 4.343936381709742e-07, + "loss": 1.3028, + "step": 442 + }, + { + "epoch": 0.4360236220472441, + "grad_norm": 10.71814250946045, + "learning_rate": 4.353876739562625e-07, + "loss": 1.263, + "step": 443 + }, + { + "epoch": 0.43700787401574803, + "grad_norm": 12.126359939575195, + "learning_rate": 4.3638170974155077e-07, + "loss": 1.8462, + "step": 444 + }, + { + "epoch": 0.43799212598425197, + "grad_norm": 15.644315719604492, + "learning_rate": 4.37375745526839e-07, + "loss": 2.1533, + "step": 445 + }, + { + "epoch": 0.4389763779527559, + "grad_norm": 12.823408126831055, + "learning_rate": 4.383697813121273e-07, + "loss": 1.5467, + "step": 446 + }, + { + "epoch": 0.43996062992125984, + "grad_norm": 12.075353622436523, + "learning_rate": 4.393638170974156e-07, + "loss": 1.4331, + "step": 447 + }, + { + "epoch": 0.4409448818897638, + "grad_norm": 10.601676940917969, + "learning_rate": 4.403578528827039e-07, + "loss": 1.4416, + "step": 448 + }, + { + "epoch": 0.4419291338582677, + "grad_norm": 12.03585433959961, + "learning_rate": 4.413518886679921e-07, + "loss": 1.5976, + "step": 449 + }, + { + "epoch": 0.44291338582677164, + "grad_norm": 12.471179962158203, + "learning_rate": 4.423459244532804e-07, + "loss": 1.8723, + "step": 450 + }, + { + "epoch": 0.4438976377952756, + "grad_norm": 11.043651580810547, + "learning_rate": 4.4333996023856864e-07, + "loss": 1.1753, + "step": 451 + }, + { + "epoch": 0.4448818897637795, + "grad_norm": 15.544292449951172, + "learning_rate": 4.4433399602385695e-07, + "loss": 2.3205, + "step": 452 + }, + { + "epoch": 0.44586614173228345, + "grad_norm": 11.868247032165527, + "learning_rate": 4.4532803180914525e-07, + "loss": 1.6467, + "step": 453 + }, + { + "epoch": 0.4468503937007874, + "grad_norm": 10.128686904907227, + "learning_rate": 4.4632206759443345e-07, + "loss": 0.9322, + "step": 454 + }, + { + "epoch": 0.4478346456692913, + "grad_norm": 14.427766799926758, + "learning_rate": 4.4731610337972176e-07, + "loss": 1.958, + "step": 455 + }, + { + "epoch": 0.44881889763779526, + "grad_norm": 14.051414489746094, + "learning_rate": 4.4831013916501e-07, + "loss": 1.8746, + "step": 456 + }, + { + "epoch": 0.4498031496062992, + "grad_norm": 11.834802627563477, + "learning_rate": 4.493041749502983e-07, + "loss": 1.4546, + "step": 457 + }, + { + "epoch": 0.4507874015748031, + "grad_norm": 11.809011459350586, + "learning_rate": 4.502982107355865e-07, + "loss": 0.9795, + "step": 458 + }, + { + "epoch": 0.45177165354330706, + "grad_norm": 13.02786922454834, + "learning_rate": 4.512922465208748e-07, + "loss": 1.5458, + "step": 459 + }, + { + "epoch": 0.45177165354330706, + "eval_Qnli-dev_cosine_accuracy": 0.640625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8912460803985596, + "eval_Qnli-dev_cosine_ap": 0.6357698436423456, + "eval_Qnli-dev_cosine_f1": 0.649645390070922, + "eval_Qnli-dev_cosine_f1_threshold": 0.770536482334137, + "eval_Qnli-dev_cosine_precision": 0.488272921108742, + "eval_Qnli-dev_cosine_recall": 0.9703389830508474, + "eval_Qnli-dev_dot_accuracy": 0.640625, + "eval_Qnli-dev_dot_accuracy_threshold": 680.3779296875, + "eval_Qnli-dev_dot_ap": 0.6343758351328452, + "eval_Qnli-dev_dot_f1": 0.6487252124645891, + "eval_Qnli-dev_dot_f1_threshold": 590.0875244140625, + "eval_Qnli-dev_dot_precision": 0.48723404255319147, + "eval_Qnli-dev_dot_recall": 0.9703389830508474, + "eval_Qnli-dev_euclidean_accuracy": 0.640625, + "eval_Qnli-dev_euclidean_accuracy_threshold": 12.908914566040039, + "eval_Qnli-dev_euclidean_ap": 0.6357656835719167, + "eval_Qnli-dev_euclidean_f1": 0.649645390070922, + "eval_Qnli-dev_euclidean_f1_threshold": 18.747791290283203, + "eval_Qnli-dev_euclidean_precision": 0.488272921108742, + "eval_Qnli-dev_euclidean_recall": 0.9703389830508474, + "eval_Qnli-dev_manhattan_accuracy": 0.638671875, + "eval_Qnli-dev_manhattan_accuracy_threshold": 260.3154602050781, + "eval_Qnli-dev_manhattan_ap": 0.643325439884054, + "eval_Qnli-dev_manhattan_f1": 0.6551226551226551, + "eval_Qnli-dev_manhattan_f1_threshold": 370.62738037109375, + "eval_Qnli-dev_manhattan_precision": 0.49671772428884026, + "eval_Qnli-dev_manhattan_recall": 0.961864406779661, + "eval_Qnli-dev_max_accuracy": 0.640625, + "eval_Qnli-dev_max_accuracy_threshold": 680.3779296875, + "eval_Qnli-dev_max_ap": 0.643325439884054, + "eval_Qnli-dev_max_f1": 0.6551226551226551, + "eval_Qnli-dev_max_f1_threshold": 590.0875244140625, + "eval_Qnli-dev_max_precision": 0.49671772428884026, + "eval_Qnli-dev_max_recall": 0.9703389830508474, + "eval_allNLI-dev_cosine_accuracy": 0.677734375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9714465141296387, + "eval_allNLI-dev_cosine_ap": 0.44846088326983674, + "eval_allNLI-dev_cosine_f1": 0.5163636363636364, + "eval_allNLI-dev_cosine_f1_threshold": 0.8693957924842834, + "eval_allNLI-dev_cosine_precision": 0.376657824933687, + "eval_allNLI-dev_cosine_recall": 0.8208092485549133, + "eval_allNLI-dev_dot_accuracy": 0.6796875, + "eval_allNLI-dev_dot_accuracy_threshold": 745.1615600585938, + "eval_allNLI-dev_dot_ap": 0.4483320077591566, + "eval_allNLI-dev_dot_f1": 0.5182481751824818, + "eval_allNLI-dev_dot_f1_threshold": 667.3764038085938, + "eval_allNLI-dev_dot_precision": 0.37866666666666665, + "eval_allNLI-dev_dot_recall": 0.8208092485549133, + "eval_allNLI-dev_euclidean_accuracy": 0.677734375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 6.618439674377441, + "eval_allNLI-dev_euclidean_ap": 0.4483085913384058, + "eval_allNLI-dev_euclidean_f1": 0.5163636363636364, + "eval_allNLI-dev_euclidean_f1_threshold": 14.158239364624023, + "eval_allNLI-dev_euclidean_precision": 0.376657824933687, + "eval_allNLI-dev_euclidean_recall": 0.8208092485549133, + "eval_allNLI-dev_manhattan_accuracy": 0.673828125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 126.3739013671875, + "eval_allNLI-dev_manhattan_ap": 0.43835460721217534, + "eval_allNLI-dev_manhattan_f1": 0.5194346289752649, + "eval_allNLI-dev_manhattan_f1_threshold": 286.7830810546875, + "eval_allNLI-dev_manhattan_precision": 0.37404580152671757, + "eval_allNLI-dev_manhattan_recall": 0.8497109826589595, + "eval_allNLI-dev_max_accuracy": 0.6796875, + "eval_allNLI-dev_max_accuracy_threshold": 745.1615600585938, + "eval_allNLI-dev_max_ap": 0.44846088326983674, + "eval_allNLI-dev_max_f1": 0.5194346289752649, + "eval_allNLI-dev_max_f1_threshold": 667.3764038085938, + "eval_allNLI-dev_max_precision": 0.37866666666666665, + "eval_allNLI-dev_max_recall": 0.8497109826589595, + "eval_loss": 1.2676286697387695, + "eval_runtime": 50.0832, + "eval_samples_per_second": 33.225, + "eval_sequential_score": 0.643325439884054, + "eval_steps_per_second": 0.14, + "eval_sts-test_pearson_cosine": 0.22108397313035918, + "eval_sts-test_pearson_dot": 0.21871092924361943, + "eval_sts-test_pearson_euclidean": 0.24847225726797056, + "eval_sts-test_pearson_manhattan": 0.2651658402517003, + "eval_sts-test_pearson_max": 0.2651658402517003, + "eval_sts-test_spearman_cosine": 0.27510018992279384, + "eval_sts-test_spearman_dot": 0.2727821193339807, + "eval_sts-test_spearman_euclidean": 0.27518333776645915, + "eval_sts-test_spearman_manhattan": 0.2935613830333265, + "eval_sts-test_spearman_max": 0.2935613830333265, + "step": 459 + }, + { + "epoch": 0.452755905511811, + "grad_norm": 13.902926445007324, + "learning_rate": 4.522862823061631e-07, + "loss": 1.6558, + "step": 460 + }, + { + "epoch": 0.453740157480315, + "grad_norm": 11.984386444091797, + "learning_rate": 4.532803180914514e-07, + "loss": 1.389, + "step": 461 + }, + { + "epoch": 0.4547244094488189, + "grad_norm": 13.723044395446777, + "learning_rate": 4.542743538767397e-07, + "loss": 1.5608, + "step": 462 + }, + { + "epoch": 0.45570866141732286, + "grad_norm": 12.504467010498047, + "learning_rate": 4.552683896620279e-07, + "loss": 1.6618, + "step": 463 + }, + { + "epoch": 0.4566929133858268, + "grad_norm": 12.590119361877441, + "learning_rate": 4.562624254473162e-07, + "loss": 1.5122, + "step": 464 + }, + { + "epoch": 0.45767716535433073, + "grad_norm": 13.59104061126709, + "learning_rate": 4.5725646123260444e-07, + "loss": 1.3602, + "step": 465 + }, + { + "epoch": 0.45866141732283466, + "grad_norm": 13.90998649597168, + "learning_rate": 4.5825049701789274e-07, + "loss": 1.6714, + "step": 466 + }, + { + "epoch": 0.4596456692913386, + "grad_norm": 11.928483009338379, + "learning_rate": 4.59244532803181e-07, + "loss": 1.0644, + "step": 467 + }, + { + "epoch": 0.46062992125984253, + "grad_norm": 53.926265716552734, + "learning_rate": 4.6023856858846925e-07, + "loss": 7.6421, + "step": 468 + }, + { + "epoch": 0.46161417322834647, + "grad_norm": 11.6729736328125, + "learning_rate": 4.6123260437375755e-07, + "loss": 1.2987, + "step": 469 + }, + { + "epoch": 0.4625984251968504, + "grad_norm": 14.65312671661377, + "learning_rate": 4.622266401590458e-07, + "loss": 1.4231, + "step": 470 + }, + { + "epoch": 0.46358267716535434, + "grad_norm": 53.9046516418457, + "learning_rate": 4.6322067594433406e-07, + "loss": 7.7424, + "step": 471 + }, + { + "epoch": 0.4645669291338583, + "grad_norm": 13.964241981506348, + "learning_rate": 4.642147117296223e-07, + "loss": 1.6811, + "step": 472 + }, + { + "epoch": 0.4655511811023622, + "grad_norm": 11.857098579406738, + "learning_rate": 4.652087475149106e-07, + "loss": 1.1814, + "step": 473 + }, + { + "epoch": 0.46653543307086615, + "grad_norm": 12.89073371887207, + "learning_rate": 4.662027833001989e-07, + "loss": 1.4486, + "step": 474 + }, + { + "epoch": 0.4675196850393701, + "grad_norm": 12.869789123535156, + "learning_rate": 4.671968190854872e-07, + "loss": 1.3892, + "step": 475 + }, + { + "epoch": 0.468503937007874, + "grad_norm": 11.521900177001953, + "learning_rate": 4.681908548707754e-07, + "loss": 1.3681, + "step": 476 + }, + { + "epoch": 0.46948818897637795, + "grad_norm": 12.49998664855957, + "learning_rate": 4.691848906560637e-07, + "loss": 1.3081, + "step": 477 + }, + { + "epoch": 0.4704724409448819, + "grad_norm": 9.533255577087402, + "learning_rate": 4.70178926441352e-07, + "loss": 0.9102, + "step": 478 + }, + { + "epoch": 0.4714566929133858, + "grad_norm": 10.81651496887207, + "learning_rate": 4.7117296222664024e-07, + "loss": 1.0992, + "step": 479 + }, + { + "epoch": 0.47244094488188976, + "grad_norm": 43.7128791809082, + "learning_rate": 4.721669980119285e-07, + "loss": 6.018, + "step": 480 + }, + { + "epoch": 0.4734251968503937, + "grad_norm": 50.324893951416016, + "learning_rate": 4.731610337972168e-07, + "loss": 6.0908, + "step": 481 + }, + { + "epoch": 0.4744094488188976, + "grad_norm": 10.95649528503418, + "learning_rate": 4.7415506958250505e-07, + "loss": 1.2245, + "step": 482 + }, + { + "epoch": 0.47539370078740156, + "grad_norm": 12.223633766174316, + "learning_rate": 4.7514910536779335e-07, + "loss": 1.4825, + "step": 483 + }, + { + "epoch": 0.4763779527559055, + "grad_norm": 12.701655387878418, + "learning_rate": 4.7614314115308155e-07, + "loss": 1.8037, + "step": 484 + }, + { + "epoch": 0.47736220472440943, + "grad_norm": 12.231987953186035, + "learning_rate": 4.771371769383698e-07, + "loss": 1.3611, + "step": 485 + }, + { + "epoch": 0.47834645669291337, + "grad_norm": 14.476117134094238, + "learning_rate": 4.781312127236581e-07, + "loss": 1.7482, + "step": 486 + }, + { + "epoch": 0.4793307086614173, + "grad_norm": 13.278433799743652, + "learning_rate": 4.791252485089464e-07, + "loss": 1.6385, + "step": 487 + }, + { + "epoch": 0.48031496062992124, + "grad_norm": 11.42531681060791, + "learning_rate": 4.801192842942347e-07, + "loss": 1.3245, + "step": 488 + }, + { + "epoch": 0.4812992125984252, + "grad_norm": 12.764735221862793, + "learning_rate": 4.811133200795229e-07, + "loss": 1.5638, + "step": 489 + }, + { + "epoch": 0.4822834645669291, + "grad_norm": 12.869999885559082, + "learning_rate": 4.821073558648112e-07, + "loss": 1.566, + "step": 490 + }, + { + "epoch": 0.48326771653543305, + "grad_norm": 14.510449409484863, + "learning_rate": 4.831013916500995e-07, + "loss": 1.9482, + "step": 491 + }, + { + "epoch": 0.484251968503937, + "grad_norm": 57.075016021728516, + "learning_rate": 4.840954274353877e-07, + "loss": 6.0859, + "step": 492 + }, + { + "epoch": 0.48523622047244097, + "grad_norm": 49.321617126464844, + "learning_rate": 4.85089463220676e-07, + "loss": 5.8754, + "step": 493 + }, + { + "epoch": 0.4862204724409449, + "grad_norm": 9.882522583007812, + "learning_rate": 4.860834990059642e-07, + "loss": 0.9964, + "step": 494 + }, + { + "epoch": 0.48720472440944884, + "grad_norm": 12.727804183959961, + "learning_rate": 4.870775347912525e-07, + "loss": 1.5949, + "step": 495 + }, + { + "epoch": 0.4881889763779528, + "grad_norm": 10.453754425048828, + "learning_rate": 4.880715705765408e-07, + "loss": 1.3167, + "step": 496 + }, + { + "epoch": 0.4891732283464567, + "grad_norm": 35.08658981323242, + "learning_rate": 4.890656063618291e-07, + "loss": 3.9345, + "step": 497 + }, + { + "epoch": 0.49015748031496065, + "grad_norm": 39.3719482421875, + "learning_rate": 4.900596421471173e-07, + "loss": 4.3886, + "step": 498 + }, + { + "epoch": 0.4911417322834646, + "grad_norm": 12.961012840270996, + "learning_rate": 4.910536779324057e-07, + "loss": 1.6124, + "step": 499 + }, + { + "epoch": 0.4921259842519685, + "grad_norm": 11.557706832885742, + "learning_rate": 4.92047713717694e-07, + "loss": 1.2145, + "step": 500 + }, + { + "epoch": 0.49311023622047245, + "grad_norm": 35.30653762817383, + "learning_rate": 4.930417495029823e-07, + "loss": 3.5499, + "step": 501 + }, + { + "epoch": 0.4940944881889764, + "grad_norm": 10.364607810974121, + "learning_rate": 4.940357852882705e-07, + "loss": 1.2999, + "step": 502 + }, + { + "epoch": 0.4950787401574803, + "grad_norm": 12.304548263549805, + "learning_rate": 4.950298210735588e-07, + "loss": 1.2375, + "step": 503 + }, + { + "epoch": 0.49606299212598426, + "grad_norm": 10.983871459960938, + "learning_rate": 4.96023856858847e-07, + "loss": 1.1606, + "step": 504 + }, + { + "epoch": 0.4970472440944882, + "grad_norm": 11.577509880065918, + "learning_rate": 4.970178926441353e-07, + "loss": 1.4634, + "step": 505 + }, + { + "epoch": 0.49803149606299213, + "grad_norm": 11.625205993652344, + "learning_rate": 4.980119284294235e-07, + "loss": 1.35, + "step": 506 + }, + { + "epoch": 0.49901574803149606, + "grad_norm": 11.956558227539062, + "learning_rate": 4.990059642147118e-07, + "loss": 1.7187, + "step": 507 + }, + { + "epoch": 0.5, + "grad_norm": 11.75649356842041, + "learning_rate": 5.000000000000001e-07, + "loss": 1.5915, + "step": 508 + }, + { + "epoch": 0.5009842519685039, + "grad_norm": 11.247126579284668, + "learning_rate": 5.009940357852883e-07, + "loss": 1.2357, + "step": 509 + }, + { + "epoch": 0.5019685039370079, + "grad_norm": 36.65633773803711, + "learning_rate": 5.019880715705767e-07, + "loss": 3.4122, + "step": 510 + }, + { + "epoch": 0.5029527559055118, + "grad_norm": 33.088661193847656, + "learning_rate": 5.029821073558649e-07, + "loss": 4.244, + "step": 511 + }, + { + "epoch": 0.5039370078740157, + "grad_norm": 8.999531745910645, + "learning_rate": 5.039761431411532e-07, + "loss": 0.9151, + "step": 512 + }, + { + "epoch": 0.5049212598425197, + "grad_norm": 10.611837387084961, + "learning_rate": 5.049701789264414e-07, + "loss": 1.4323, + "step": 513 + }, + { + "epoch": 0.5059055118110236, + "grad_norm": 11.011503219604492, + "learning_rate": 5.059642147117297e-07, + "loss": 1.4824, + "step": 514 + }, + { + "epoch": 0.5068897637795275, + "grad_norm": 11.72793197631836, + "learning_rate": 5.06958250497018e-07, + "loss": 1.339, + "step": 515 + }, + { + "epoch": 0.5078740157480315, + "grad_norm": 45.36782455444336, + "learning_rate": 5.079522862823062e-07, + "loss": 4.1658, + "step": 516 + }, + { + "epoch": 0.5088582677165354, + "grad_norm": 11.602323532104492, + "learning_rate": 5.089463220675945e-07, + "loss": 1.3062, + "step": 517 + }, + { + "epoch": 0.5098425196850394, + "grad_norm": 10.920750617980957, + "learning_rate": 5.099403578528828e-07, + "loss": 1.2905, + "step": 518 + }, + { + "epoch": 0.5108267716535433, + "grad_norm": 12.382309913635254, + "learning_rate": 5.10934393638171e-07, + "loss": 1.1487, + "step": 519 + }, + { + "epoch": 0.5118110236220472, + "grad_norm": 23.62725067138672, + "learning_rate": 5.119284294234593e-07, + "loss": 2.8652, + "step": 520 + }, + { + "epoch": 0.5127952755905512, + "grad_norm": 10.770047187805176, + "learning_rate": 5.129224652087476e-07, + "loss": 1.2634, + "step": 521 + }, + { + "epoch": 0.5137795275590551, + "grad_norm": 11.524735450744629, + "learning_rate": 5.139165009940358e-07, + "loss": 1.6745, + "step": 522 + }, + { + "epoch": 0.514763779527559, + "grad_norm": 13.119950294494629, + "learning_rate": 5.149105367793241e-07, + "loss": 1.6548, + "step": 523 + }, + { + "epoch": 0.515748031496063, + "grad_norm": 20.820924758911133, + "learning_rate": 5.159045725646124e-07, + "loss": 2.4204, + "step": 524 + }, + { + "epoch": 0.5167322834645669, + "grad_norm": 13.70715045928955, + "learning_rate": 5.168986083499007e-07, + "loss": 1.7201, + "step": 525 + }, + { + "epoch": 0.5177165354330708, + "grad_norm": 12.159992218017578, + "learning_rate": 5.178926441351889e-07, + "loss": 1.761, + "step": 526 + }, + { + "epoch": 0.5187007874015748, + "grad_norm": 22.431594848632812, + "learning_rate": 5.188866799204771e-07, + "loss": 2.7098, + "step": 527 + }, + { + "epoch": 0.5196850393700787, + "grad_norm": 11.434654235839844, + "learning_rate": 5.198807157057656e-07, + "loss": 1.6425, + "step": 528 + }, + { + "epoch": 0.5206692913385826, + "grad_norm": 11.650453567504883, + "learning_rate": 5.208747514910538e-07, + "loss": 1.2466, + "step": 529 + }, + { + "epoch": 0.5216535433070866, + "grad_norm": 11.456934928894043, + "learning_rate": 5.21868787276342e-07, + "loss": 1.3339, + "step": 530 + }, + { + "epoch": 0.5226377952755905, + "grad_norm": 10.777106285095215, + "learning_rate": 5.228628230616304e-07, + "loss": 1.2398, + "step": 531 + }, + { + "epoch": 0.5236220472440944, + "grad_norm": 38.83426284790039, + "learning_rate": 5.238568588469186e-07, + "loss": 3.5325, + "step": 532 + }, + { + "epoch": 0.5246062992125984, + "grad_norm": 9.635952949523926, + "learning_rate": 5.248508946322069e-07, + "loss": 1.1303, + "step": 533 + }, + { + "epoch": 0.5255905511811023, + "grad_norm": 9.629429817199707, + "learning_rate": 5.258449304174951e-07, + "loss": 1.2601, + "step": 534 + }, + { + "epoch": 0.5265748031496063, + "grad_norm": 12.506965637207031, + "learning_rate": 5.268389662027834e-07, + "loss": 1.5762, + "step": 535 + }, + { + "epoch": 0.5275590551181102, + "grad_norm": 11.253185272216797, + "learning_rate": 5.278330019880717e-07, + "loss": 1.3992, + "step": 536 + }, + { + "epoch": 0.5285433070866141, + "grad_norm": 12.544111251831055, + "learning_rate": 5.288270377733599e-07, + "loss": 1.7125, + "step": 537 + }, + { + "epoch": 0.5295275590551181, + "grad_norm": 38.10435104370117, + "learning_rate": 5.298210735586483e-07, + "loss": 3.6759, + "step": 538 + }, + { + "epoch": 0.530511811023622, + "grad_norm": 12.76169490814209, + "learning_rate": 5.308151093439365e-07, + "loss": 1.5468, + "step": 539 + }, + { + "epoch": 0.531496062992126, + "grad_norm": 12.69034194946289, + "learning_rate": 5.318091451292247e-07, + "loss": 1.4316, + "step": 540 + }, + { + "epoch": 0.53248031496063, + "grad_norm": 10.690762519836426, + "learning_rate": 5.32803180914513e-07, + "loss": 1.2797, + "step": 541 + }, + { + "epoch": 0.5334645669291339, + "grad_norm": 12.846982955932617, + "learning_rate": 5.337972166998013e-07, + "loss": 1.9122, + "step": 542 + }, + { + "epoch": 0.5344488188976378, + "grad_norm": 15.378868103027344, + "learning_rate": 5.347912524850895e-07, + "loss": 2.0367, + "step": 543 + }, + { + "epoch": 0.5354330708661418, + "grad_norm": 30.10799217224121, + "learning_rate": 5.357852882703778e-07, + "loss": 3.3029, + "step": 544 + }, + { + "epoch": 0.5364173228346457, + "grad_norm": 47.62705993652344, + "learning_rate": 5.367793240556661e-07, + "loss": 3.9263, + "step": 545 + }, + { + "epoch": 0.5374015748031497, + "grad_norm": 28.87757110595703, + "learning_rate": 5.377733598409544e-07, + "loss": 3.0101, + "step": 546 + }, + { + "epoch": 0.5383858267716536, + "grad_norm": 33.99931335449219, + "learning_rate": 5.387673956262426e-07, + "loss": 3.3555, + "step": 547 + }, + { + "epoch": 0.5393700787401575, + "grad_norm": 10.475042343139648, + "learning_rate": 5.397614314115308e-07, + "loss": 1.2068, + "step": 548 + }, + { + "epoch": 0.5403543307086615, + "grad_norm": 8.499237060546875, + "learning_rate": 5.407554671968192e-07, + "loss": 1.1566, + "step": 549 + }, + { + "epoch": 0.5413385826771654, + "grad_norm": 10.652088165283203, + "learning_rate": 5.417495029821074e-07, + "loss": 1.2773, + "step": 550 + }, + { + "epoch": 0.5423228346456693, + "grad_norm": 10.508785247802734, + "learning_rate": 5.427435387673957e-07, + "loss": 1.4047, + "step": 551 + }, + { + "epoch": 0.5433070866141733, + "grad_norm": 10.353549003601074, + "learning_rate": 5.43737574552684e-07, + "loss": 1.6048, + "step": 552 + }, + { + "epoch": 0.5442913385826772, + "grad_norm": 8.851251602172852, + "learning_rate": 5.447316103379722e-07, + "loss": 1.217, + "step": 553 + }, + { + "epoch": 0.5452755905511811, + "grad_norm": 11.29096508026123, + "learning_rate": 5.457256461232605e-07, + "loss": 1.8104, + "step": 554 + }, + { + "epoch": 0.5462598425196851, + "grad_norm": 12.565424919128418, + "learning_rate": 5.467196819085487e-07, + "loss": 1.687, + "step": 555 + }, + { + "epoch": 0.547244094488189, + "grad_norm": 11.023751258850098, + "learning_rate": 5.477137176938371e-07, + "loss": 1.6702, + "step": 556 + }, + { + "epoch": 0.5482283464566929, + "grad_norm": 12.920904159545898, + "learning_rate": 5.487077534791253e-07, + "loss": 1.7011, + "step": 557 + }, + { + "epoch": 0.5492125984251969, + "grad_norm": 12.08708667755127, + "learning_rate": 5.497017892644135e-07, + "loss": 1.7341, + "step": 558 + }, + { + "epoch": 0.5501968503937008, + "grad_norm": 10.975556373596191, + "learning_rate": 5.50695825049702e-07, + "loss": 1.5006, + "step": 559 + }, + { + "epoch": 0.5511811023622047, + "grad_norm": 11.018404960632324, + "learning_rate": 5.516898608349902e-07, + "loss": 1.2778, + "step": 560 + }, + { + "epoch": 0.5521653543307087, + "grad_norm": 11.709118843078613, + "learning_rate": 5.526838966202784e-07, + "loss": 1.5081, + "step": 561 + }, + { + "epoch": 0.5531496062992126, + "grad_norm": 11.321682929992676, + "learning_rate": 5.536779324055667e-07, + "loss": 1.2398, + "step": 562 + }, + { + "epoch": 0.5541338582677166, + "grad_norm": 10.15671157836914, + "learning_rate": 5.54671968190855e-07, + "loss": 1.1054, + "step": 563 + }, + { + "epoch": 0.5551181102362205, + "grad_norm": 45.36001205444336, + "learning_rate": 5.556660039761433e-07, + "loss": 4.0185, + "step": 564 + }, + { + "epoch": 0.5561023622047244, + "grad_norm": 10.333076477050781, + "learning_rate": 5.566600397614315e-07, + "loss": 1.0427, + "step": 565 + }, + { + "epoch": 0.5570866141732284, + "grad_norm": 11.81332015991211, + "learning_rate": 5.576540755467198e-07, + "loss": 1.3934, + "step": 566 + }, + { + "epoch": 0.5580708661417323, + "grad_norm": 12.600215911865234, + "learning_rate": 5.586481113320081e-07, + "loss": 1.2378, + "step": 567 + }, + { + "epoch": 0.5590551181102362, + "grad_norm": 10.777389526367188, + "learning_rate": 5.596421471172963e-07, + "loss": 1.022, + "step": 568 + }, + { + "epoch": 0.5600393700787402, + "grad_norm": 9.640018463134766, + "learning_rate": 5.606361829025846e-07, + "loss": 0.9001, + "step": 569 + }, + { + "epoch": 0.5610236220472441, + "grad_norm": 13.254591941833496, + "learning_rate": 5.616302186878729e-07, + "loss": 1.3279, + "step": 570 + }, + { + "epoch": 0.562007874015748, + "grad_norm": 12.040815353393555, + "learning_rate": 5.626242544731611e-07, + "loss": 1.2889, + "step": 571 + }, + { + "epoch": 0.562992125984252, + "grad_norm": 10.555000305175781, + "learning_rate": 5.636182902584494e-07, + "loss": 0.9383, + "step": 572 + }, + { + "epoch": 0.5639763779527559, + "grad_norm": 15.663253784179688, + "learning_rate": 5.646123260437377e-07, + "loss": 1.749, + "step": 573 + }, + { + "epoch": 0.5649606299212598, + "grad_norm": 9.658116340637207, + "learning_rate": 5.656063618290259e-07, + "loss": 0.7669, + "step": 574 + }, + { + "epoch": 0.5659448818897638, + "grad_norm": 10.09415054321289, + "learning_rate": 5.666003976143142e-07, + "loss": 0.9355, + "step": 575 + }, + { + "epoch": 0.5669291338582677, + "grad_norm": 14.240106582641602, + "learning_rate": 5.675944333996024e-07, + "loss": 1.3596, + "step": 576 + }, + { + "epoch": 0.5679133858267716, + "grad_norm": 56.997859954833984, + "learning_rate": 5.685884691848908e-07, + "loss": 5.5102, + "step": 577 + }, + { + "epoch": 0.5688976377952756, + "grad_norm": 11.803168296813965, + "learning_rate": 5.69582504970179e-07, + "loss": 0.7984, + "step": 578 + }, + { + "epoch": 0.5698818897637795, + "grad_norm": 10.7587251663208, + "learning_rate": 5.705765407554672e-07, + "loss": 0.8871, + "step": 579 + }, + { + "epoch": 0.5708661417322834, + "grad_norm": 12.7469482421875, + "learning_rate": 5.715705765407556e-07, + "loss": 1.1151, + "step": 580 + }, + { + "epoch": 0.5718503937007874, + "grad_norm": 10.369434356689453, + "learning_rate": 5.725646123260438e-07, + "loss": 0.9502, + "step": 581 + }, + { + "epoch": 0.5728346456692913, + "grad_norm": 35.74211502075195, + "learning_rate": 5.735586481113321e-07, + "loss": 3.6492, + "step": 582 + }, + { + "epoch": 0.5738188976377953, + "grad_norm": 35.057254791259766, + "learning_rate": 5.745526838966203e-07, + "loss": 3.4262, + "step": 583 + }, + { + "epoch": 0.5748031496062992, + "grad_norm": 11.208961486816406, + "learning_rate": 5.755467196819086e-07, + "loss": 1.3362, + "step": 584 + }, + { + "epoch": 0.5757874015748031, + "grad_norm": 9.760231018066406, + "learning_rate": 5.765407554671969e-07, + "loss": 0.9015, + "step": 585 + }, + { + "epoch": 0.5767716535433071, + "grad_norm": 13.841927528381348, + "learning_rate": 5.775347912524851e-07, + "loss": 1.5884, + "step": 586 + }, + { + "epoch": 0.577755905511811, + "grad_norm": 11.57701301574707, + "learning_rate": 5.785288270377734e-07, + "loss": 1.109, + "step": 587 + }, + { + "epoch": 0.5787401574803149, + "grad_norm": 10.33425521850586, + "learning_rate": 5.795228628230618e-07, + "loss": 1.041, + "step": 588 + }, + { + "epoch": 0.5797244094488189, + "grad_norm": 11.956677436828613, + "learning_rate": 5.8051689860835e-07, + "loss": 1.4892, + "step": 589 + }, + { + "epoch": 0.5807086614173228, + "grad_norm": 10.908195495605469, + "learning_rate": 5.815109343936383e-07, + "loss": 1.2623, + "step": 590 + }, + { + "epoch": 0.5816929133858267, + "grad_norm": 12.974413871765137, + "learning_rate": 5.825049701789266e-07, + "loss": 1.5302, + "step": 591 + }, + { + "epoch": 0.5826771653543307, + "grad_norm": 10.713995933532715, + "learning_rate": 5.834990059642148e-07, + "loss": 1.3517, + "step": 592 + }, + { + "epoch": 0.5836614173228346, + "grad_norm": 7.986886501312256, + "learning_rate": 5.844930417495031e-07, + "loss": 0.6166, + "step": 593 + }, + { + "epoch": 0.5846456692913385, + "grad_norm": 12.688620567321777, + "learning_rate": 5.854870775347914e-07, + "loss": 1.6761, + "step": 594 + }, + { + "epoch": 0.5856299212598425, + "grad_norm": 10.7589693069458, + "learning_rate": 5.864811133200797e-07, + "loss": 1.1115, + "step": 595 + }, + { + "epoch": 0.5866141732283464, + "grad_norm": 10.918403625488281, + "learning_rate": 5.874751491053679e-07, + "loss": 1.2945, + "step": 596 + }, + { + "epoch": 0.5875984251968503, + "grad_norm": 12.224095344543457, + "learning_rate": 5.884691848906561e-07, + "loss": 1.4378, + "step": 597 + }, + { + "epoch": 0.5885826771653543, + "grad_norm": 10.335618019104004, + "learning_rate": 5.894632206759445e-07, + "loss": 0.9928, + "step": 598 + }, + { + "epoch": 0.5895669291338582, + "grad_norm": 13.077963829040527, + "learning_rate": 5.904572564612327e-07, + "loss": 0.9898, + "step": 599 + }, + { + "epoch": 0.5905511811023622, + "grad_norm": 48.9807243347168, + "learning_rate": 5.914512922465209e-07, + "loss": 4.6887, + "step": 600 + }, + { + "epoch": 0.5915354330708661, + "grad_norm": 11.930185317993164, + "learning_rate": 5.924453280318093e-07, + "loss": 1.2254, + "step": 601 + }, + { + "epoch": 0.59251968503937, + "grad_norm": 12.545085906982422, + "learning_rate": 5.934393638170975e-07, + "loss": 1.2707, + "step": 602 + }, + { + "epoch": 0.593503937007874, + "grad_norm": 16.840160369873047, + "learning_rate": 5.944333996023858e-07, + "loss": 1.8289, + "step": 603 + }, + { + "epoch": 0.594488188976378, + "grad_norm": 10.321748733520508, + "learning_rate": 5.95427435387674e-07, + "loss": 0.7801, + "step": 604 + }, + { + "epoch": 0.5954724409448819, + "grad_norm": 11.279508590698242, + "learning_rate": 5.964214711729623e-07, + "loss": 0.9111, + "step": 605 + }, + { + "epoch": 0.5964566929133859, + "grad_norm": 11.96855354309082, + "learning_rate": 5.974155069582506e-07, + "loss": 1.1405, + "step": 606 + }, + { + "epoch": 0.5974409448818898, + "grad_norm": 10.992218971252441, + "learning_rate": 5.984095427435388e-07, + "loss": 1.0497, + "step": 607 + }, + { + "epoch": 0.5984251968503937, + "grad_norm": 10.756412506103516, + "learning_rate": 5.994035785288272e-07, + "loss": 1.0792, + "step": 608 + }, + { + "epoch": 0.5994094488188977, + "grad_norm": 10.584625244140625, + "learning_rate": 6.003976143141154e-07, + "loss": 0.9699, + "step": 609 + }, + { + "epoch": 0.6003937007874016, + "grad_norm": 13.49753475189209, + "learning_rate": 6.013916500994036e-07, + "loss": 0.9398, + "step": 610 + }, + { + "epoch": 0.6013779527559056, + "grad_norm": 13.567956924438477, + "learning_rate": 6.02385685884692e-07, + "loss": 1.5483, + "step": 611 + }, + { + "epoch": 0.6023622047244095, + "grad_norm": 10.190022468566895, + "learning_rate": 6.033797216699802e-07, + "loss": 0.997, + "step": 612 + }, + { + "epoch": 0.6023622047244095, + "eval_Qnli-dev_cosine_accuracy": 0.646484375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8648326992988586, + "eval_Qnli-dev_cosine_ap": 0.667348431183779, + "eval_Qnli-dev_cosine_f1": 0.6574074074074073, + "eval_Qnli-dev_cosine_f1_threshold": 0.7951723337173462, + "eval_Qnli-dev_cosine_precision": 0.5169902912621359, + "eval_Qnli-dev_cosine_recall": 0.902542372881356, + "eval_Qnli-dev_dot_accuracy": 0.646484375, + "eval_Qnli-dev_dot_accuracy_threshold": 661.619873046875, + "eval_Qnli-dev_dot_ap": 0.6663803020252368, + "eval_Qnli-dev_dot_f1": 0.6573643410852713, + "eval_Qnli-dev_dot_f1_threshold": 609.5113525390625, + "eval_Qnli-dev_dot_precision": 0.5183374083129584, + "eval_Qnli-dev_dot_recall": 0.8983050847457628, + "eval_Qnli-dev_euclidean_accuracy": 0.646484375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 14.383621215820312, + "eval_Qnli-dev_euclidean_ap": 0.6675778194412583, + "eval_Qnli-dev_euclidean_f1": 0.6574500768049155, + "eval_Qnli-dev_euclidean_f1_threshold": 17.748760223388672, + "eval_Qnli-dev_euclidean_precision": 0.5156626506024097, + "eval_Qnli-dev_euclidean_recall": 0.9067796610169492, + "eval_Qnli-dev_manhattan_accuracy": 0.662109375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 295.7828369140625, + "eval_Qnli-dev_manhattan_ap": 0.6700872776325868, + "eval_Qnli-dev_manhattan_f1": 0.6732026143790849, + "eval_Qnli-dev_manhattan_f1_threshold": 347.7679443359375, + "eval_Qnli-dev_manhattan_precision": 0.5478723404255319, + "eval_Qnli-dev_manhattan_recall": 0.8728813559322034, + "eval_Qnli-dev_max_accuracy": 0.662109375, + "eval_Qnli-dev_max_accuracy_threshold": 661.619873046875, + "eval_Qnli-dev_max_ap": 0.6700872776325868, + "eval_Qnli-dev_max_f1": 0.6732026143790849, + "eval_Qnli-dev_max_f1_threshold": 609.5113525390625, + "eval_Qnli-dev_max_precision": 0.5478723404255319, + "eval_Qnli-dev_max_recall": 0.9067796610169492, + "eval_allNLI-dev_cosine_accuracy": 0.67578125, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9710712432861328, + "eval_allNLI-dev_cosine_ap": 0.4550292124468429, + "eval_allNLI-dev_cosine_f1": 0.5226480836236935, + "eval_allNLI-dev_cosine_f1_threshold": 0.8640801906585693, + "eval_allNLI-dev_cosine_precision": 0.3740648379052369, + "eval_allNLI-dev_cosine_recall": 0.8670520231213873, + "eval_allNLI-dev_dot_accuracy": 0.67578125, + "eval_allNLI-dev_dot_accuracy_threshold": 744.9188232421875, + "eval_allNLI-dev_dot_ap": 0.4554218742381212, + "eval_allNLI-dev_dot_f1": 0.5226480836236935, + "eval_allNLI-dev_dot_f1_threshold": 661.9685668945312, + "eval_allNLI-dev_dot_precision": 0.3740648379052369, + "eval_allNLI-dev_dot_recall": 0.8670520231213873, + "eval_allNLI-dev_euclidean_accuracy": 0.67578125, + "eval_allNLI-dev_euclidean_accuracy_threshold": 6.661540985107422, + "eval_allNLI-dev_euclidean_ap": 0.45513348048034996, + "eval_allNLI-dev_euclidean_f1": 0.5226480836236935, + "eval_allNLI-dev_euclidean_f1_threshold": 14.431058883666992, + "eval_allNLI-dev_euclidean_precision": 0.3740648379052369, + "eval_allNLI-dev_euclidean_recall": 0.8670520231213873, + "eval_allNLI-dev_manhattan_accuracy": 0.67578125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 137.76480102539062, + "eval_allNLI-dev_manhattan_ap": 0.45337533571519506, + "eval_allNLI-dev_manhattan_f1": 0.5220458553791887, + "eval_allNLI-dev_manhattan_f1_threshold": 293.4437255859375, + "eval_allNLI-dev_manhattan_precision": 0.3756345177664975, + "eval_allNLI-dev_manhattan_recall": 0.8554913294797688, + "eval_allNLI-dev_max_accuracy": 0.67578125, + "eval_allNLI-dev_max_accuracy_threshold": 744.9188232421875, + "eval_allNLI-dev_max_ap": 0.4554218742381212, + "eval_allNLI-dev_max_f1": 0.5226480836236935, + "eval_allNLI-dev_max_f1_threshold": 661.9685668945312, + "eval_allNLI-dev_max_precision": 0.3756345177664975, + "eval_allNLI-dev_max_recall": 0.8670520231213873, + "eval_loss": 1.0046995878219604, + "eval_runtime": 50.2527, + "eval_samples_per_second": 33.113, + "eval_sequential_score": 0.6700872776325868, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.34857372903377226, + "eval_sts-test_pearson_dot": 0.34499075210005525, + "eval_sts-test_pearson_euclidean": 0.3698824430134462, + "eval_sts-test_pearson_manhattan": 0.3809531131406821, + "eval_sts-test_pearson_max": 0.3809531131406821, + "eval_sts-test_spearman_cosine": 0.3979645095653129, + "eval_sts-test_spearman_dot": 0.39459004070009446, + "eval_sts-test_spearman_euclidean": 0.3981739454686246, + "eval_sts-test_spearman_manhattan": 0.4083220547605681, + "eval_sts-test_spearman_max": 0.4083220547605681, + "step": 612 + }, + { + "epoch": 0.6033464566929134, + "grad_norm": 10.152923583984375, + "learning_rate": 6.043737574552684e-07, + "loss": 0.8358, + "step": 613 + }, + { + "epoch": 0.6043307086614174, + "grad_norm": 11.065042495727539, + "learning_rate": 6.053677932405567e-07, + "loss": 1.211, + "step": 614 + }, + { + "epoch": 0.6053149606299213, + "grad_norm": 69.22620391845703, + "learning_rate": 6.06361829025845e-07, + "loss": 6.7813, + "step": 615 + }, + { + "epoch": 0.6062992125984252, + "grad_norm": 10.513162612915039, + "learning_rate": 6.073558648111333e-07, + "loss": 1.1229, + "step": 616 + }, + { + "epoch": 0.6072834645669292, + "grad_norm": 10.532771110534668, + "learning_rate": 6.083499005964215e-07, + "loss": 1.0317, + "step": 617 + }, + { + "epoch": 0.6082677165354331, + "grad_norm": 11.072421073913574, + "learning_rate": 6.093439363817097e-07, + "loss": 1.2123, + "step": 618 + }, + { + "epoch": 0.609251968503937, + "grad_norm": 12.958882331848145, + "learning_rate": 6.103379721669982e-07, + "loss": 1.4073, + "step": 619 + }, + { + "epoch": 0.610236220472441, + "grad_norm": 9.524628639221191, + "learning_rate": 6.113320079522864e-07, + "loss": 0.9951, + "step": 620 + }, + { + "epoch": 0.6112204724409449, + "grad_norm": 10.776963233947754, + "learning_rate": 6.123260437375747e-07, + "loss": 1.3166, + "step": 621 + }, + { + "epoch": 0.6122047244094488, + "grad_norm": 40.76365661621094, + "learning_rate": 6.13320079522863e-07, + "loss": 4.5204, + "step": 622 + }, + { + "epoch": 0.6131889763779528, + "grad_norm": 8.555766105651855, + "learning_rate": 6.143141153081512e-07, + "loss": 0.6539, + "step": 623 + }, + { + "epoch": 0.6141732283464567, + "grad_norm": 10.885683059692383, + "learning_rate": 6.153081510934395e-07, + "loss": 1.1959, + "step": 624 + }, + { + "epoch": 0.6151574803149606, + "grad_norm": 42.833946228027344, + "learning_rate": 6.163021868787277e-07, + "loss": 4.2551, + "step": 625 + }, + { + "epoch": 0.6161417322834646, + "grad_norm": 11.948238372802734, + "learning_rate": 6.172962226640161e-07, + "loss": 1.2459, + "step": 626 + }, + { + "epoch": 0.6171259842519685, + "grad_norm": 11.74181842803955, + "learning_rate": 6.182902584493043e-07, + "loss": 1.3758, + "step": 627 + }, + { + "epoch": 0.6181102362204725, + "grad_norm": 12.536766052246094, + "learning_rate": 6.192842942345925e-07, + "loss": 1.0524, + "step": 628 + }, + { + "epoch": 0.6190944881889764, + "grad_norm": 12.458625793457031, + "learning_rate": 6.202783300198809e-07, + "loss": 1.5197, + "step": 629 + }, + { + "epoch": 0.6200787401574803, + "grad_norm": 12.924437522888184, + "learning_rate": 6.212723658051691e-07, + "loss": 1.0201, + "step": 630 + }, + { + "epoch": 0.6210629921259843, + "grad_norm": 10.728304862976074, + "learning_rate": 6.222664015904573e-07, + "loss": 0.9007, + "step": 631 + }, + { + "epoch": 0.6220472440944882, + "grad_norm": 9.886204719543457, + "learning_rate": 6.232604373757457e-07, + "loss": 0.8418, + "step": 632 + }, + { + "epoch": 0.6230314960629921, + "grad_norm": 13.076040267944336, + "learning_rate": 6.242544731610339e-07, + "loss": 1.4343, + "step": 633 + }, + { + "epoch": 0.6240157480314961, + "grad_norm": 9.723533630371094, + "learning_rate": 6.252485089463222e-07, + "loss": 0.5292, + "step": 634 + }, + { + "epoch": 0.625, + "grad_norm": 10.228574752807617, + "learning_rate": 6.262425447316104e-07, + "loss": 0.8549, + "step": 635 + }, + { + "epoch": 0.6259842519685039, + "grad_norm": 10.32657241821289, + "learning_rate": 6.272365805168987e-07, + "loss": 0.8703, + "step": 636 + }, + { + "epoch": 0.6269685039370079, + "grad_norm": 12.475861549377441, + "learning_rate": 6.28230616302187e-07, + "loss": 0.9911, + "step": 637 + }, + { + "epoch": 0.6279527559055118, + "grad_norm": 14.592931747436523, + "learning_rate": 6.292246520874752e-07, + "loss": 1.3342, + "step": 638 + }, + { + "epoch": 0.6289370078740157, + "grad_norm": 12.569700241088867, + "learning_rate": 6.302186878727636e-07, + "loss": 1.1332, + "step": 639 + }, + { + "epoch": 0.6299212598425197, + "grad_norm": 38.98273468017578, + "learning_rate": 6.312127236580518e-07, + "loss": 3.9965, + "step": 640 + }, + { + "epoch": 0.6309055118110236, + "grad_norm": 12.703550338745117, + "learning_rate": 6.3220675944334e-07, + "loss": 0.7236, + "step": 641 + }, + { + "epoch": 0.6318897637795275, + "grad_norm": 11.266120910644531, + "learning_rate": 6.332007952286283e-07, + "loss": 0.9079, + "step": 642 + }, + { + "epoch": 0.6328740157480315, + "grad_norm": 12.65903377532959, + "learning_rate": 6.341948310139166e-07, + "loss": 1.0967, + "step": 643 + }, + { + "epoch": 0.6338582677165354, + "grad_norm": 14.934013366699219, + "learning_rate": 6.351888667992048e-07, + "loss": 1.4183, + "step": 644 + }, + { + "epoch": 0.6348425196850394, + "grad_norm": 13.014033317565918, + "learning_rate": 6.361829025844931e-07, + "loss": 1.3841, + "step": 645 + }, + { + "epoch": 0.6358267716535433, + "grad_norm": 14.226466178894043, + "learning_rate": 6.371769383697813e-07, + "loss": 1.2982, + "step": 646 + }, + { + "epoch": 0.6368110236220472, + "grad_norm": 11.135141372680664, + "learning_rate": 6.381709741550698e-07, + "loss": 0.9048, + "step": 647 + }, + { + "epoch": 0.6377952755905512, + "grad_norm": 9.338624954223633, + "learning_rate": 6.39165009940358e-07, + "loss": 0.7918, + "step": 648 + }, + { + "epoch": 0.6387795275590551, + "grad_norm": 7.227178573608398, + "learning_rate": 6.401590457256462e-07, + "loss": 0.3685, + "step": 649 + }, + { + "epoch": 0.639763779527559, + "grad_norm": 9.620096206665039, + "learning_rate": 6.411530815109346e-07, + "loss": 0.6949, + "step": 650 + }, + { + "epoch": 0.640748031496063, + "grad_norm": 56.73809814453125, + "learning_rate": 6.421471172962228e-07, + "loss": 5.1568, + "step": 651 + }, + { + "epoch": 0.6417322834645669, + "grad_norm": 14.595137596130371, + "learning_rate": 6.431411530815111e-07, + "loss": 1.3943, + "step": 652 + }, + { + "epoch": 0.6427165354330708, + "grad_norm": 9.840551376342773, + "learning_rate": 6.441351888667994e-07, + "loss": 0.8608, + "step": 653 + }, + { + "epoch": 0.6437007874015748, + "grad_norm": 10.686992645263672, + "learning_rate": 6.451292246520876e-07, + "loss": 0.8197, + "step": 654 + }, + { + "epoch": 0.6446850393700787, + "grad_norm": 11.54585075378418, + "learning_rate": 6.461232604373759e-07, + "loss": 0.822, + "step": 655 + }, + { + "epoch": 0.6456692913385826, + "grad_norm": 40.4345588684082, + "learning_rate": 6.471172962226641e-07, + "loss": 3.2918, + "step": 656 + }, + { + "epoch": 0.6466535433070866, + "grad_norm": 8.040945053100586, + "learning_rate": 6.481113320079524e-07, + "loss": 0.5596, + "step": 657 + }, + { + "epoch": 0.6476377952755905, + "grad_norm": 37.2348747253418, + "learning_rate": 6.491053677932407e-07, + "loss": 4.1499, + "step": 658 + }, + { + "epoch": 0.6486220472440944, + "grad_norm": 10.862921714782715, + "learning_rate": 6.500994035785289e-07, + "loss": 1.0279, + "step": 659 + }, + { + "epoch": 0.6496062992125984, + "grad_norm": 13.189752578735352, + "learning_rate": 6.510934393638173e-07, + "loss": 1.1506, + "step": 660 + }, + { + "epoch": 0.6505905511811023, + "grad_norm": 13.04338264465332, + "learning_rate": 6.520874751491055e-07, + "loss": 1.1673, + "step": 661 + }, + { + "epoch": 0.6515748031496063, + "grad_norm": 10.274197578430176, + "learning_rate": 6.530815109343937e-07, + "loss": 0.96, + "step": 662 + }, + { + "epoch": 0.6525590551181102, + "grad_norm": 33.91826629638672, + "learning_rate": 6.54075546719682e-07, + "loss": 3.5414, + "step": 663 + }, + { + "epoch": 0.6535433070866141, + "grad_norm": 9.1599702835083, + "learning_rate": 6.550695825049703e-07, + "loss": 0.6599, + "step": 664 + }, + { + "epoch": 0.6545275590551181, + "grad_norm": 35.5181770324707, + "learning_rate": 6.560636182902586e-07, + "loss": 3.5518, + "step": 665 + }, + { + "epoch": 0.655511811023622, + "grad_norm": 12.839941024780273, + "learning_rate": 6.570576540755468e-07, + "loss": 1.1906, + "step": 666 + }, + { + "epoch": 0.656496062992126, + "grad_norm": 21.329296112060547, + "learning_rate": 6.580516898608351e-07, + "loss": 2.1353, + "step": 667 + }, + { + "epoch": 0.65748031496063, + "grad_norm": 9.478302955627441, + "learning_rate": 6.590457256461234e-07, + "loss": 0.7083, + "step": 668 + }, + { + "epoch": 0.6584645669291339, + "grad_norm": 40.32704162597656, + "learning_rate": 6.600397614314116e-07, + "loss": 2.9425, + "step": 669 + }, + { + "epoch": 0.6594488188976378, + "grad_norm": 9.107198715209961, + "learning_rate": 6.610337972166998e-07, + "loss": 0.9433, + "step": 670 + }, + { + "epoch": 0.6604330708661418, + "grad_norm": 15.187662124633789, + "learning_rate": 6.620278330019882e-07, + "loss": 1.8499, + "step": 671 + }, + { + "epoch": 0.6614173228346457, + "grad_norm": 11.691488265991211, + "learning_rate": 6.630218687872764e-07, + "loss": 1.1614, + "step": 672 + }, + { + "epoch": 0.6624015748031497, + "grad_norm": 10.95936107635498, + "learning_rate": 6.640159045725647e-07, + "loss": 1.0474, + "step": 673 + }, + { + "epoch": 0.6633858267716536, + "grad_norm": 11.567349433898926, + "learning_rate": 6.65009940357853e-07, + "loss": 1.2895, + "step": 674 + }, + { + "epoch": 0.6643700787401575, + "grad_norm": 10.821932792663574, + "learning_rate": 6.660039761431412e-07, + "loss": 0.9789, + "step": 675 + }, + { + "epoch": 0.6653543307086615, + "grad_norm": 9.049556732177734, + "learning_rate": 6.669980119284295e-07, + "loss": 0.7719, + "step": 676 + }, + { + "epoch": 0.6663385826771654, + "grad_norm": 11.720834732055664, + "learning_rate": 6.679920477137177e-07, + "loss": 1.2203, + "step": 677 + }, + { + "epoch": 0.6673228346456693, + "grad_norm": 12.704697608947754, + "learning_rate": 6.689860834990062e-07, + "loss": 1.0516, + "step": 678 + }, + { + "epoch": 0.6683070866141733, + "grad_norm": 29.334060668945312, + "learning_rate": 6.699801192842944e-07, + "loss": 2.5514, + "step": 679 + }, + { + "epoch": 0.6692913385826772, + "grad_norm": 10.399552345275879, + "learning_rate": 6.709741550695826e-07, + "loss": 0.7346, + "step": 680 + }, + { + "epoch": 0.6702755905511811, + "grad_norm": 10.699383735656738, + "learning_rate": 6.71968190854871e-07, + "loss": 1.0245, + "step": 681 + }, + { + "epoch": 0.6712598425196851, + "grad_norm": 32.639041900634766, + "learning_rate": 6.729622266401592e-07, + "loss": 2.8005, + "step": 682 + }, + { + "epoch": 0.672244094488189, + "grad_norm": 12.40882682800293, + "learning_rate": 6.739562624254474e-07, + "loss": 1.3212, + "step": 683 + }, + { + "epoch": 0.6732283464566929, + "grad_norm": 11.128901481628418, + "learning_rate": 6.749502982107357e-07, + "loss": 0.95, + "step": 684 + }, + { + "epoch": 0.6742125984251969, + "grad_norm": 10.618715286254883, + "learning_rate": 6.75944333996024e-07, + "loss": 1.0483, + "step": 685 + }, + { + "epoch": 0.6751968503937008, + "grad_norm": 9.804398536682129, + "learning_rate": 6.769383697813123e-07, + "loss": 0.8504, + "step": 686 + }, + { + "epoch": 0.6761811023622047, + "grad_norm": 25.071557998657227, + "learning_rate": 6.779324055666005e-07, + "loss": 2.281, + "step": 687 + }, + { + "epoch": 0.6771653543307087, + "grad_norm": 19.744482040405273, + "learning_rate": 6.789264413518888e-07, + "loss": 1.8153, + "step": 688 + }, + { + "epoch": 0.6781496062992126, + "grad_norm": 11.338679313659668, + "learning_rate": 6.799204771371771e-07, + "loss": 1.3652, + "step": 689 + }, + { + "epoch": 0.6791338582677166, + "grad_norm": 11.757587432861328, + "learning_rate": 6.809145129224653e-07, + "loss": 1.0949, + "step": 690 + }, + { + "epoch": 0.6801181102362205, + "grad_norm": 10.77193546295166, + "learning_rate": 6.819085487077536e-07, + "loss": 1.2196, + "step": 691 + }, + { + "epoch": 0.6811023622047244, + "grad_norm": 10.206846237182617, + "learning_rate": 6.829025844930419e-07, + "loss": 0.7995, + "step": 692 + }, + { + "epoch": 0.6820866141732284, + "grad_norm": 14.147713661193848, + "learning_rate": 6.838966202783301e-07, + "loss": 1.5108, + "step": 693 + }, + { + "epoch": 0.6830708661417323, + "grad_norm": 10.401080131530762, + "learning_rate": 6.848906560636184e-07, + "loss": 0.7933, + "step": 694 + }, + { + "epoch": 0.6840551181102362, + "grad_norm": 12.020344734191895, + "learning_rate": 6.858846918489067e-07, + "loss": 1.2367, + "step": 695 + }, + { + "epoch": 0.6850393700787402, + "grad_norm": 9.998419761657715, + "learning_rate": 6.86878727634195e-07, + "loss": 1.0352, + "step": 696 + }, + { + "epoch": 0.6860236220472441, + "grad_norm": 11.320432662963867, + "learning_rate": 6.878727634194832e-07, + "loss": 1.1709, + "step": 697 + }, + { + "epoch": 0.687007874015748, + "grad_norm": 12.676666259765625, + "learning_rate": 6.888667992047714e-07, + "loss": 1.452, + "step": 698 + }, + { + "epoch": 0.687992125984252, + "grad_norm": 11.981403350830078, + "learning_rate": 6.898608349900598e-07, + "loss": 0.8497, + "step": 699 + }, + { + "epoch": 0.6889763779527559, + "grad_norm": 29.784114837646484, + "learning_rate": 6.90854870775348e-07, + "loss": 2.8109, + "step": 700 + }, + { + "epoch": 0.6899606299212598, + "grad_norm": 32.6761474609375, + "learning_rate": 6.918489065606362e-07, + "loss": 2.6196, + "step": 701 + }, + { + "epoch": 0.6909448818897638, + "grad_norm": 12.863127708435059, + "learning_rate": 6.928429423459246e-07, + "loss": 1.4556, + "step": 702 + }, + { + "epoch": 0.6919291338582677, + "grad_norm": 12.314940452575684, + "learning_rate": 6.938369781312128e-07, + "loss": 1.3494, + "step": 703 + }, + { + "epoch": 0.6929133858267716, + "grad_norm": 20.81646728515625, + "learning_rate": 6.948310139165011e-07, + "loss": 1.6624, + "step": 704 + }, + { + "epoch": 0.6938976377952756, + "grad_norm": 20.603694915771484, + "learning_rate": 6.958250497017893e-07, + "loss": 1.6169, + "step": 705 + }, + { + "epoch": 0.6948818897637795, + "grad_norm": 8.17772388458252, + "learning_rate": 6.968190854870776e-07, + "loss": 0.5565, + "step": 706 + }, + { + "epoch": 0.6958661417322834, + "grad_norm": 8.934789657592773, + "learning_rate": 6.97813121272366e-07, + "loss": 0.8594, + "step": 707 + }, + { + "epoch": 0.6968503937007874, + "grad_norm": 10.263940811157227, + "learning_rate": 6.988071570576542e-07, + "loss": 0.8551, + "step": 708 + }, + { + "epoch": 0.6978346456692913, + "grad_norm": 11.96685791015625, + "learning_rate": 6.998011928429426e-07, + "loss": 1.1693, + "step": 709 + }, + { + "epoch": 0.6988188976377953, + "grad_norm": 10.99656867980957, + "learning_rate": 7.007952286282308e-07, + "loss": 1.0514, + "step": 710 + }, + { + "epoch": 0.6998031496062992, + "grad_norm": 12.418295860290527, + "learning_rate": 7.01789264413519e-07, + "loss": 1.1862, + "step": 711 + }, + { + "epoch": 0.7007874015748031, + "grad_norm": 11.22188949584961, + "learning_rate": 7.027833001988073e-07, + "loss": 0.8359, + "step": 712 + }, + { + "epoch": 0.7017716535433071, + "grad_norm": 9.169300079345703, + "learning_rate": 7.037773359840956e-07, + "loss": 0.7692, + "step": 713 + }, + { + "epoch": 0.702755905511811, + "grad_norm": 22.9448299407959, + "learning_rate": 7.047713717693838e-07, + "loss": 1.815, + "step": 714 + }, + { + "epoch": 0.7037401574803149, + "grad_norm": 9.696531295776367, + "learning_rate": 7.057654075546721e-07, + "loss": 1.0751, + "step": 715 + }, + { + "epoch": 0.7047244094488189, + "grad_norm": 9.424071311950684, + "learning_rate": 7.067594433399604e-07, + "loss": 0.6526, + "step": 716 + }, + { + "epoch": 0.7057086614173228, + "grad_norm": 12.614093780517578, + "learning_rate": 7.077534791252487e-07, + "loss": 1.1617, + "step": 717 + }, + { + "epoch": 0.7066929133858267, + "grad_norm": 11.695466041564941, + "learning_rate": 7.087475149105369e-07, + "loss": 1.0783, + "step": 718 + }, + { + "epoch": 0.7076771653543307, + "grad_norm": 10.782369613647461, + "learning_rate": 7.097415506958251e-07, + "loss": 0.7916, + "step": 719 + }, + { + "epoch": 0.7086614173228346, + "grad_norm": 13.475589752197266, + "learning_rate": 7.107355864811135e-07, + "loss": 1.3039, + "step": 720 + }, + { + "epoch": 0.7096456692913385, + "grad_norm": 14.771034240722656, + "learning_rate": 7.117296222664017e-07, + "loss": 1.1156, + "step": 721 + }, + { + "epoch": 0.7106299212598425, + "grad_norm": 12.65285587310791, + "learning_rate": 7.1272365805169e-07, + "loss": 1.0529, + "step": 722 + }, + { + "epoch": 0.7116141732283464, + "grad_norm": 10.499075889587402, + "learning_rate": 7.137176938369783e-07, + "loss": 0.8265, + "step": 723 + }, + { + "epoch": 0.7125984251968503, + "grad_norm": 9.90308952331543, + "learning_rate": 7.147117296222665e-07, + "loss": 0.8019, + "step": 724 + }, + { + "epoch": 0.7135826771653543, + "grad_norm": 9.007514953613281, + "learning_rate": 7.157057654075548e-07, + "loss": 0.6116, + "step": 725 + }, + { + "epoch": 0.7145669291338582, + "grad_norm": 13.224435806274414, + "learning_rate": 7.16699801192843e-07, + "loss": 1.135, + "step": 726 + }, + { + "epoch": 0.7155511811023622, + "grad_norm": 11.456241607666016, + "learning_rate": 7.176938369781313e-07, + "loss": 0.7692, + "step": 727 + }, + { + "epoch": 0.7165354330708661, + "grad_norm": 30.24581527709961, + "learning_rate": 7.186878727634196e-07, + "loss": 2.3559, + "step": 728 + }, + { + "epoch": 0.71751968503937, + "grad_norm": 13.747896194458008, + "learning_rate": 7.196819085487078e-07, + "loss": 1.352, + "step": 729 + }, + { + "epoch": 0.718503937007874, + "grad_norm": 33.27655029296875, + "learning_rate": 7.206759443339962e-07, + "loss": 2.823, + "step": 730 + }, + { + "epoch": 0.719488188976378, + "grad_norm": 12.033122062683105, + "learning_rate": 7.216699801192844e-07, + "loss": 1.0067, + "step": 731 + }, + { + "epoch": 0.7204724409448819, + "grad_norm": 10.001436233520508, + "learning_rate": 7.226640159045726e-07, + "loss": 0.9077, + "step": 732 + }, + { + "epoch": 0.7214566929133859, + "grad_norm": 11.733288764953613, + "learning_rate": 7.236580516898609e-07, + "loss": 1.0933, + "step": 733 + }, + { + "epoch": 0.7224409448818898, + "grad_norm": 10.980072975158691, + "learning_rate": 7.246520874751492e-07, + "loss": 0.8174, + "step": 734 + }, + { + "epoch": 0.7234251968503937, + "grad_norm": 12.679802894592285, + "learning_rate": 7.256461232604375e-07, + "loss": 1.2212, + "step": 735 + }, + { + "epoch": 0.7244094488188977, + "grad_norm": 12.101704597473145, + "learning_rate": 7.266401590457257e-07, + "loss": 1.1557, + "step": 736 + }, + { + "epoch": 0.7253937007874016, + "grad_norm": 11.133832931518555, + "learning_rate": 7.27634194831014e-07, + "loss": 0.6191, + "step": 737 + }, + { + "epoch": 0.7263779527559056, + "grad_norm": 20.73958969116211, + "learning_rate": 7.286282306163024e-07, + "loss": 1.7437, + "step": 738 + }, + { + "epoch": 0.7273622047244095, + "grad_norm": 9.562580108642578, + "learning_rate": 7.296222664015906e-07, + "loss": 0.8977, + "step": 739 + }, + { + "epoch": 0.7283464566929134, + "grad_norm": 12.212095260620117, + "learning_rate": 7.306163021868788e-07, + "loss": 1.0782, + "step": 740 + }, + { + "epoch": 0.7293307086614174, + "grad_norm": 11.410839080810547, + "learning_rate": 7.316103379721672e-07, + "loss": 0.8985, + "step": 741 + }, + { + "epoch": 0.7303149606299213, + "grad_norm": 15.690488815307617, + "learning_rate": 7.326043737574554e-07, + "loss": 1.4867, + "step": 742 + }, + { + "epoch": 0.7312992125984252, + "grad_norm": 11.45366096496582, + "learning_rate": 7.335984095427437e-07, + "loss": 0.7497, + "step": 743 + }, + { + "epoch": 0.7322834645669292, + "grad_norm": 9.350363731384277, + "learning_rate": 7.34592445328032e-07, + "loss": 0.6433, + "step": 744 + }, + { + "epoch": 0.7332677165354331, + "grad_norm": 24.7938289642334, + "learning_rate": 7.355864811133202e-07, + "loss": 1.4175, + "step": 745 + }, + { + "epoch": 0.734251968503937, + "grad_norm": 12.32961368560791, + "learning_rate": 7.365805168986085e-07, + "loss": 1.1896, + "step": 746 + }, + { + "epoch": 0.735236220472441, + "grad_norm": 23.96776008605957, + "learning_rate": 7.375745526838967e-07, + "loss": 1.9867, + "step": 747 + }, + { + "epoch": 0.7362204724409449, + "grad_norm": 10.38709545135498, + "learning_rate": 7.385685884691851e-07, + "loss": 0.8968, + "step": 748 + }, + { + "epoch": 0.7372047244094488, + "grad_norm": 9.817655563354492, + "learning_rate": 7.395626242544733e-07, + "loss": 0.7265, + "step": 749 + }, + { + "epoch": 0.7381889763779528, + "grad_norm": 11.957969665527344, + "learning_rate": 7.405566600397615e-07, + "loss": 0.9418, + "step": 750 + }, + { + "epoch": 0.7391732283464567, + "grad_norm": 12.698554992675781, + "learning_rate": 7.415506958250499e-07, + "loss": 1.3717, + "step": 751 + }, + { + "epoch": 0.7401574803149606, + "grad_norm": 28.135526657104492, + "learning_rate": 7.425447316103381e-07, + "loss": 2.1774, + "step": 752 + }, + { + "epoch": 0.7411417322834646, + "grad_norm": 11.899801254272461, + "learning_rate": 7.435387673956263e-07, + "loss": 1.0854, + "step": 753 + }, + { + "epoch": 0.7421259842519685, + "grad_norm": 12.33936595916748, + "learning_rate": 7.445328031809146e-07, + "loss": 0.9777, + "step": 754 + }, + { + "epoch": 0.7431102362204725, + "grad_norm": 13.418231010437012, + "learning_rate": 7.455268389662029e-07, + "loss": 1.2721, + "step": 755 + }, + { + "epoch": 0.7440944881889764, + "grad_norm": 10.435457229614258, + "learning_rate": 7.465208747514912e-07, + "loss": 0.7114, + "step": 756 + }, + { + "epoch": 0.7450787401574803, + "grad_norm": 13.19214916229248, + "learning_rate": 7.475149105367794e-07, + "loss": 1.4036, + "step": 757 + }, + { + "epoch": 0.7460629921259843, + "grad_norm": 12.965828895568848, + "learning_rate": 7.485089463220677e-07, + "loss": 1.1742, + "step": 758 + }, + { + "epoch": 0.7470472440944882, + "grad_norm": 14.7980375289917, + "learning_rate": 7.49502982107356e-07, + "loss": 0.9351, + "step": 759 + }, + { + "epoch": 0.7480314960629921, + "grad_norm": 9.050884246826172, + "learning_rate": 7.504970178926442e-07, + "loss": 0.5537, + "step": 760 + }, + { + "epoch": 0.7490157480314961, + "grad_norm": 13.177655220031738, + "learning_rate": 7.514910536779325e-07, + "loss": 0.8688, + "step": 761 + }, + { + "epoch": 0.75, + "grad_norm": 46.27436447143555, + "learning_rate": 7.524850894632208e-07, + "loss": 3.0053, + "step": 762 + }, + { + "epoch": 0.7509842519685039, + "grad_norm": 48.34501647949219, + "learning_rate": 7.53479125248509e-07, + "loss": 3.3743, + "step": 763 + }, + { + "epoch": 0.7519685039370079, + "grad_norm": 26.593921661376953, + "learning_rate": 7.544731610337973e-07, + "loss": 1.9928, + "step": 764 + }, + { + "epoch": 0.7529527559055118, + "grad_norm": 13.116415977478027, + "learning_rate": 7.554671968190856e-07, + "loss": 1.5118, + "step": 765 + }, + { + "epoch": 0.7529527559055118, + "eval_Qnli-dev_cosine_accuracy": 0.666015625, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8375362157821655, + "eval_Qnli-dev_cosine_ap": 0.6688267117653419, + "eval_Qnli-dev_cosine_f1": 0.663594470046083, + "eval_Qnli-dev_cosine_f1_threshold": 0.776394248008728, + "eval_Qnli-dev_cosine_precision": 0.5204819277108433, + "eval_Qnli-dev_cosine_recall": 0.9152542372881356, + "eval_Qnli-dev_dot_accuracy": 0.666015625, + "eval_Qnli-dev_dot_accuracy_threshold": 642.4522705078125, + "eval_Qnli-dev_dot_ap": 0.667969160193219, + "eval_Qnli-dev_dot_f1": 0.6636085626911314, + "eval_Qnli-dev_dot_f1_threshold": 591.8570556640625, + "eval_Qnli-dev_dot_precision": 0.5191387559808612, + "eval_Qnli-dev_dot_recall": 0.9194915254237288, + "eval_Qnli-dev_euclidean_accuracy": 0.66796875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.774551391601562, + "eval_Qnli-dev_euclidean_ap": 0.6688954739628344, + "eval_Qnli-dev_euclidean_f1": 0.662557781201849, + "eval_Qnli-dev_euclidean_f1_threshold": 18.4862060546875, + "eval_Qnli-dev_euclidean_precision": 0.5205811138014528, + "eval_Qnli-dev_euclidean_recall": 0.9110169491525424, + "eval_Qnli-dev_manhattan_accuracy": 0.666015625, + "eval_Qnli-dev_manhattan_accuracy_threshold": 322.94683837890625, + "eval_Qnli-dev_manhattan_ap": 0.6781736723001364, + "eval_Qnli-dev_manhattan_f1": 0.6762360446570973, + "eval_Qnli-dev_manhattan_f1_threshold": 371.5932922363281, + "eval_Qnli-dev_manhattan_precision": 0.5421994884910486, + "eval_Qnli-dev_manhattan_recall": 0.8983050847457628, + "eval_Qnli-dev_max_accuracy": 0.66796875, + "eval_Qnli-dev_max_accuracy_threshold": 642.4522705078125, + "eval_Qnli-dev_max_ap": 0.6781736723001364, + "eval_Qnli-dev_max_f1": 0.6762360446570973, + "eval_Qnli-dev_max_f1_threshold": 591.8570556640625, + "eval_Qnli-dev_max_precision": 0.5421994884910486, + "eval_Qnli-dev_max_recall": 0.9194915254237288, + "eval_allNLI-dev_cosine_accuracy": 0.685546875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9454631209373474, + "eval_allNLI-dev_cosine_ap": 0.4792378857888552, + "eval_allNLI-dev_cosine_f1": 0.529531568228106, + "eval_allNLI-dev_cosine_f1_threshold": 0.881622314453125, + "eval_allNLI-dev_cosine_precision": 0.4088050314465409, + "eval_allNLI-dev_cosine_recall": 0.7514450867052023, + "eval_allNLI-dev_dot_accuracy": 0.685546875, + "eval_allNLI-dev_dot_accuracy_threshold": 726.0874633789062, + "eval_allNLI-dev_dot_ap": 0.4773554395168431, + "eval_allNLI-dev_dot_f1": 0.5287356321839081, + "eval_allNLI-dev_dot_f1_threshold": 667.7103271484375, + "eval_allNLI-dev_dot_precision": 0.3954154727793696, + "eval_allNLI-dev_dot_recall": 0.7976878612716763, + "eval_allNLI-dev_euclidean_accuracy": 0.685546875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.15554428100586, + "eval_allNLI-dev_euclidean_ap": 0.479012627080778, + "eval_allNLI-dev_euclidean_f1": 0.5303643724696356, + "eval_allNLI-dev_euclidean_f1_threshold": 13.481329917907715, + "eval_allNLI-dev_euclidean_precision": 0.40809968847352024, + "eval_allNLI-dev_euclidean_recall": 0.7572254335260116, + "eval_allNLI-dev_manhattan_accuracy": 0.681640625, + "eval_allNLI-dev_manhattan_accuracy_threshold": 179.65867614746094, + "eval_allNLI-dev_manhattan_ap": 0.4774414549549294, + "eval_allNLI-dev_manhattan_f1": 0.5375, + "eval_allNLI-dev_manhattan_f1_threshold": 272.7696838378906, + "eval_allNLI-dev_manhattan_precision": 0.4201954397394137, + "eval_allNLI-dev_manhattan_recall": 0.7456647398843931, + "eval_allNLI-dev_max_accuracy": 0.685546875, + "eval_allNLI-dev_max_accuracy_threshold": 726.0874633789062, + "eval_allNLI-dev_max_ap": 0.4792378857888552, + "eval_allNLI-dev_max_f1": 0.5375, + "eval_allNLI-dev_max_f1_threshold": 667.7103271484375, + "eval_allNLI-dev_max_precision": 0.4201954397394137, + "eval_allNLI-dev_max_recall": 0.7976878612716763, + "eval_loss": 0.9341774582862854, + "eval_runtime": 50.0311, + "eval_samples_per_second": 33.259, + "eval_sequential_score": 0.6781736723001364, + "eval_steps_per_second": 0.14, + "eval_sts-test_pearson_cosine": 0.4008630217130614, + "eval_sts-test_pearson_dot": 0.39680226288992326, + "eval_sts-test_pearson_euclidean": 0.42558645063013534, + "eval_sts-test_pearson_manhattan": 0.43690206761360884, + "eval_sts-test_pearson_max": 0.43690206761360884, + "eval_sts-test_spearman_cosine": 0.45136395222060033, + "eval_sts-test_spearman_dot": 0.4467416839340055, + "eval_sts-test_spearman_euclidean": 0.45137661682757413, + "eval_sts-test_spearman_manhattan": 0.4609070244585321, + "eval_sts-test_spearman_max": 0.4609070244585321, + "step": 765 + }, + { + "epoch": 0.7539370078740157, + "grad_norm": 12.811962127685547, + "learning_rate": 7.56461232604374e-07, + "loss": 1.1213, + "step": 766 + }, + { + "epoch": 0.7549212598425197, + "grad_norm": 28.99581527709961, + "learning_rate": 7.574552683896621e-07, + "loss": 2.1312, + "step": 767 + }, + { + "epoch": 0.7559055118110236, + "grad_norm": 14.349875450134277, + "learning_rate": 7.584493041749503e-07, + "loss": 1.3739, + "step": 768 + }, + { + "epoch": 0.7568897637795275, + "grad_norm": 11.058197021484375, + "learning_rate": 7.594433399602388e-07, + "loss": 0.8819, + "step": 769 + }, + { + "epoch": 0.7578740157480315, + "grad_norm": 10.843620300292969, + "learning_rate": 7.60437375745527e-07, + "loss": 0.9069, + "step": 770 + }, + { + "epoch": 0.7588582677165354, + "grad_norm": 10.030068397521973, + "learning_rate": 7.614314115308152e-07, + "loss": 0.935, + "step": 771 + }, + { + "epoch": 0.7598425196850394, + "grad_norm": 9.304388999938965, + "learning_rate": 7.624254473161036e-07, + "loss": 0.7874, + "step": 772 + }, + { + "epoch": 0.7608267716535433, + "grad_norm": 25.917118072509766, + "learning_rate": 7.634194831013918e-07, + "loss": 1.9942, + "step": 773 + }, + { + "epoch": 0.7618110236220472, + "grad_norm": 11.599132537841797, + "learning_rate": 7.644135188866801e-07, + "loss": 1.1711, + "step": 774 + }, + { + "epoch": 0.7627952755905512, + "grad_norm": 10.600054740905762, + "learning_rate": 7.654075546719683e-07, + "loss": 0.8407, + "step": 775 + }, + { + "epoch": 0.7637795275590551, + "grad_norm": 14.485662460327148, + "learning_rate": 7.664015904572566e-07, + "loss": 1.5171, + "step": 776 + }, + { + "epoch": 0.764763779527559, + "grad_norm": 8.246170997619629, + "learning_rate": 7.673956262425449e-07, + "loss": 0.5308, + "step": 777 + }, + { + "epoch": 0.765748031496063, + "grad_norm": 13.491275787353516, + "learning_rate": 7.683896620278331e-07, + "loss": 1.4107, + "step": 778 + }, + { + "epoch": 0.7667322834645669, + "grad_norm": 11.929235458374023, + "learning_rate": 7.693836978131215e-07, + "loss": 1.1766, + "step": 779 + }, + { + "epoch": 0.7677165354330708, + "grad_norm": 12.884323120117188, + "learning_rate": 7.703777335984097e-07, + "loss": 1.326, + "step": 780 + }, + { + "epoch": 0.7687007874015748, + "grad_norm": 10.983901023864746, + "learning_rate": 7.713717693836979e-07, + "loss": 0.7371, + "step": 781 + }, + { + "epoch": 0.7696850393700787, + "grad_norm": 12.778090476989746, + "learning_rate": 7.723658051689862e-07, + "loss": 1.0504, + "step": 782 + }, + { + "epoch": 0.7706692913385826, + "grad_norm": 12.948322296142578, + "learning_rate": 7.733598409542745e-07, + "loss": 1.1458, + "step": 783 + }, + { + "epoch": 0.7716535433070866, + "grad_norm": 9.125856399536133, + "learning_rate": 7.743538767395627e-07, + "loss": 0.7242, + "step": 784 + }, + { + "epoch": 0.7726377952755905, + "grad_norm": 10.288677215576172, + "learning_rate": 7.75347912524851e-07, + "loss": 0.8113, + "step": 785 + }, + { + "epoch": 0.7736220472440944, + "grad_norm": 14.06745433807373, + "learning_rate": 7.763419483101393e-07, + "loss": 1.3808, + "step": 786 + }, + { + "epoch": 0.7746062992125984, + "grad_norm": 9.976495742797852, + "learning_rate": 7.773359840954276e-07, + "loss": 0.7584, + "step": 787 + }, + { + "epoch": 0.7755905511811023, + "grad_norm": 13.068366050720215, + "learning_rate": 7.783300198807158e-07, + "loss": 1.226, + "step": 788 + }, + { + "epoch": 0.7765748031496063, + "grad_norm": 11.98946475982666, + "learning_rate": 7.79324055666004e-07, + "loss": 1.0599, + "step": 789 + }, + { + "epoch": 0.7775590551181102, + "grad_norm": 37.414520263671875, + "learning_rate": 7.803180914512924e-07, + "loss": 2.9348, + "step": 790 + }, + { + "epoch": 0.7785433070866141, + "grad_norm": 11.82547378540039, + "learning_rate": 7.813121272365806e-07, + "loss": 1.0849, + "step": 791 + }, + { + "epoch": 0.7795275590551181, + "grad_norm": 9.449908256530762, + "learning_rate": 7.823061630218689e-07, + "loss": 0.5362, + "step": 792 + }, + { + "epoch": 0.780511811023622, + "grad_norm": 14.662030220031738, + "learning_rate": 7.833001988071572e-07, + "loss": 1.3765, + "step": 793 + }, + { + "epoch": 0.781496062992126, + "grad_norm": 10.446459770202637, + "learning_rate": 7.842942345924454e-07, + "loss": 0.6824, + "step": 794 + }, + { + "epoch": 0.78248031496063, + "grad_norm": 9.619561195373535, + "learning_rate": 7.852882703777337e-07, + "loss": 0.6009, + "step": 795 + }, + { + "epoch": 0.7834645669291339, + "grad_norm": 29.211288452148438, + "learning_rate": 7.862823061630219e-07, + "loss": 2.3853, + "step": 796 + }, + { + "epoch": 0.7844488188976378, + "grad_norm": 12.475735664367676, + "learning_rate": 7.872763419483102e-07, + "loss": 1.0571, + "step": 797 + }, + { + "epoch": 0.7854330708661418, + "grad_norm": 12.86826229095459, + "learning_rate": 7.882703777335986e-07, + "loss": 0.9172, + "step": 798 + }, + { + "epoch": 0.7864173228346457, + "grad_norm": 10.496902465820312, + "learning_rate": 7.892644135188868e-07, + "loss": 0.7915, + "step": 799 + }, + { + "epoch": 0.7874015748031497, + "grad_norm": 10.799955368041992, + "learning_rate": 7.902584493041752e-07, + "loss": 0.827, + "step": 800 + }, + { + "epoch": 0.7883858267716536, + "grad_norm": 10.648172378540039, + "learning_rate": 7.912524850894634e-07, + "loss": 0.8465, + "step": 801 + }, + { + "epoch": 0.7893700787401575, + "grad_norm": 25.809648513793945, + "learning_rate": 7.922465208747516e-07, + "loss": 2.3489, + "step": 802 + }, + { + "epoch": 0.7903543307086615, + "grad_norm": 9.341686248779297, + "learning_rate": 7.932405566600399e-07, + "loss": 0.6506, + "step": 803 + }, + { + "epoch": 0.7913385826771654, + "grad_norm": 11.539505004882812, + "learning_rate": 7.942345924453282e-07, + "loss": 0.8346, + "step": 804 + }, + { + "epoch": 0.7923228346456693, + "grad_norm": 9.639948844909668, + "learning_rate": 7.952286282306165e-07, + "loss": 0.6249, + "step": 805 + }, + { + "epoch": 0.7933070866141733, + "grad_norm": 14.152978897094727, + "learning_rate": 7.962226640159047e-07, + "loss": 1.0557, + "step": 806 + }, + { + "epoch": 0.7942913385826772, + "grad_norm": 10.899728775024414, + "learning_rate": 7.97216699801193e-07, + "loss": 0.7552, + "step": 807 + }, + { + "epoch": 0.7952755905511811, + "grad_norm": 13.07314682006836, + "learning_rate": 7.982107355864813e-07, + "loss": 1.281, + "step": 808 + }, + { + "epoch": 0.7962598425196851, + "grad_norm": 10.976669311523438, + "learning_rate": 7.992047713717695e-07, + "loss": 0.7846, + "step": 809 + }, + { + "epoch": 0.797244094488189, + "grad_norm": 31.689002990722656, + "learning_rate": 8.001988071570577e-07, + "loss": 2.6403, + "step": 810 + }, + { + "epoch": 0.7982283464566929, + "grad_norm": 7.491154193878174, + "learning_rate": 8.011928429423461e-07, + "loss": 0.3679, + "step": 811 + }, + { + "epoch": 0.7992125984251969, + "grad_norm": 25.290111541748047, + "learning_rate": 8.021868787276343e-07, + "loss": 1.9118, + "step": 812 + }, + { + "epoch": 0.8001968503937008, + "grad_norm": 29.633962631225586, + "learning_rate": 8.031809145129226e-07, + "loss": 2.5911, + "step": 813 + }, + { + "epoch": 0.8011811023622047, + "grad_norm": 13.148744583129883, + "learning_rate": 8.041749502982109e-07, + "loss": 1.1783, + "step": 814 + }, + { + "epoch": 0.8021653543307087, + "grad_norm": 12.22188949584961, + "learning_rate": 8.051689860834991e-07, + "loss": 0.9347, + "step": 815 + }, + { + "epoch": 0.8031496062992126, + "grad_norm": 8.539156913757324, + "learning_rate": 8.061630218687874e-07, + "loss": 0.5311, + "step": 816 + }, + { + "epoch": 0.8041338582677166, + "grad_norm": 10.084023475646973, + "learning_rate": 8.071570576540756e-07, + "loss": 0.7092, + "step": 817 + }, + { + "epoch": 0.8051181102362205, + "grad_norm": 9.478177070617676, + "learning_rate": 8.08151093439364e-07, + "loss": 0.8384, + "step": 818 + }, + { + "epoch": 0.8061023622047244, + "grad_norm": 9.305712699890137, + "learning_rate": 8.091451292246522e-07, + "loss": 0.514, + "step": 819 + }, + { + "epoch": 0.8070866141732284, + "grad_norm": 7.318828105926514, + "learning_rate": 8.101391650099404e-07, + "loss": 0.3638, + "step": 820 + }, + { + "epoch": 0.8080708661417323, + "grad_norm": 25.613597869873047, + "learning_rate": 8.111332007952288e-07, + "loss": 1.9376, + "step": 821 + }, + { + "epoch": 0.8090551181102362, + "grad_norm": 12.352375030517578, + "learning_rate": 8.12127236580517e-07, + "loss": 0.9177, + "step": 822 + }, + { + "epoch": 0.8100393700787402, + "grad_norm": 11.09453010559082, + "learning_rate": 8.131212723658052e-07, + "loss": 0.8293, + "step": 823 + }, + { + "epoch": 0.8110236220472441, + "grad_norm": 10.947206497192383, + "learning_rate": 8.141153081510935e-07, + "loss": 0.7269, + "step": 824 + }, + { + "epoch": 0.812007874015748, + "grad_norm": 10.038667678833008, + "learning_rate": 8.151093439363818e-07, + "loss": 0.664, + "step": 825 + }, + { + "epoch": 0.812992125984252, + "grad_norm": 9.964850425720215, + "learning_rate": 8.161033797216701e-07, + "loss": 0.6205, + "step": 826 + }, + { + "epoch": 0.8139763779527559, + "grad_norm": 9.162919044494629, + "learning_rate": 8.170974155069583e-07, + "loss": 0.6562, + "step": 827 + }, + { + "epoch": 0.8149606299212598, + "grad_norm": 9.911048889160156, + "learning_rate": 8.180914512922467e-07, + "loss": 0.6576, + "step": 828 + }, + { + "epoch": 0.8159448818897638, + "grad_norm": 13.729418754577637, + "learning_rate": 8.19085487077535e-07, + "loss": 0.9931, + "step": 829 + }, + { + "epoch": 0.8169291338582677, + "grad_norm": 13.092864036560059, + "learning_rate": 8.200795228628232e-07, + "loss": 1.1707, + "step": 830 + }, + { + "epoch": 0.8179133858267716, + "grad_norm": 11.316834449768066, + "learning_rate": 8.210735586481115e-07, + "loss": 0.8635, + "step": 831 + }, + { + "epoch": 0.8188976377952756, + "grad_norm": 10.211715698242188, + "learning_rate": 8.220675944333998e-07, + "loss": 0.7274, + "step": 832 + }, + { + "epoch": 0.8198818897637795, + "grad_norm": 24.996658325195312, + "learning_rate": 8.23061630218688e-07, + "loss": 1.6808, + "step": 833 + }, + { + "epoch": 0.8208661417322834, + "grad_norm": 28.608722686767578, + "learning_rate": 8.240556660039763e-07, + "loss": 1.8309, + "step": 834 + }, + { + "epoch": 0.8218503937007874, + "grad_norm": 9.821938514709473, + "learning_rate": 8.250497017892646e-07, + "loss": 0.6191, + "step": 835 + }, + { + "epoch": 0.8228346456692913, + "grad_norm": 12.319458961486816, + "learning_rate": 8.260437375745529e-07, + "loss": 1.0789, + "step": 836 + }, + { + "epoch": 0.8238188976377953, + "grad_norm": 14.616786003112793, + "learning_rate": 8.270377733598411e-07, + "loss": 1.1637, + "step": 837 + }, + { + "epoch": 0.8248031496062992, + "grad_norm": 10.99319076538086, + "learning_rate": 8.280318091451293e-07, + "loss": 0.7813, + "step": 838 + }, + { + "epoch": 0.8257874015748031, + "grad_norm": 12.11503791809082, + "learning_rate": 8.290258449304177e-07, + "loss": 1.0403, + "step": 839 + }, + { + "epoch": 0.8267716535433071, + "grad_norm": 10.753040313720703, + "learning_rate": 8.300198807157059e-07, + "loss": 0.7656, + "step": 840 + }, + { + "epoch": 0.827755905511811, + "grad_norm": 12.77871036529541, + "learning_rate": 8.310139165009941e-07, + "loss": 0.9994, + "step": 841 + }, + { + "epoch": 0.8287401574803149, + "grad_norm": 10.987958908081055, + "learning_rate": 8.320079522862825e-07, + "loss": 1.009, + "step": 842 + }, + { + "epoch": 0.8297244094488189, + "grad_norm": 10.836406707763672, + "learning_rate": 8.330019880715707e-07, + "loss": 0.8585, + "step": 843 + }, + { + "epoch": 0.8307086614173228, + "grad_norm": 11.661853790283203, + "learning_rate": 8.33996023856859e-07, + "loss": 0.8847, + "step": 844 + }, + { + "epoch": 0.8316929133858267, + "grad_norm": 10.705239295959473, + "learning_rate": 8.349900596421472e-07, + "loss": 0.8321, + "step": 845 + }, + { + "epoch": 0.8326771653543307, + "grad_norm": 12.334227561950684, + "learning_rate": 8.359840954274355e-07, + "loss": 1.2605, + "step": 846 + }, + { + "epoch": 0.8336614173228346, + "grad_norm": 11.797635078430176, + "learning_rate": 8.369781312127238e-07, + "loss": 1.0609, + "step": 847 + }, + { + "epoch": 0.8346456692913385, + "grad_norm": 24.433855056762695, + "learning_rate": 8.37972166998012e-07, + "loss": 2.0115, + "step": 848 + }, + { + "epoch": 0.8356299212598425, + "grad_norm": 20.614023208618164, + "learning_rate": 8.389662027833004e-07, + "loss": 1.2952, + "step": 849 + }, + { + "epoch": 0.8366141732283464, + "grad_norm": 10.158178329467773, + "learning_rate": 8.399602385685886e-07, + "loss": 0.6999, + "step": 850 + }, + { + "epoch": 0.8375984251968503, + "grad_norm": 10.152667999267578, + "learning_rate": 8.409542743538768e-07, + "loss": 0.7006, + "step": 851 + }, + { + "epoch": 0.8385826771653543, + "grad_norm": 10.248065948486328, + "learning_rate": 8.419483101391651e-07, + "loss": 0.927, + "step": 852 + }, + { + "epoch": 0.8395669291338582, + "grad_norm": 18.727346420288086, + "learning_rate": 8.429423459244534e-07, + "loss": 1.2083, + "step": 853 + }, + { + "epoch": 0.8405511811023622, + "grad_norm": 10.210419654846191, + "learning_rate": 8.439363817097416e-07, + "loss": 0.608, + "step": 854 + }, + { + "epoch": 0.8415354330708661, + "grad_norm": 11.704425811767578, + "learning_rate": 8.449304174950299e-07, + "loss": 0.8478, + "step": 855 + }, + { + "epoch": 0.84251968503937, + "grad_norm": 15.504826545715332, + "learning_rate": 8.459244532803182e-07, + "loss": 1.5731, + "step": 856 + }, + { + "epoch": 0.843503937007874, + "grad_norm": 26.414409637451172, + "learning_rate": 8.469184890656066e-07, + "loss": 1.6353, + "step": 857 + }, + { + "epoch": 0.844488188976378, + "grad_norm": 9.277349472045898, + "learning_rate": 8.479125248508948e-07, + "loss": 0.7862, + "step": 858 + }, + { + "epoch": 0.8454724409448819, + "grad_norm": 11.632246017456055, + "learning_rate": 8.48906560636183e-07, + "loss": 0.8909, + "step": 859 + }, + { + "epoch": 0.8464566929133859, + "grad_norm": 12.540164947509766, + "learning_rate": 8.499005964214714e-07, + "loss": 1.1719, + "step": 860 + }, + { + "epoch": 0.8474409448818898, + "grad_norm": 12.841619491577148, + "learning_rate": 8.508946322067596e-07, + "loss": 1.2722, + "step": 861 + }, + { + "epoch": 0.8484251968503937, + "grad_norm": 11.645380973815918, + "learning_rate": 8.518886679920479e-07, + "loss": 1.0022, + "step": 862 + }, + { + "epoch": 0.8494094488188977, + "grad_norm": 15.473668098449707, + "learning_rate": 8.528827037773362e-07, + "loss": 1.5307, + "step": 863 + }, + { + "epoch": 0.8503937007874016, + "grad_norm": 12.68653392791748, + "learning_rate": 8.538767395626244e-07, + "loss": 1.0162, + "step": 864 + }, + { + "epoch": 0.8513779527559056, + "grad_norm": 11.079401016235352, + "learning_rate": 8.548707753479127e-07, + "loss": 0.6827, + "step": 865 + }, + { + "epoch": 0.8523622047244095, + "grad_norm": 10.115388870239258, + "learning_rate": 8.558648111332009e-07, + "loss": 0.7744, + "step": 866 + }, + { + "epoch": 0.8533464566929134, + "grad_norm": 12.911059379577637, + "learning_rate": 8.568588469184892e-07, + "loss": 1.2011, + "step": 867 + }, + { + "epoch": 0.8543307086614174, + "grad_norm": 11.86655330657959, + "learning_rate": 8.578528827037775e-07, + "loss": 0.9219, + "step": 868 + }, + { + "epoch": 0.8553149606299213, + "grad_norm": 10.42223834991455, + "learning_rate": 8.588469184890657e-07, + "loss": 0.7636, + "step": 869 + }, + { + "epoch": 0.8562992125984252, + "grad_norm": 14.084538459777832, + "learning_rate": 8.598409542743541e-07, + "loss": 1.5061, + "step": 870 + }, + { + "epoch": 0.8572834645669292, + "grad_norm": 12.389713287353516, + "learning_rate": 8.608349900596423e-07, + "loss": 1.5569, + "step": 871 + }, + { + "epoch": 0.8582677165354331, + "grad_norm": 10.80617618560791, + "learning_rate": 8.618290258449305e-07, + "loss": 0.5896, + "step": 872 + }, + { + "epoch": 0.859251968503937, + "grad_norm": 12.295326232910156, + "learning_rate": 8.628230616302189e-07, + "loss": 1.1918, + "step": 873 + }, + { + "epoch": 0.860236220472441, + "grad_norm": 10.111541748046875, + "learning_rate": 8.638170974155071e-07, + "loss": 0.8572, + "step": 874 + }, + { + "epoch": 0.8612204724409449, + "grad_norm": 13.957104682922363, + "learning_rate": 8.648111332007954e-07, + "loss": 1.0421, + "step": 875 + }, + { + "epoch": 0.8622047244094488, + "grad_norm": 28.831130981445312, + "learning_rate": 8.658051689860836e-07, + "loss": 2.4599, + "step": 876 + }, + { + "epoch": 0.8631889763779528, + "grad_norm": 8.383614540100098, + "learning_rate": 8.667992047713719e-07, + "loss": 0.55, + "step": 877 + }, + { + "epoch": 0.8641732283464567, + "grad_norm": 22.71127700805664, + "learning_rate": 8.677932405566602e-07, + "loss": 1.2829, + "step": 878 + }, + { + "epoch": 0.8651574803149606, + "grad_norm": 10.82112979888916, + "learning_rate": 8.687872763419484e-07, + "loss": 0.7808, + "step": 879 + }, + { + "epoch": 0.8661417322834646, + "grad_norm": 20.141550064086914, + "learning_rate": 8.697813121272366e-07, + "loss": 1.7712, + "step": 880 + }, + { + "epoch": 0.8671259842519685, + "grad_norm": 11.213445663452148, + "learning_rate": 8.70775347912525e-07, + "loss": 0.7456, + "step": 881 + }, + { + "epoch": 0.8681102362204725, + "grad_norm": 14.254484176635742, + "learning_rate": 8.717693836978132e-07, + "loss": 1.2805, + "step": 882 + }, + { + "epoch": 0.8690944881889764, + "grad_norm": 24.12900733947754, + "learning_rate": 8.727634194831015e-07, + "loss": 2.1927, + "step": 883 + }, + { + "epoch": 0.8700787401574803, + "grad_norm": 10.013153076171875, + "learning_rate": 8.737574552683898e-07, + "loss": 0.855, + "step": 884 + }, + { + "epoch": 0.8710629921259843, + "grad_norm": 10.102697372436523, + "learning_rate": 8.74751491053678e-07, + "loss": 0.667, + "step": 885 + }, + { + "epoch": 0.8720472440944882, + "grad_norm": 14.909455299377441, + "learning_rate": 8.757455268389663e-07, + "loss": 1.1097, + "step": 886 + }, + { + "epoch": 0.8730314960629921, + "grad_norm": 23.29335594177246, + "learning_rate": 8.767395626242545e-07, + "loss": 1.8795, + "step": 887 + }, + { + "epoch": 0.8740157480314961, + "grad_norm": 11.177861213684082, + "learning_rate": 8.77733598409543e-07, + "loss": 0.6767, + "step": 888 + }, + { + "epoch": 0.875, + "grad_norm": 10.940074920654297, + "learning_rate": 8.787276341948312e-07, + "loss": 0.7549, + "step": 889 + }, + { + "epoch": 0.8759842519685039, + "grad_norm": 10.625170707702637, + "learning_rate": 8.797216699801194e-07, + "loss": 0.8616, + "step": 890 + }, + { + "epoch": 0.8769685039370079, + "grad_norm": 27.984130859375, + "learning_rate": 8.807157057654078e-07, + "loss": 1.9461, + "step": 891 + }, + { + "epoch": 0.8779527559055118, + "grad_norm": 12.983888626098633, + "learning_rate": 8.81709741550696e-07, + "loss": 1.2694, + "step": 892 + }, + { + "epoch": 0.8789370078740157, + "grad_norm": 17.34275245666504, + "learning_rate": 8.827037773359842e-07, + "loss": 1.825, + "step": 893 + }, + { + "epoch": 0.8799212598425197, + "grad_norm": 11.462705612182617, + "learning_rate": 8.836978131212726e-07, + "loss": 0.9218, + "step": 894 + }, + { + "epoch": 0.8809055118110236, + "grad_norm": 12.501668930053711, + "learning_rate": 8.846918489065608e-07, + "loss": 1.0297, + "step": 895 + }, + { + "epoch": 0.8818897637795275, + "grad_norm": 9.415837287902832, + "learning_rate": 8.856858846918491e-07, + "loss": 0.609, + "step": 896 + }, + { + "epoch": 0.8828740157480315, + "grad_norm": 11.89690113067627, + "learning_rate": 8.866799204771373e-07, + "loss": 0.9638, + "step": 897 + }, + { + "epoch": 0.8838582677165354, + "grad_norm": 10.2057523727417, + "learning_rate": 8.876739562624256e-07, + "loss": 0.5521, + "step": 898 + }, + { + "epoch": 0.8848425196850394, + "grad_norm": 14.843278884887695, + "learning_rate": 8.886679920477139e-07, + "loss": 1.3365, + "step": 899 + }, + { + "epoch": 0.8858267716535433, + "grad_norm": 12.098006248474121, + "learning_rate": 8.896620278330021e-07, + "loss": 0.8443, + "step": 900 + }, + { + "epoch": 0.8868110236220472, + "grad_norm": 9.75728988647461, + "learning_rate": 8.906560636182905e-07, + "loss": 0.7848, + "step": 901 + }, + { + "epoch": 0.8877952755905512, + "grad_norm": 11.828933715820312, + "learning_rate": 8.916500994035787e-07, + "loss": 1.0733, + "step": 902 + }, + { + "epoch": 0.8887795275590551, + "grad_norm": 10.241612434387207, + "learning_rate": 8.926441351888669e-07, + "loss": 0.5657, + "step": 903 + }, + { + "epoch": 0.889763779527559, + "grad_norm": 18.475008010864258, + "learning_rate": 8.936381709741552e-07, + "loss": 1.8081, + "step": 904 + }, + { + "epoch": 0.890748031496063, + "grad_norm": 11.45572566986084, + "learning_rate": 8.946322067594435e-07, + "loss": 0.8232, + "step": 905 + }, + { + "epoch": 0.8917322834645669, + "grad_norm": 10.6580810546875, + "learning_rate": 8.956262425447318e-07, + "loss": 0.6159, + "step": 906 + }, + { + "epoch": 0.8927165354330708, + "grad_norm": 12.259444236755371, + "learning_rate": 8.9662027833002e-07, + "loss": 0.9832, + "step": 907 + }, + { + "epoch": 0.8937007874015748, + "grad_norm": 20.447729110717773, + "learning_rate": 8.976143141153082e-07, + "loss": 1.1375, + "step": 908 + }, + { + "epoch": 0.8946850393700787, + "grad_norm": 16.43317413330078, + "learning_rate": 8.986083499005966e-07, + "loss": 1.4182, + "step": 909 + }, + { + "epoch": 0.8956692913385826, + "grad_norm": 14.764389991760254, + "learning_rate": 8.996023856858848e-07, + "loss": 1.2287, + "step": 910 + }, + { + "epoch": 0.8966535433070866, + "grad_norm": 13.386956214904785, + "learning_rate": 9.00596421471173e-07, + "loss": 1.0915, + "step": 911 + }, + { + "epoch": 0.8976377952755905, + "grad_norm": 11.347896575927734, + "learning_rate": 9.015904572564614e-07, + "loss": 0.8116, + "step": 912 + }, + { + "epoch": 0.8986220472440944, + "grad_norm": 10.63943862915039, + "learning_rate": 9.025844930417496e-07, + "loss": 0.6824, + "step": 913 + }, + { + "epoch": 0.8996062992125984, + "grad_norm": 10.412869453430176, + "learning_rate": 9.035785288270379e-07, + "loss": 0.8888, + "step": 914 + }, + { + "epoch": 0.9005905511811023, + "grad_norm": 9.667635917663574, + "learning_rate": 9.045725646123262e-07, + "loss": 0.5974, + "step": 915 + }, + { + "epoch": 0.9015748031496063, + "grad_norm": 12.988663673400879, + "learning_rate": 9.055666003976144e-07, + "loss": 1.1766, + "step": 916 + }, + { + "epoch": 0.9025590551181102, + "grad_norm": 10.166474342346191, + "learning_rate": 9.065606361829028e-07, + "loss": 0.9415, + "step": 917 + }, + { + "epoch": 0.9035433070866141, + "grad_norm": 9.530787467956543, + "learning_rate": 9.07554671968191e-07, + "loss": 0.6387, + "step": 918 + }, + { + "epoch": 0.9035433070866141, + "eval_Qnli-dev_cosine_accuracy": 0.669921875, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8597075343132019, + "eval_Qnli-dev_cosine_ap": 0.6901424139844872, + "eval_Qnli-dev_cosine_f1": 0.6784565916398714, + "eval_Qnli-dev_cosine_f1_threshold": 0.7796358466148376, + "eval_Qnli-dev_cosine_precision": 0.5466321243523317, + "eval_Qnli-dev_cosine_recall": 0.8940677966101694, + "eval_Qnli-dev_dot_accuracy": 0.671875, + "eval_Qnli-dev_dot_accuracy_threshold": 652.4218139648438, + "eval_Qnli-dev_dot_ap": 0.6875184110680547, + "eval_Qnli-dev_dot_f1": 0.678688524590164, + "eval_Qnli-dev_dot_f1_threshold": 597.8012084960938, + "eval_Qnli-dev_dot_precision": 0.553475935828877, + "eval_Qnli-dev_dot_recall": 0.8771186440677966, + "eval_Qnli-dev_euclidean_accuracy": 0.669921875, + "eval_Qnli-dev_euclidean_accuracy_threshold": 14.654603004455566, + "eval_Qnli-dev_euclidean_ap": 0.6902963827034869, + "eval_Qnli-dev_euclidean_f1": 0.6781789638932496, + "eval_Qnli-dev_euclidean_f1_threshold": 18.631183624267578, + "eval_Qnli-dev_euclidean_precision": 0.5386533665835411, + "eval_Qnli-dev_euclidean_recall": 0.9152542372881356, + "eval_Qnli-dev_manhattan_accuracy": 0.677734375, + "eval_Qnli-dev_manhattan_accuracy_threshold": 298.08416748046875, + "eval_Qnli-dev_manhattan_ap": 0.6933835859762669, + "eval_Qnli-dev_manhattan_f1": 0.6864686468646866, + "eval_Qnli-dev_manhattan_f1_threshold": 372.546142578125, + "eval_Qnli-dev_manhattan_precision": 0.5621621621621622, + "eval_Qnli-dev_manhattan_recall": 0.8813559322033898, + "eval_Qnli-dev_max_accuracy": 0.677734375, + "eval_Qnli-dev_max_accuracy_threshold": 652.4218139648438, + "eval_Qnli-dev_max_ap": 0.6933835859762669, + "eval_Qnli-dev_max_f1": 0.6864686468646866, + "eval_Qnli-dev_max_f1_threshold": 597.8012084960938, + "eval_Qnli-dev_max_precision": 0.5621621621621622, + "eval_Qnli-dev_max_recall": 0.9152542372881356, + "eval_allNLI-dev_cosine_accuracy": 0.68359375, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.943246603012085, + "eval_allNLI-dev_cosine_ap": 0.4774083935976875, + "eval_allNLI-dev_cosine_f1": 0.5239179954441913, + "eval_allNLI-dev_cosine_f1_threshold": 0.8804351091384888, + "eval_allNLI-dev_cosine_precision": 0.4323308270676692, + "eval_allNLI-dev_cosine_recall": 0.6647398843930635, + "eval_allNLI-dev_dot_accuracy": 0.6875, + "eval_allNLI-dev_dot_accuracy_threshold": 717.8801879882812, + "eval_allNLI-dev_dot_ap": 0.4770885052177509, + "eval_allNLI-dev_dot_f1": 0.5245098039215687, + "eval_allNLI-dev_dot_f1_threshold": 679.1788330078125, + "eval_allNLI-dev_dot_precision": 0.4553191489361702, + "eval_allNLI-dev_dot_recall": 0.6184971098265896, + "eval_allNLI-dev_euclidean_accuracy": 0.68359375, + "eval_allNLI-dev_euclidean_accuracy_threshold": 9.316292762756348, + "eval_allNLI-dev_euclidean_ap": 0.4772019601787133, + "eval_allNLI-dev_euclidean_f1": 0.5239179954441913, + "eval_allNLI-dev_euclidean_f1_threshold": 13.514263153076172, + "eval_allNLI-dev_euclidean_precision": 0.4323308270676692, + "eval_allNLI-dev_euclidean_recall": 0.6647398843930635, + "eval_allNLI-dev_manhattan_accuracy": 0.6875, + "eval_allNLI-dev_manhattan_accuracy_threshold": 198.4757843017578, + "eval_allNLI-dev_manhattan_ap": 0.48352872945893066, + "eval_allNLI-dev_manhattan_f1": 0.5378973105134474, + "eval_allNLI-dev_manhattan_f1_threshold": 268.8349609375, + "eval_allNLI-dev_manhattan_precision": 0.4661016949152542, + "eval_allNLI-dev_manhattan_recall": 0.6358381502890174, + "eval_allNLI-dev_max_accuracy": 0.6875, + "eval_allNLI-dev_max_accuracy_threshold": 717.8801879882812, + "eval_allNLI-dev_max_ap": 0.48352872945893066, + "eval_allNLI-dev_max_f1": 0.5378973105134474, + "eval_allNLI-dev_max_f1_threshold": 679.1788330078125, + "eval_allNLI-dev_max_precision": 0.4661016949152542, + "eval_allNLI-dev_max_recall": 0.6647398843930635, + "eval_loss": 0.7856447100639343, + "eval_runtime": 50.2067, + "eval_samples_per_second": 33.143, + "eval_sequential_score": 0.6933835859762669, + "eval_steps_per_second": 0.139, + "eval_sts-test_pearson_cosine": 0.4755590947571911, + "eval_sts-test_pearson_dot": 0.47017549814130327, + "eval_sts-test_pearson_euclidean": 0.497277223110132, + "eval_sts-test_pearson_manhattan": 0.5069429381583643, + "eval_sts-test_pearson_max": 0.5069429381583643, + "eval_sts-test_spearman_cosine": 0.5146741882286727, + "eval_sts-test_spearman_dot": 0.5090295326243299, + "eval_sts-test_spearman_euclidean": 0.5151112290474211, + "eval_sts-test_spearman_manhattan": 0.5226267558854928, + "eval_sts-test_spearman_max": 0.5226267558854928, + "step": 918 + }, + { + "epoch": 0.9045275590551181, + "grad_norm": 11.6918363571167, + "learning_rate": 9.085487077534794e-07, + "loss": 0.7342, + "step": 919 + }, + { + "epoch": 0.905511811023622, + "grad_norm": 14.327659606933594, + "learning_rate": 9.095427435387676e-07, + "loss": 1.2232, + "step": 920 + }, + { + "epoch": 0.906496062992126, + "grad_norm": 21.73860740661621, + "learning_rate": 9.105367793240558e-07, + "loss": 1.4883, + "step": 921 + }, + { + "epoch": 0.90748031496063, + "grad_norm": 26.939565658569336, + "learning_rate": 9.115308151093442e-07, + "loss": 1.4453, + "step": 922 + }, + { + "epoch": 0.9084645669291339, + "grad_norm": 9.518061637878418, + "learning_rate": 9.125248508946324e-07, + "loss": 0.665, + "step": 923 + }, + { + "epoch": 0.9094488188976378, + "grad_norm": 11.462472915649414, + "learning_rate": 9.135188866799206e-07, + "loss": 0.8973, + "step": 924 + }, + { + "epoch": 0.9104330708661418, + "grad_norm": 10.186049461364746, + "learning_rate": 9.145129224652089e-07, + "loss": 0.7578, + "step": 925 + }, + { + "epoch": 0.9114173228346457, + "grad_norm": 11.482391357421875, + "learning_rate": 9.155069582504972e-07, + "loss": 0.8693, + "step": 926 + }, + { + "epoch": 0.9124015748031497, + "grad_norm": 13.015597343444824, + "learning_rate": 9.165009940357855e-07, + "loss": 1.0055, + "step": 927 + }, + { + "epoch": 0.9133858267716536, + "grad_norm": 9.612940788269043, + "learning_rate": 9.174950298210737e-07, + "loss": 0.4451, + "step": 928 + }, + { + "epoch": 0.9143700787401575, + "grad_norm": 22.400720596313477, + "learning_rate": 9.18489065606362e-07, + "loss": 1.3435, + "step": 929 + }, + { + "epoch": 0.9153543307086615, + "grad_norm": 14.427326202392578, + "learning_rate": 9.194831013916503e-07, + "loss": 1.0979, + "step": 930 + }, + { + "epoch": 0.9163385826771654, + "grad_norm": 12.939393043518066, + "learning_rate": 9.204771371769385e-07, + "loss": 1.0552, + "step": 931 + }, + { + "epoch": 0.9173228346456693, + "grad_norm": 13.793437004089355, + "learning_rate": 9.214711729622268e-07, + "loss": 0.8224, + "step": 932 + }, + { + "epoch": 0.9183070866141733, + "grad_norm": 37.37275314331055, + "learning_rate": 9.224652087475151e-07, + "loss": 2.824, + "step": 933 + }, + { + "epoch": 0.9192913385826772, + "grad_norm": 15.605339050292969, + "learning_rate": 9.234592445328033e-07, + "loss": 1.3514, + "step": 934 + }, + { + "epoch": 0.9202755905511811, + "grad_norm": 14.879430770874023, + "learning_rate": 9.244532803180916e-07, + "loss": 1.3339, + "step": 935 + }, + { + "epoch": 0.9212598425196851, + "grad_norm": 10.399064064025879, + "learning_rate": 9.254473161033799e-07, + "loss": 0.8439, + "step": 936 + }, + { + "epoch": 0.922244094488189, + "grad_norm": 10.231853485107422, + "learning_rate": 9.264413518886681e-07, + "loss": 0.6325, + "step": 937 + }, + { + "epoch": 0.9232283464566929, + "grad_norm": 10.755455017089844, + "learning_rate": 9.274353876739564e-07, + "loss": 0.7714, + "step": 938 + }, + { + "epoch": 0.9242125984251969, + "grad_norm": 8.56434154510498, + "learning_rate": 9.284294234592446e-07, + "loss": 0.4552, + "step": 939 + }, + { + "epoch": 0.9251968503937008, + "grad_norm": 15.114951133728027, + "learning_rate": 9.29423459244533e-07, + "loss": 1.3962, + "step": 940 + }, + { + "epoch": 0.9261811023622047, + "grad_norm": 17.96543312072754, + "learning_rate": 9.304174950298212e-07, + "loss": 1.3079, + "step": 941 + }, + { + "epoch": 0.9271653543307087, + "grad_norm": 11.526895523071289, + "learning_rate": 9.314115308151094e-07, + "loss": 0.8963, + "step": 942 + }, + { + "epoch": 0.9281496062992126, + "grad_norm": 12.170365333557129, + "learning_rate": 9.324055666003978e-07, + "loss": 0.7712, + "step": 943 + }, + { + "epoch": 0.9291338582677166, + "grad_norm": 10.663252830505371, + "learning_rate": 9.33399602385686e-07, + "loss": 0.7079, + "step": 944 + }, + { + "epoch": 0.9301181102362205, + "grad_norm": 12.715139389038086, + "learning_rate": 9.343936381709743e-07, + "loss": 1.2151, + "step": 945 + }, + { + "epoch": 0.9311023622047244, + "grad_norm": 9.071114540100098, + "learning_rate": 9.353876739562625e-07, + "loss": 0.5961, + "step": 946 + }, + { + "epoch": 0.9320866141732284, + "grad_norm": 20.959735870361328, + "learning_rate": 9.363817097415509e-07, + "loss": 1.5555, + "step": 947 + }, + { + "epoch": 0.9330708661417323, + "grad_norm": 9.442753791809082, + "learning_rate": 9.373757455268392e-07, + "loss": 0.6374, + "step": 948 + }, + { + "epoch": 0.9340551181102362, + "grad_norm": 12.313002586364746, + "learning_rate": 9.383697813121274e-07, + "loss": 0.8514, + "step": 949 + }, + { + "epoch": 0.9350393700787402, + "grad_norm": 10.137872695922852, + "learning_rate": 9.393638170974157e-07, + "loss": 1.0144, + "step": 950 + }, + { + "epoch": 0.9360236220472441, + "grad_norm": 8.325979232788086, + "learning_rate": 9.40357852882704e-07, + "loss": 0.346, + "step": 951 + }, + { + "epoch": 0.937007874015748, + "grad_norm": 10.4571533203125, + "learning_rate": 9.413518886679922e-07, + "loss": 0.7938, + "step": 952 + }, + { + "epoch": 0.937992125984252, + "grad_norm": 10.583680152893066, + "learning_rate": 9.423459244532805e-07, + "loss": 0.7822, + "step": 953 + }, + { + "epoch": 0.9389763779527559, + "grad_norm": 30.63351058959961, + "learning_rate": 9.433399602385688e-07, + "loss": 2.5079, + "step": 954 + }, + { + "epoch": 0.9399606299212598, + "grad_norm": 9.195967674255371, + "learning_rate": 9.44333996023857e-07, + "loss": 0.4717, + "step": 955 + }, + { + "epoch": 0.9409448818897638, + "grad_norm": 29.021352767944336, + "learning_rate": 9.453280318091453e-07, + "loss": 2.047, + "step": 956 + }, + { + "epoch": 0.9419291338582677, + "grad_norm": 26.134672164916992, + "learning_rate": 9.463220675944336e-07, + "loss": 1.4548, + "step": 957 + }, + { + "epoch": 0.9429133858267716, + "grad_norm": 9.519295692443848, + "learning_rate": 9.473161033797219e-07, + "loss": 0.6623, + "step": 958 + }, + { + "epoch": 0.9438976377952756, + "grad_norm": 11.477132797241211, + "learning_rate": 9.483101391650101e-07, + "loss": 0.8172, + "step": 959 + }, + { + "epoch": 0.9448818897637795, + "grad_norm": 11.666595458984375, + "learning_rate": 9.493041749502983e-07, + "loss": 0.9362, + "step": 960 + }, + { + "epoch": 0.9458661417322834, + "grad_norm": 24.117143630981445, + "learning_rate": 9.502982107355867e-07, + "loss": 1.6731, + "step": 961 + }, + { + "epoch": 0.9468503937007874, + "grad_norm": 8.999564170837402, + "learning_rate": 9.512922465208749e-07, + "loss": 0.4495, + "step": 962 + }, + { + "epoch": 0.9478346456692913, + "grad_norm": 8.393959999084473, + "learning_rate": 9.522862823061631e-07, + "loss": 0.5375, + "step": 963 + }, + { + "epoch": 0.9488188976377953, + "grad_norm": 11.88058853149414, + "learning_rate": 9.532803180914515e-07, + "loss": 1.3343, + "step": 964 + }, + { + "epoch": 0.9498031496062992, + "grad_norm": 9.698150634765625, + "learning_rate": 9.542743538767396e-07, + "loss": 0.5332, + "step": 965 + }, + { + "epoch": 0.9507874015748031, + "grad_norm": 10.911288261413574, + "learning_rate": 9.55268389662028e-07, + "loss": 1.0183, + "step": 966 + }, + { + "epoch": 0.9517716535433071, + "grad_norm": 10.970593452453613, + "learning_rate": 9.562624254473162e-07, + "loss": 0.6058, + "step": 967 + }, + { + "epoch": 0.952755905511811, + "grad_norm": 10.558159828186035, + "learning_rate": 9.572564612326044e-07, + "loss": 0.6536, + "step": 968 + }, + { + "epoch": 0.9537401574803149, + "grad_norm": 12.001582145690918, + "learning_rate": 9.582504970178928e-07, + "loss": 1.0448, + "step": 969 + }, + { + "epoch": 0.9547244094488189, + "grad_norm": 12.771181106567383, + "learning_rate": 9.59244532803181e-07, + "loss": 0.9479, + "step": 970 + }, + { + "epoch": 0.9557086614173228, + "grad_norm": 12.624911308288574, + "learning_rate": 9.602385685884694e-07, + "loss": 0.8316, + "step": 971 + }, + { + "epoch": 0.9566929133858267, + "grad_norm": 11.48292064666748, + "learning_rate": 9.612326043737576e-07, + "loss": 1.0847, + "step": 972 + }, + { + "epoch": 0.9576771653543307, + "grad_norm": 15.085115432739258, + "learning_rate": 9.622266401590458e-07, + "loss": 1.3262, + "step": 973 + }, + { + "epoch": 0.9586614173228346, + "grad_norm": 10.267237663269043, + "learning_rate": 9.632206759443342e-07, + "loss": 0.6488, + "step": 974 + }, + { + "epoch": 0.9596456692913385, + "grad_norm": 12.869492530822754, + "learning_rate": 9.642147117296224e-07, + "loss": 0.7577, + "step": 975 + }, + { + "epoch": 0.9606299212598425, + "grad_norm": 13.7593994140625, + "learning_rate": 9.652087475149109e-07, + "loss": 1.0546, + "step": 976 + }, + { + "epoch": 0.9616141732283464, + "grad_norm": 11.3866548538208, + "learning_rate": 9.66202783300199e-07, + "loss": 0.9759, + "step": 977 + }, + { + "epoch": 0.9625984251968503, + "grad_norm": 10.745214462280273, + "learning_rate": 9.671968190854873e-07, + "loss": 0.526, + "step": 978 + }, + { + "epoch": 0.9635826771653543, + "grad_norm": 11.974394798278809, + "learning_rate": 9.681908548707755e-07, + "loss": 0.9726, + "step": 979 + }, + { + "epoch": 0.9645669291338582, + "grad_norm": 10.351767539978027, + "learning_rate": 9.691848906560639e-07, + "loss": 0.7035, + "step": 980 + }, + { + "epoch": 0.9655511811023622, + "grad_norm": 7.831402778625488, + "learning_rate": 9.70178926441352e-07, + "loss": 0.4028, + "step": 981 + }, + { + "epoch": 0.9665354330708661, + "grad_norm": 13.162599563598633, + "learning_rate": 9.711729622266403e-07, + "loss": 0.889, + "step": 982 + }, + { + "epoch": 0.96751968503937, + "grad_norm": 11.184473991394043, + "learning_rate": 9.721669980119285e-07, + "loss": 0.6391, + "step": 983 + }, + { + "epoch": 0.968503937007874, + "grad_norm": 29.158571243286133, + "learning_rate": 9.731610337972169e-07, + "loss": 2.2124, + "step": 984 + }, + { + "epoch": 0.969488188976378, + "grad_norm": 35.266605377197266, + "learning_rate": 9.74155069582505e-07, + "loss": 2.5108, + "step": 985 + }, + { + "epoch": 0.9704724409448819, + "grad_norm": 8.492383003234863, + "learning_rate": 9.751491053677933e-07, + "loss": 0.5352, + "step": 986 + }, + { + "epoch": 0.9714566929133859, + "grad_norm": 11.822407722473145, + "learning_rate": 9.761431411530817e-07, + "loss": 0.7982, + "step": 987 + }, + { + "epoch": 0.9724409448818898, + "grad_norm": 11.544654846191406, + "learning_rate": 9.771371769383699e-07, + "loss": 0.8057, + "step": 988 + }, + { + "epoch": 0.9734251968503937, + "grad_norm": 11.718117713928223, + "learning_rate": 9.781312127236583e-07, + "loss": 0.6363, + "step": 989 + }, + { + "epoch": 0.9744094488188977, + "grad_norm": 21.24821662902832, + "learning_rate": 9.791252485089465e-07, + "loss": 1.4105, + "step": 990 + }, + { + "epoch": 0.9753937007874016, + "grad_norm": 11.616889953613281, + "learning_rate": 9.801192842942347e-07, + "loss": 0.6527, + "step": 991 + }, + { + "epoch": 0.9763779527559056, + "grad_norm": 10.165470123291016, + "learning_rate": 9.811133200795231e-07, + "loss": 0.7418, + "step": 992 + }, + { + "epoch": 0.9773622047244095, + "grad_norm": 23.416990280151367, + "learning_rate": 9.821073558648113e-07, + "loss": 1.5734, + "step": 993 + }, + { + "epoch": 0.9783464566929134, + "grad_norm": 10.304159164428711, + "learning_rate": 9.831013916500995e-07, + "loss": 0.512, + "step": 994 + }, + { + "epoch": 0.9793307086614174, + "grad_norm": 11.679780006408691, + "learning_rate": 9.84095427435388e-07, + "loss": 0.7346, + "step": 995 + }, + { + "epoch": 0.9803149606299213, + "grad_norm": 10.360453605651855, + "learning_rate": 9.850894632206761e-07, + "loss": 0.6094, + "step": 996 + }, + { + "epoch": 0.9812992125984252, + "grad_norm": 14.059850692749023, + "learning_rate": 9.860834990059645e-07, + "loss": 0.9234, + "step": 997 + }, + { + "epoch": 0.9822834645669292, + "grad_norm": 27.811769485473633, + "learning_rate": 9.870775347912527e-07, + "loss": 2.1518, + "step": 998 + }, + { + "epoch": 0.9832677165354331, + "grad_norm": 7.412935733795166, + "learning_rate": 9.88071570576541e-07, + "loss": 0.458, + "step": 999 + }, + { + "epoch": 0.984251968503937, + "grad_norm": 12.724889755249023, + "learning_rate": 9.890656063618291e-07, + "loss": 1.0281, + "step": 1000 + }, + { + "epoch": 0.985236220472441, + "grad_norm": 10.296438217163086, + "learning_rate": 9.900596421471175e-07, + "loss": 0.735, + "step": 1001 + }, + { + "epoch": 0.9862204724409449, + "grad_norm": 12.345407485961914, + "learning_rate": 9.910536779324057e-07, + "loss": 1.1242, + "step": 1002 + }, + { + "epoch": 0.9872047244094488, + "grad_norm": 17.737520217895508, + "learning_rate": 9.92047713717694e-07, + "loss": 1.3979, + "step": 1003 + }, + { + "epoch": 0.9881889763779528, + "grad_norm": 11.95985221862793, + "learning_rate": 9.930417495029821e-07, + "loss": 0.8926, + "step": 1004 + }, + { + "epoch": 0.9891732283464567, + "grad_norm": 21.35641860961914, + "learning_rate": 9.940357852882705e-07, + "loss": 2.1105, + "step": 1005 + }, + { + "epoch": 0.9901574803149606, + "grad_norm": 11.893367767333984, + "learning_rate": 9.950298210735587e-07, + "loss": 0.6443, + "step": 1006 + }, + { + "epoch": 0.9911417322834646, + "grad_norm": 20.44769859313965, + "learning_rate": 9.96023856858847e-07, + "loss": 1.6493, + "step": 1007 + }, + { + "epoch": 0.9921259842519685, + "grad_norm": 15.562771797180176, + "learning_rate": 9.970178926441354e-07, + "loss": 1.125, + "step": 1008 + }, + { + "epoch": 0.9931102362204725, + "grad_norm": 7.511135578155518, + "learning_rate": 9.980119284294236e-07, + "loss": 0.3277, + "step": 1009 + }, + { + "epoch": 0.9940944881889764, + "grad_norm": 11.336944580078125, + "learning_rate": 9.99005964214712e-07, + "loss": 0.8848, + "step": 1010 + }, + { + "epoch": 0.9950787401574803, + "grad_norm": 10.511725425720215, + "learning_rate": 1.0000000000000002e-06, + "loss": 0.6624, + "step": 1011 + }, + { + "epoch": 0.9960629921259843, + "grad_norm": 11.216442108154297, + "learning_rate": 9.999994102361802e-07, + "loss": 0.7913, + "step": 1012 + }, + { + "epoch": 0.9970472440944882, + "grad_norm": 16.446813583374023, + "learning_rate": 9.999976409461164e-07, + "loss": 1.2572, + "step": 1013 + }, + { + "epoch": 0.9980314960629921, + "grad_norm": 13.772931098937988, + "learning_rate": 9.999946921339963e-07, + "loss": 1.2533, + "step": 1014 + }, + { + "epoch": 0.9990157480314961, + "grad_norm": 13.333099365234375, + "learning_rate": 9.999905638068e-07, + "loss": 0.7953, + "step": 1015 + }, + { + "epoch": 1.0, + "grad_norm": 11.766683578491211, + "learning_rate": 9.999852559742985e-07, + "loss": 0.3578, + "step": 1016 + }, + { + "epoch": 1.000984251968504, + "grad_norm": 14.253616333007812, + "learning_rate": 9.999787686490555e-07, + "loss": 1.1694, + "step": 1017 + }, + { + "epoch": 1.0019685039370079, + "grad_norm": 13.263395309448242, + "learning_rate": 9.99971101846426e-07, + "loss": 1.0959, + "step": 1018 + }, + { + "epoch": 1.0029527559055118, + "grad_norm": 14.207919120788574, + "learning_rate": 9.999622555845573e-07, + "loss": 0.8922, + "step": 1019 + }, + { + "epoch": 1.0039370078740157, + "grad_norm": 12.13494873046875, + "learning_rate": 9.999522298843875e-07, + "loss": 0.7743, + "step": 1020 + }, + { + "epoch": 1.0049212598425197, + "grad_norm": 9.635050773620605, + "learning_rate": 9.99941024769647e-07, + "loss": 0.5631, + "step": 1021 + }, + { + "epoch": 1.0059055118110236, + "grad_norm": 17.427759170532227, + "learning_rate": 9.999286402668578e-07, + "loss": 1.2144, + "step": 1022 + }, + { + "epoch": 1.0068897637795275, + "grad_norm": 8.528889656066895, + "learning_rate": 9.999150764053334e-07, + "loss": 0.5034, + "step": 1023 + }, + { + "epoch": 1.0078740157480315, + "grad_norm": 11.485081672668457, + "learning_rate": 9.999003332171784e-07, + "loss": 0.7687, + "step": 1024 + }, + { + "epoch": 1.0088582677165354, + "grad_norm": 11.63568115234375, + "learning_rate": 9.998844107372894e-07, + "loss": 0.7181, + "step": 1025 + }, + { + "epoch": 1.0098425196850394, + "grad_norm": 20.09282112121582, + "learning_rate": 9.99867309003354e-07, + "loss": 1.0367, + "step": 1026 + }, + { + "epoch": 1.0108267716535433, + "grad_norm": 10.6008882522583, + "learning_rate": 9.998490280558511e-07, + "loss": 0.8523, + "step": 1027 + }, + { + "epoch": 1.0118110236220472, + "grad_norm": 15.000754356384277, + "learning_rate": 9.998295679380506e-07, + "loss": 1.1932, + "step": 1028 + }, + { + "epoch": 1.0127952755905512, + "grad_norm": 20.114559173583984, + "learning_rate": 9.998089286960135e-07, + "loss": 1.3118, + "step": 1029 + }, + { + "epoch": 1.013779527559055, + "grad_norm": 10.466768264770508, + "learning_rate": 9.99787110378592e-07, + "loss": 0.8769, + "step": 1030 + }, + { + "epoch": 1.014763779527559, + "grad_norm": 11.994022369384766, + "learning_rate": 9.997641130374284e-07, + "loss": 0.8931, + "step": 1031 + }, + { + "epoch": 1.015748031496063, + "grad_norm": 11.19292163848877, + "learning_rate": 9.997399367269565e-07, + "loss": 0.8208, + "step": 1032 + }, + { + "epoch": 1.016732283464567, + "grad_norm": 10.797515869140625, + "learning_rate": 9.997145815044003e-07, + "loss": 0.7876, + "step": 1033 + }, + { + "epoch": 1.0177165354330708, + "grad_norm": 12.783084869384766, + "learning_rate": 9.99688047429774e-07, + "loss": 1.1651, + "step": 1034 + }, + { + "epoch": 1.0187007874015748, + "grad_norm": 11.415165901184082, + "learning_rate": 9.996603345658824e-07, + "loss": 0.8233, + "step": 1035 + }, + { + "epoch": 1.0196850393700787, + "grad_norm": 11.22139835357666, + "learning_rate": 9.996314429783204e-07, + "loss": 0.7586, + "step": 1036 + }, + { + "epoch": 1.0206692913385826, + "grad_norm": 15.440783500671387, + "learning_rate": 9.99601372735473e-07, + "loss": 0.8531, + "step": 1037 + }, + { + "epoch": 1.0216535433070866, + "grad_norm": 27.25291633605957, + "learning_rate": 9.995701239085143e-07, + "loss": 1.81, + "step": 1038 + }, + { + "epoch": 1.0226377952755905, + "grad_norm": 11.31596565246582, + "learning_rate": 9.99537696571409e-07, + "loss": 0.601, + "step": 1039 + }, + { + "epoch": 1.0236220472440944, + "grad_norm": 10.041638374328613, + "learning_rate": 9.995040908009107e-07, + "loss": 0.6086, + "step": 1040 + }, + { + "epoch": 1.0246062992125984, + "grad_norm": 10.827730178833008, + "learning_rate": 9.994693066765622e-07, + "loss": 0.6538, + "step": 1041 + }, + { + "epoch": 1.0255905511811023, + "grad_norm": 9.742436408996582, + "learning_rate": 9.994333442806959e-07, + "loss": 0.5518, + "step": 1042 + }, + { + "epoch": 1.0265748031496063, + "grad_norm": 15.172879219055176, + "learning_rate": 9.993962036984325e-07, + "loss": 1.249, + "step": 1043 + }, + { + "epoch": 1.0275590551181102, + "grad_norm": 9.799712181091309, + "learning_rate": 9.99357885017682e-07, + "loss": 0.5059, + "step": 1044 + }, + { + "epoch": 1.0285433070866141, + "grad_norm": 10.319066047668457, + "learning_rate": 9.993183883291425e-07, + "loss": 0.6202, + "step": 1045 + }, + { + "epoch": 1.029527559055118, + "grad_norm": 12.552563667297363, + "learning_rate": 9.992777137263007e-07, + "loss": 0.8073, + "step": 1046 + }, + { + "epoch": 1.030511811023622, + "grad_norm": 10.50554084777832, + "learning_rate": 9.99235861305431e-07, + "loss": 0.4438, + "step": 1047 + }, + { + "epoch": 1.031496062992126, + "grad_norm": 16.503042221069336, + "learning_rate": 9.991928311655956e-07, + "loss": 1.4425, + "step": 1048 + }, + { + "epoch": 1.0324803149606299, + "grad_norm": 10.766325950622559, + "learning_rate": 9.991486234086447e-07, + "loss": 0.3772, + "step": 1049 + }, + { + "epoch": 1.0334645669291338, + "grad_norm": 7.945598125457764, + "learning_rate": 9.991032381392158e-07, + "loss": 0.4225, + "step": 1050 + }, + { + "epoch": 1.0344488188976377, + "grad_norm": 13.435383796691895, + "learning_rate": 9.990566754647328e-07, + "loss": 0.7363, + "step": 1051 + }, + { + "epoch": 1.0354330708661417, + "grad_norm": 10.391773223876953, + "learning_rate": 9.990089354954073e-07, + "loss": 0.4342, + "step": 1052 + }, + { + "epoch": 1.0364173228346456, + "grad_norm": 11.232378005981445, + "learning_rate": 9.989600183442373e-07, + "loss": 0.8763, + "step": 1053 + }, + { + "epoch": 1.0374015748031495, + "grad_norm": 12.171229362487793, + "learning_rate": 9.98909924127007e-07, + "loss": 0.8974, + "step": 1054 + }, + { + "epoch": 1.0383858267716535, + "grad_norm": 13.529887199401855, + "learning_rate": 9.988586529622864e-07, + "loss": 0.9175, + "step": 1055 + }, + { + "epoch": 1.0393700787401574, + "grad_norm": 12.536489486694336, + "learning_rate": 9.98806204971432e-07, + "loss": 0.9145, + "step": 1056 + }, + { + "epoch": 1.0403543307086613, + "grad_norm": 9.952815055847168, + "learning_rate": 9.987525802785847e-07, + "loss": 0.7247, + "step": 1057 + }, + { + "epoch": 1.0413385826771653, + "grad_norm": 10.323770523071289, + "learning_rate": 9.986977790106715e-07, + "loss": 0.6066, + "step": 1058 + }, + { + "epoch": 1.0423228346456692, + "grad_norm": 10.691637992858887, + "learning_rate": 9.986418012974042e-07, + "loss": 0.5892, + "step": 1059 + }, + { + "epoch": 1.0433070866141732, + "grad_norm": 27.982858657836914, + "learning_rate": 9.985846472712786e-07, + "loss": 2.1779, + "step": 1060 + }, + { + "epoch": 1.044291338582677, + "grad_norm": 11.58721923828125, + "learning_rate": 9.985263170675756e-07, + "loss": 0.7973, + "step": 1061 + }, + { + "epoch": 1.045275590551181, + "grad_norm": 8.09024715423584, + "learning_rate": 9.98466810824359e-07, + "loss": 0.4354, + "step": 1062 + }, + { + "epoch": 1.046259842519685, + "grad_norm": 23.964033126831055, + "learning_rate": 9.984061286824774e-07, + "loss": 1.2032, + "step": 1063 + }, + { + "epoch": 1.047244094488189, + "grad_norm": 15.369793891906738, + "learning_rate": 9.983442707855616e-07, + "loss": 1.088, + "step": 1064 + }, + { + "epoch": 1.0482283464566928, + "grad_norm": 8.0208101272583, + "learning_rate": 9.982812372800262e-07, + "loss": 0.3944, + "step": 1065 + }, + { + "epoch": 1.0492125984251968, + "grad_norm": 8.56423568725586, + "learning_rate": 9.982170283150678e-07, + "loss": 0.5178, + "step": 1066 + }, + { + "epoch": 1.0501968503937007, + "grad_norm": 14.80748462677002, + "learning_rate": 9.981516440426659e-07, + "loss": 1.0818, + "step": 1067 + }, + { + "epoch": 1.0511811023622046, + "grad_norm": 12.083704948425293, + "learning_rate": 9.98085084617581e-07, + "loss": 0.8308, + "step": 1068 + }, + { + "epoch": 1.0521653543307086, + "grad_norm": 20.59059715270996, + "learning_rate": 9.980173501973558e-07, + "loss": 1.54, + "step": 1069 + }, + { + "epoch": 1.0531496062992125, + "grad_norm": 11.229269981384277, + "learning_rate": 9.97948440942314e-07, + "loss": 0.8444, + "step": 1070 + }, + { + "epoch": 1.0541338582677164, + "grad_norm": 9.534320831298828, + "learning_rate": 9.978783570155597e-07, + "loss": 0.4829, + "step": 1071 + }, + { + "epoch": 1.0541338582677164, + "eval_Qnli-dev_cosine_accuracy": 0.677734375, + "eval_Qnli-dev_cosine_accuracy_threshold": 0.8417912721633911, + "eval_Qnli-dev_cosine_ap": 0.6886087408755974, + "eval_Qnli-dev_cosine_f1": 0.6799999999999999, + "eval_Qnli-dev_cosine_f1_threshold": 0.7771409749984741, + "eval_Qnli-dev_cosine_precision": 0.5604395604395604, + "eval_Qnli-dev_cosine_recall": 0.864406779661017, + "eval_Qnli-dev_dot_accuracy": 0.681640625, + "eval_Qnli-dev_dot_accuracy_threshold": 642.3902587890625, + "eval_Qnli-dev_dot_ap": 0.688149265594685, + "eval_Qnli-dev_dot_f1": 0.6780238500851788, + "eval_Qnli-dev_dot_f1_threshold": 596.5299072265625, + "eval_Qnli-dev_dot_precision": 0.5669515669515669, + "eval_Qnli-dev_dot_recall": 0.8432203389830508, + "eval_Qnli-dev_euclidean_accuracy": 0.677734375, + "eval_Qnli-dev_euclidean_accuracy_threshold": 15.544126510620117, + "eval_Qnli-dev_euclidean_ap": 0.6885091394341563, + "eval_Qnli-dev_euclidean_f1": 0.6821963394342762, + "eval_Qnli-dev_euclidean_f1_threshold": 18.467456817626953, + "eval_Qnli-dev_euclidean_precision": 0.5616438356164384, + "eval_Qnli-dev_euclidean_recall": 0.8686440677966102, + "eval_Qnli-dev_manhattan_accuracy": 0.689453125, + "eval_Qnli-dev_manhattan_accuracy_threshold": 328.31695556640625, + "eval_Qnli-dev_manhattan_ap": 0.6916225946572983, + "eval_Qnli-dev_manhattan_f1": 0.6816720257234726, + "eval_Qnli-dev_manhattan_f1_threshold": 396.9043884277344, + "eval_Qnli-dev_manhattan_precision": 0.5492227979274611, + "eval_Qnli-dev_manhattan_recall": 0.8983050847457628, + "eval_Qnli-dev_max_accuracy": 0.689453125, + "eval_Qnli-dev_max_accuracy_threshold": 642.3902587890625, + "eval_Qnli-dev_max_ap": 0.6916225946572983, + "eval_Qnli-dev_max_f1": 0.6821963394342762, + "eval_Qnli-dev_max_f1_threshold": 596.5299072265625, + "eval_Qnli-dev_max_precision": 0.5669515669515669, + "eval_Qnli-dev_max_recall": 0.8983050847457628, + "eval_allNLI-dev_cosine_accuracy": 0.6875, + "eval_allNLI-dev_cosine_accuracy_threshold": 0.9252105951309204, + "eval_allNLI-dev_cosine_ap": 0.49265349936764846, + "eval_allNLI-dev_cosine_f1": 0.5250431778929189, + "eval_allNLI-dev_cosine_f1_threshold": 0.8086467385292053, + "eval_allNLI-dev_cosine_precision": 0.37438423645320196, + "eval_allNLI-dev_cosine_recall": 0.8786127167630058, + "eval_allNLI-dev_dot_accuracy": 0.6875, + "eval_allNLI-dev_dot_accuracy_threshold": 703.1842041015625, + "eval_allNLI-dev_dot_ap": 0.4926134658014375, + "eval_allNLI-dev_dot_f1": 0.5259515570934257, + "eval_allNLI-dev_dot_f1_threshold": 617.3992309570312, + "eval_allNLI-dev_dot_precision": 0.37530864197530867, + "eval_allNLI-dev_dot_recall": 0.8786127167630058, + "eval_allNLI-dev_euclidean_accuracy": 0.6875, + "eval_allNLI-dev_euclidean_accuracy_threshold": 10.687753677368164, + "eval_allNLI-dev_euclidean_ap": 0.49249691805198476, + "eval_allNLI-dev_euclidean_f1": 0.5241379310344828, + "eval_allNLI-dev_euclidean_f1_threshold": 17.097484588623047, + "eval_allNLI-dev_euclidean_precision": 0.37346437346437344, + "eval_allNLI-dev_euclidean_recall": 0.8786127167630058, + "eval_allNLI-dev_manhattan_accuracy": 0.689453125, + "eval_allNLI-dev_manhattan_accuracy_threshold": 225.8095703125, + "eval_allNLI-dev_manhattan_ap": 0.49426585546140256, + "eval_allNLI-dev_manhattan_f1": 0.5268292682926828, + "eval_allNLI-dev_manhattan_f1_threshold": 291.3158264160156, + "eval_allNLI-dev_manhattan_precision": 0.45569620253164556, + "eval_allNLI-dev_manhattan_recall": 0.6242774566473989, + "eval_allNLI-dev_max_accuracy": 0.689453125, + "eval_allNLI-dev_max_accuracy_threshold": 703.1842041015625, + "eval_allNLI-dev_max_ap": 0.49426585546140256, + "eval_allNLI-dev_max_f1": 0.5268292682926828, + "eval_allNLI-dev_max_f1_threshold": 617.3992309570312, + "eval_allNLI-dev_max_precision": 0.45569620253164556, + "eval_allNLI-dev_max_recall": 0.8786127167630058, + "eval_loss": 0.7322226166725159, + "eval_runtime": 50.082, + "eval_samples_per_second": 33.226, + "eval_sequential_score": 0.6916225946572983, + "eval_steps_per_second": 0.14, + "eval_sts-test_pearson_cosine": 0.556294209379771, + "eval_sts-test_pearson_dot": 0.554452173690328, + "eval_sts-test_pearson_euclidean": 0.5742151059777929, + "eval_sts-test_pearson_manhattan": 0.5858768099408169, + "eval_sts-test_pearson_max": 0.5858768099408169, + "eval_sts-test_spearman_cosine": 0.5792614909830714, + "eval_sts-test_spearman_dot": 0.5782108656193855, + "eval_sts-test_spearman_euclidean": 0.579772863646991, + "eval_sts-test_spearman_manhattan": 0.5911280934637061, + "eval_sts-test_spearman_max": 0.5911280934637061, + "step": 1071 + }, + { + "epoch": 1.0551181102362204, + "grad_norm": 8.921422958374023, + "learning_rate": 9.978070985829782e-07, + "loss": 0.495, + "step": 1072 + }, + { + "epoch": 1.0561023622047243, + "grad_norm": 11.355947494506836, + "learning_rate": 9.97734665813234e-07, + "loss": 0.8591, + "step": 1073 + }, + { + "epoch": 1.0570866141732282, + "grad_norm": 6.873776435852051, + "learning_rate": 9.976610588777717e-07, + "loss": 0.327, + "step": 1074 + }, + { + "epoch": 1.0580708661417322, + "grad_norm": 13.124799728393555, + "learning_rate": 9.975862779508149e-07, + "loss": 0.7161, + "step": 1075 + }, + { + "epoch": 1.0590551181102361, + "grad_norm": 8.418201446533203, + "learning_rate": 9.975103232093656e-07, + "loss": 0.6374, + "step": 1076 + }, + { + "epoch": 1.06003937007874, + "grad_norm": 17.85619354248047, + "learning_rate": 9.974331948332048e-07, + "loss": 1.1748, + "step": 1077 + }, + { + "epoch": 1.061023622047244, + "grad_norm": 24.725465774536133, + "learning_rate": 9.97354893004891e-07, + "loss": 1.7501, + "step": 1078 + }, + { + "epoch": 1.0620078740157481, + "grad_norm": 9.624460220336914, + "learning_rate": 9.972754179097602e-07, + "loss": 0.5544, + "step": 1079 + }, + { + "epoch": 1.0629921259842519, + "grad_norm": 10.143649101257324, + "learning_rate": 9.971947697359257e-07, + "loss": 0.6265, + "step": 1080 + }, + { + "epoch": 1.063976377952756, + "grad_norm": 19.294435501098633, + "learning_rate": 9.971129486742777e-07, + "loss": 1.6517, + "step": 1081 + }, + { + "epoch": 1.0649606299212597, + "grad_norm": 9.823873519897461, + "learning_rate": 9.970299549184814e-07, + "loss": 0.7457, + "step": 1082 + }, + { + "epoch": 1.0659448818897639, + "grad_norm": 10.644150733947754, + "learning_rate": 9.969457886649791e-07, + "loss": 0.7492, + "step": 1083 + }, + { + "epoch": 1.0669291338582678, + "grad_norm": 11.440357208251953, + "learning_rate": 9.968604501129877e-07, + "loss": 0.8013, + "step": 1084 + }, + { + "epoch": 1.0679133858267718, + "grad_norm": 4.15566873550415, + "learning_rate": 9.967739394644982e-07, + "loss": 0.1619, + "step": 1085 + }, + { + "epoch": 1.0688976377952757, + "grad_norm": 8.881417274475098, + "learning_rate": 9.966862569242773e-07, + "loss": 0.5057, + "step": 1086 + }, + { + "epoch": 1.0698818897637796, + "grad_norm": 8.335419654846191, + "learning_rate": 9.965974026998642e-07, + "loss": 0.4712, + "step": 1087 + }, + { + "epoch": 1.0708661417322836, + "grad_norm": 11.294657707214355, + "learning_rate": 9.965073770015722e-07, + "loss": 0.8382, + "step": 1088 + }, + { + "epoch": 1.0718503937007875, + "grad_norm": 9.385281562805176, + "learning_rate": 9.964161800424873e-07, + "loss": 0.6045, + "step": 1089 + }, + { + "epoch": 1.0728346456692914, + "grad_norm": 10.363978385925293, + "learning_rate": 9.963238120384677e-07, + "loss": 0.6117, + "step": 1090 + }, + { + "epoch": 1.0738188976377954, + "grad_norm": 10.40096378326416, + "learning_rate": 9.962302732081432e-07, + "loss": 0.7028, + "step": 1091 + }, + { + "epoch": 1.0748031496062993, + "grad_norm": 19.598663330078125, + "learning_rate": 9.961355637729153e-07, + "loss": 1.2376, + "step": 1092 + }, + { + "epoch": 1.0757874015748032, + "grad_norm": 11.906511306762695, + "learning_rate": 9.960396839569557e-07, + "loss": 1.045, + "step": 1093 + }, + { + "epoch": 1.0767716535433072, + "grad_norm": 19.7343692779541, + "learning_rate": 9.95942633987207e-07, + "loss": 1.1152, + "step": 1094 + }, + { + "epoch": 1.077755905511811, + "grad_norm": 10.810567855834961, + "learning_rate": 9.95844414093381e-07, + "loss": 0.5572, + "step": 1095 + }, + { + "epoch": 1.078740157480315, + "grad_norm": 10.99404239654541, + "learning_rate": 9.957450245079588e-07, + "loss": 0.7047, + "step": 1096 + }, + { + "epoch": 1.079724409448819, + "grad_norm": 17.423023223876953, + "learning_rate": 9.9564446546619e-07, + "loss": 1.4233, + "step": 1097 + }, + { + "epoch": 1.080708661417323, + "grad_norm": 13.35729694366455, + "learning_rate": 9.955427372060928e-07, + "loss": 0.8478, + "step": 1098 + }, + { + "epoch": 1.0816929133858268, + "grad_norm": 10.397942543029785, + "learning_rate": 9.954398399684515e-07, + "loss": 0.6851, + "step": 1099 + }, + { + "epoch": 1.0826771653543308, + "grad_norm": 9.397493362426758, + "learning_rate": 9.953357739968192e-07, + "loss": 0.4462, + "step": 1100 + }, + { + "epoch": 1.0836614173228347, + "grad_norm": 22.470144271850586, + "learning_rate": 9.952305395375137e-07, + "loss": 2.1139, + "step": 1101 + }, + { + "epoch": 1.0846456692913387, + "grad_norm": 10.985520362854004, + "learning_rate": 9.951241368396194e-07, + "loss": 0.8097, + "step": 1102 + }, + { + "epoch": 1.0856299212598426, + "grad_norm": 12.094350814819336, + "learning_rate": 9.950165661549856e-07, + "loss": 1.0912, + "step": 1103 + }, + { + "epoch": 1.0866141732283465, + "grad_norm": 15.580087661743164, + "learning_rate": 9.949078277382262e-07, + "loss": 1.1922, + "step": 1104 + }, + { + "epoch": 1.0875984251968505, + "grad_norm": 7.17451810836792, + "learning_rate": 9.947979218467193e-07, + "loss": 0.3888, + "step": 1105 + }, + { + "epoch": 1.0885826771653544, + "grad_norm": 11.2852144241333, + "learning_rate": 9.946868487406057e-07, + "loss": 0.7842, + "step": 1106 + }, + { + "epoch": 1.0895669291338583, + "grad_norm": 4.887936592102051, + "learning_rate": 9.945746086827896e-07, + "loss": 0.1422, + "step": 1107 + }, + { + "epoch": 1.0905511811023623, + "grad_norm": 11.22752571105957, + "learning_rate": 9.94461201938937e-07, + "loss": 0.6949, + "step": 1108 + }, + { + "epoch": 1.0915354330708662, + "grad_norm": 11.954117774963379, + "learning_rate": 9.943466287774756e-07, + "loss": 0.819, + "step": 1109 + }, + { + "epoch": 1.0925196850393701, + "grad_norm": 8.453436851501465, + "learning_rate": 9.942308894695932e-07, + "loss": 0.4947, + "step": 1110 + }, + { + "epoch": 1.093503937007874, + "grad_norm": 6.822066783905029, + "learning_rate": 9.941139842892391e-07, + "loss": 0.3346, + "step": 1111 + }, + { + "epoch": 1.094488188976378, + "grad_norm": 16.205955505371094, + "learning_rate": 9.939959135131214e-07, + "loss": 1.1459, + "step": 1112 + }, + { + "epoch": 1.095472440944882, + "grad_norm": 6.679828643798828, + "learning_rate": 9.938766774207064e-07, + "loss": 0.3276, + "step": 1113 + }, + { + "epoch": 1.0964566929133859, + "grad_norm": 11.79887580871582, + "learning_rate": 9.937562762942202e-07, + "loss": 0.7464, + "step": 1114 + }, + { + "epoch": 1.0974409448818898, + "grad_norm": 14.634333610534668, + "learning_rate": 9.936347104186452e-07, + "loss": 0.8906, + "step": 1115 + }, + { + "epoch": 1.0984251968503937, + "grad_norm": 29.19684600830078, + "learning_rate": 9.935119800817215e-07, + "loss": 1.9711, + "step": 1116 + }, + { + "epoch": 1.0994094488188977, + "grad_norm": 9.913980484008789, + "learning_rate": 9.933880855739446e-07, + "loss": 0.6403, + "step": 1117 + }, + { + "epoch": 1.1003937007874016, + "grad_norm": 26.225311279296875, + "learning_rate": 9.932630271885662e-07, + "loss": 1.3684, + "step": 1118 + }, + { + "epoch": 1.1013779527559056, + "grad_norm": 14.971549987792969, + "learning_rate": 9.93136805221593e-07, + "loss": 1.2074, + "step": 1119 + }, + { + "epoch": 1.1023622047244095, + "grad_norm": 10.119375228881836, + "learning_rate": 9.930094199717848e-07, + "loss": 0.5098, + "step": 1120 + }, + { + "epoch": 1.1033464566929134, + "grad_norm": 10.07550048828125, + "learning_rate": 9.92880871740656e-07, + "loss": 0.5498, + "step": 1121 + }, + { + "epoch": 1.1043307086614174, + "grad_norm": 9.61234188079834, + "learning_rate": 9.927511608324732e-07, + "loss": 0.3848, + "step": 1122 + }, + { + "epoch": 1.1053149606299213, + "grad_norm": 24.639860153198242, + "learning_rate": 9.926202875542546e-07, + "loss": 2.0202, + "step": 1123 + }, + { + "epoch": 1.1062992125984252, + "grad_norm": 10.644918441772461, + "learning_rate": 9.924882522157706e-07, + "loss": 0.5944, + "step": 1124 + }, + { + "epoch": 1.1072834645669292, + "grad_norm": 6.669126987457275, + "learning_rate": 9.923550551295411e-07, + "loss": 0.3266, + "step": 1125 + }, + { + "epoch": 1.108267716535433, + "grad_norm": 19.6352596282959, + "learning_rate": 9.922206966108367e-07, + "loss": 1.0289, + "step": 1126 + }, + { + "epoch": 1.109251968503937, + "grad_norm": 23.757041931152344, + "learning_rate": 9.920851769776761e-07, + "loss": 1.0807, + "step": 1127 + }, + { + "epoch": 1.110236220472441, + "grad_norm": 9.98759651184082, + "learning_rate": 9.919484965508276e-07, + "loss": 0.6155, + "step": 1128 + }, + { + "epoch": 1.111220472440945, + "grad_norm": 18.757999420166016, + "learning_rate": 9.918106556538057e-07, + "loss": 1.1686, + "step": 1129 + }, + { + "epoch": 1.1122047244094488, + "grad_norm": 13.33157730102539, + "learning_rate": 9.91671654612872e-07, + "loss": 1.0762, + "step": 1130 + }, + { + "epoch": 1.1131889763779528, + "grad_norm": 10.404326438903809, + "learning_rate": 9.91531493757035e-07, + "loss": 0.6781, + "step": 1131 + }, + { + "epoch": 1.1141732283464567, + "grad_norm": 10.745798110961914, + "learning_rate": 9.913901734180476e-07, + "loss": 0.6144, + "step": 1132 + }, + { + "epoch": 1.1151574803149606, + "grad_norm": 12.104547500610352, + "learning_rate": 9.912476939304069e-07, + "loss": 0.8022, + "step": 1133 + }, + { + "epoch": 1.1161417322834646, + "grad_norm": 9.081506729125977, + "learning_rate": 9.911040556313543e-07, + "loss": 0.5213, + "step": 1134 + }, + { + "epoch": 1.1171259842519685, + "grad_norm": 10.227931022644043, + "learning_rate": 9.909592588608738e-07, + "loss": 0.6014, + "step": 1135 + }, + { + "epoch": 1.1181102362204725, + "grad_norm": 12.51789379119873, + "learning_rate": 9.908133039616911e-07, + "loss": 0.901, + "step": 1136 + }, + { + "epoch": 1.1190944881889764, + "grad_norm": 12.91264533996582, + "learning_rate": 9.90666191279274e-07, + "loss": 0.9938, + "step": 1137 + }, + { + "epoch": 1.1200787401574803, + "grad_norm": 25.061851501464844, + "learning_rate": 9.905179211618296e-07, + "loss": 1.8173, + "step": 1138 + }, + { + "epoch": 1.1210629921259843, + "grad_norm": 7.622464179992676, + "learning_rate": 9.903684939603054e-07, + "loss": 0.5572, + "step": 1139 + }, + { + "epoch": 1.1220472440944882, + "grad_norm": 11.782830238342285, + "learning_rate": 9.902179100283875e-07, + "loss": 0.7489, + "step": 1140 + }, + { + "epoch": 1.1230314960629921, + "grad_norm": 9.14368724822998, + "learning_rate": 9.900661697224995e-07, + "loss": 0.4338, + "step": 1141 + }, + { + "epoch": 1.124015748031496, + "grad_norm": 7.549107551574707, + "learning_rate": 9.899132734018024e-07, + "loss": 0.3086, + "step": 1142 + }, + { + "epoch": 1.125, + "grad_norm": 10.181684494018555, + "learning_rate": 9.897592214281936e-07, + "loss": 0.6942, + "step": 1143 + }, + { + "epoch": 1.125984251968504, + "grad_norm": 12.112625122070312, + "learning_rate": 9.896040141663054e-07, + "loss": 0.7665, + "step": 1144 + }, + { + "epoch": 1.1269685039370079, + "grad_norm": 9.436166763305664, + "learning_rate": 9.894476519835052e-07, + "loss": 0.2734, + "step": 1145 + }, + { + "epoch": 1.1279527559055118, + "grad_norm": 13.32164192199707, + "learning_rate": 9.892901352498932e-07, + "loss": 0.9961, + "step": 1146 + }, + { + "epoch": 1.1289370078740157, + "grad_norm": 12.448330879211426, + "learning_rate": 9.891314643383037e-07, + "loss": 0.5258, + "step": 1147 + }, + { + "epoch": 1.1299212598425197, + "grad_norm": 12.589882850646973, + "learning_rate": 9.88971639624301e-07, + "loss": 0.7122, + "step": 1148 + }, + { + "epoch": 1.1309055118110236, + "grad_norm": 9.494569778442383, + "learning_rate": 9.888106614861823e-07, + "loss": 0.3747, + "step": 1149 + }, + { + "epoch": 1.1318897637795275, + "grad_norm": 11.688382148742676, + "learning_rate": 9.886485303049734e-07, + "loss": 0.6397, + "step": 1150 + }, + { + "epoch": 1.1328740157480315, + "grad_norm": 10.375199317932129, + "learning_rate": 9.884852464644305e-07, + "loss": 0.5504, + "step": 1151 + }, + { + "epoch": 1.1338582677165354, + "grad_norm": 10.527824401855469, + "learning_rate": 9.883208103510367e-07, + "loss": 0.5572, + "step": 1152 + }, + { + "epoch": 1.1348425196850394, + "grad_norm": 14.038769721984863, + "learning_rate": 9.88155222354004e-07, + "loss": 0.7828, + "step": 1153 + }, + { + "epoch": 1.1358267716535433, + "grad_norm": 22.79770851135254, + "learning_rate": 9.879884828652698e-07, + "loss": 1.0443, + "step": 1154 + }, + { + "epoch": 1.1368110236220472, + "grad_norm": 13.345325469970703, + "learning_rate": 9.878205922794972e-07, + "loss": 1.0731, + "step": 1155 + }, + { + "epoch": 1.1377952755905512, + "grad_norm": 25.466354370117188, + "learning_rate": 9.876515509940742e-07, + "loss": 1.1341, + "step": 1156 + }, + { + "epoch": 1.138779527559055, + "grad_norm": 8.98187255859375, + "learning_rate": 9.874813594091122e-07, + "loss": 0.391, + "step": 1157 + }, + { + "epoch": 1.139763779527559, + "grad_norm": 22.255481719970703, + "learning_rate": 9.873100179274453e-07, + "loss": 1.462, + "step": 1158 + }, + { + "epoch": 1.140748031496063, + "grad_norm": 12.306983947753906, + "learning_rate": 9.871375269546292e-07, + "loss": 0.8131, + "step": 1159 + }, + { + "epoch": 1.141732283464567, + "grad_norm": 9.708534240722656, + "learning_rate": 9.869638868989407e-07, + "loss": 0.7323, + "step": 1160 + }, + { + "epoch": 1.1427165354330708, + "grad_norm": 10.042935371398926, + "learning_rate": 9.867890981713764e-07, + "loss": 0.5473, + "step": 1161 + }, + { + "epoch": 1.1437007874015748, + "grad_norm": 12.677626609802246, + "learning_rate": 9.866131611856514e-07, + "loss": 0.7973, + "step": 1162 + }, + { + "epoch": 1.1446850393700787, + "grad_norm": 10.884305000305176, + "learning_rate": 9.864360763581994e-07, + "loss": 0.5875, + "step": 1163 + }, + { + "epoch": 1.1456692913385826, + "grad_norm": 12.529742240905762, + "learning_rate": 9.862578441081699e-07, + "loss": 0.9248, + "step": 1164 + }, + { + "epoch": 1.1466535433070866, + "grad_norm": 11.557649612426758, + "learning_rate": 9.860784648574288e-07, + "loss": 0.6898, + "step": 1165 + }, + { + "epoch": 1.1476377952755905, + "grad_norm": 21.74925422668457, + "learning_rate": 9.858979390305573e-07, + "loss": 1.4924, + "step": 1166 + }, + { + "epoch": 1.1486220472440944, + "grad_norm": 12.445868492126465, + "learning_rate": 9.857162670548502e-07, + "loss": 0.8908, + "step": 1167 + }, + { + "epoch": 1.1496062992125984, + "grad_norm": 11.005480766296387, + "learning_rate": 9.855334493603147e-07, + "loss": 0.564, + "step": 1168 + }, + { + "epoch": 1.1505905511811023, + "grad_norm": 11.636656761169434, + "learning_rate": 9.853494863796705e-07, + "loss": 0.3779, + "step": 1169 + }, + { + "epoch": 1.1515748031496063, + "grad_norm": 13.485095977783203, + "learning_rate": 9.851643785483478e-07, + "loss": 1.0715, + "step": 1170 + }, + { + "epoch": 1.1525590551181102, + "grad_norm": 9.2525634765625, + "learning_rate": 9.849781263044867e-07, + "loss": 0.4366, + "step": 1171 + }, + { + "epoch": 1.1535433070866141, + "grad_norm": 10.915475845336914, + "learning_rate": 9.847907300889358e-07, + "loss": 0.6391, + "step": 1172 + }, + { + "epoch": 1.154527559055118, + "grad_norm": 18.239994049072266, + "learning_rate": 9.846021903452519e-07, + "loss": 1.2133, + "step": 1173 + }, + { + "epoch": 1.155511811023622, + "grad_norm": 23.0247802734375, + "learning_rate": 9.844125075196983e-07, + "loss": 1.4135, + "step": 1174 + }, + { + "epoch": 1.156496062992126, + "grad_norm": 12.33239459991455, + "learning_rate": 9.842216820612438e-07, + "loss": 0.7748, + "step": 1175 + }, + { + "epoch": 1.1574803149606299, + "grad_norm": 9.904097557067871, + "learning_rate": 9.840297144215615e-07, + "loss": 0.544, + "step": 1176 + }, + { + "epoch": 1.1584645669291338, + "grad_norm": 9.18204402923584, + "learning_rate": 9.838366050550287e-07, + "loss": 0.5168, + "step": 1177 + }, + { + "epoch": 1.1594488188976377, + "grad_norm": 10.689743041992188, + "learning_rate": 9.836423544187245e-07, + "loss": 0.6931, + "step": 1178 + }, + { + "epoch": 1.1604330708661417, + "grad_norm": 11.55836296081543, + "learning_rate": 9.834469629724293e-07, + "loss": 0.87, + "step": 1179 + }, + { + "epoch": 1.1614173228346456, + "grad_norm": 13.83465576171875, + "learning_rate": 9.832504311786238e-07, + "loss": 0.9842, + "step": 1180 + }, + { + "epoch": 1.1624015748031495, + "grad_norm": 8.401900291442871, + "learning_rate": 9.830527595024885e-07, + "loss": 0.3614, + "step": 1181 + }, + { + "epoch": 1.1633858267716535, + "grad_norm": 9.143038749694824, + "learning_rate": 9.82853948411901e-07, + "loss": 0.4167, + "step": 1182 + }, + { + "epoch": 1.1643700787401574, + "grad_norm": 7.40506649017334, + "learning_rate": 9.826539983774365e-07, + "loss": 0.3688, + "step": 1183 + }, + { + "epoch": 1.1653543307086613, + "grad_norm": 10.054327964782715, + "learning_rate": 9.824529098723657e-07, + "loss": 0.5431, + "step": 1184 + }, + { + "epoch": 1.1663385826771653, + "grad_norm": 9.978111267089844, + "learning_rate": 9.822506833726537e-07, + "loss": 0.6127, + "step": 1185 + }, + { + "epoch": 1.1673228346456692, + "grad_norm": 11.50671100616455, + "learning_rate": 9.8204731935696e-07, + "loss": 0.8693, + "step": 1186 + }, + { + "epoch": 1.1683070866141732, + "grad_norm": 10.58636474609375, + "learning_rate": 9.818428183066356e-07, + "loss": 0.7596, + "step": 1187 + }, + { + "epoch": 1.169291338582677, + "grad_norm": 12.282295227050781, + "learning_rate": 9.816371807057237e-07, + "loss": 0.724, + "step": 1188 + }, + { + "epoch": 1.170275590551181, + "grad_norm": 12.36290454864502, + "learning_rate": 9.81430407040957e-07, + "loss": 0.9105, + "step": 1189 + }, + { + "epoch": 1.171259842519685, + "grad_norm": 7.814232349395752, + "learning_rate": 9.812224978017573e-07, + "loss": 0.3941, + "step": 1190 + }, + { + "epoch": 1.172244094488189, + "grad_norm": 20.87270164489746, + "learning_rate": 9.810134534802346e-07, + "loss": 1.1768, + "step": 1191 + }, + { + "epoch": 1.1732283464566928, + "grad_norm": 10.50306510925293, + "learning_rate": 9.80803274571185e-07, + "loss": 0.5509, + "step": 1192 + }, + { + "epoch": 1.1742125984251968, + "grad_norm": 21.016460418701172, + "learning_rate": 9.80591961572091e-07, + "loss": 1.1616, + "step": 1193 + }, + { + "epoch": 1.1751968503937007, + "grad_norm": 12.868950843811035, + "learning_rate": 9.803795149831183e-07, + "loss": 0.6835, + "step": 1194 + }, + { + "epoch": 1.1761811023622046, + "grad_norm": 8.271536827087402, + "learning_rate": 9.801659353071165e-07, + "loss": 0.4379, + "step": 1195 + }, + { + "epoch": 1.1771653543307086, + "grad_norm": 11.406146049499512, + "learning_rate": 9.799512230496172e-07, + "loss": 0.5453, + "step": 1196 + }, + { + "epoch": 1.1781496062992125, + "grad_norm": 10.591256141662598, + "learning_rate": 9.79735378718832e-07, + "loss": 0.5505, + "step": 1197 + }, + { + "epoch": 1.1791338582677164, + "grad_norm": 12.822675704956055, + "learning_rate": 9.795184028256534e-07, + "loss": 0.7472, + "step": 1198 + }, + { + "epoch": 1.1801181102362204, + "grad_norm": 10.516703605651855, + "learning_rate": 9.793002958836506e-07, + "loss": 0.3541, + "step": 1199 + }, + { + "epoch": 1.1811023622047245, + "grad_norm": 14.338447570800781, + "learning_rate": 9.790810584090712e-07, + "loss": 0.796, + "step": 1200 + }, + { + "epoch": 1.1820866141732282, + "grad_norm": 10.278166770935059, + "learning_rate": 9.78860690920838e-07, + "loss": 0.558, + "step": 1201 + }, + { + "epoch": 1.1830708661417324, + "grad_norm": 13.116827011108398, + "learning_rate": 9.78639193940549e-07, + "loss": 0.8679, + "step": 1202 + }, + { + "epoch": 1.1840551181102361, + "grad_norm": 10.748214721679688, + "learning_rate": 9.784165679924752e-07, + "loss": 0.7619, + "step": 1203 + }, + { + "epoch": 1.1850393700787403, + "grad_norm": 10.780762672424316, + "learning_rate": 9.781928136035604e-07, + "loss": 0.7039, + "step": 1204 + }, + { + "epoch": 1.186023622047244, + "grad_norm": 11.740593910217285, + "learning_rate": 9.779679313034182e-07, + "loss": 0.7166, + "step": 1205 + }, + { + "epoch": 1.1870078740157481, + "grad_norm": 11.709891319274902, + "learning_rate": 9.777419216243333e-07, + "loss": 0.6982, + "step": 1206 + }, + { + "epoch": 1.1879921259842519, + "grad_norm": 9.765851020812988, + "learning_rate": 9.77514785101258e-07, + "loss": 0.4206, + "step": 1207 + }, + { + "epoch": 1.188976377952756, + "grad_norm": 9.983769416809082, + "learning_rate": 9.77286522271812e-07, + "loss": 0.6361, + "step": 1208 + }, + { + "epoch": 1.1899606299212597, + "grad_norm": 9.72842025756836, + "learning_rate": 9.77057133676281e-07, + "loss": 0.6248, + "step": 1209 + }, + { + "epoch": 1.1909448818897639, + "grad_norm": 11.863199234008789, + "learning_rate": 9.768266198576147e-07, + "loss": 0.7933, + "step": 1210 + }, + { + "epoch": 1.1919291338582676, + "grad_norm": 9.37264347076416, + "learning_rate": 9.765949813614273e-07, + "loss": 0.5985, + "step": 1211 + }, + { + "epoch": 1.1929133858267718, + "grad_norm": 14.158349990844727, + "learning_rate": 9.763622187359942e-07, + "loss": 0.6147, + "step": 1212 + }, + { + "epoch": 1.1938976377952755, + "grad_norm": 10.92676830291748, + "learning_rate": 9.761283325322515e-07, + "loss": 0.6085, + "step": 1213 + }, + { + "epoch": 1.1948818897637796, + "grad_norm": 9.06302261352539, + "learning_rate": 9.758933233037952e-07, + "loss": 0.6713, + "step": 1214 + }, + { + "epoch": 1.1958661417322836, + "grad_norm": 25.792692184448242, + "learning_rate": 9.756571916068793e-07, + "loss": 1.0315, + "step": 1215 + }, + { + "epoch": 1.1968503937007875, + "grad_norm": 34.984886169433594, + "learning_rate": 9.754199380004145e-07, + "loss": 2.0024, + "step": 1216 + }, + { + "epoch": 1.1978346456692914, + "grad_norm": 25.739473342895508, + "learning_rate": 9.75181563045967e-07, + "loss": 1.6034, + "step": 1217 + }, + { + "epoch": 1.1988188976377954, + "grad_norm": 22.122873306274414, + "learning_rate": 9.749420673077574e-07, + "loss": 1.7407, + "step": 1218 + }, + { + "epoch": 1.1998031496062993, + "grad_norm": 14.874014854431152, + "learning_rate": 9.747014513526588e-07, + "loss": 1.2014, + "step": 1219 + }, + { + "epoch": 1.2007874015748032, + "grad_norm": 24.43715476989746, + "learning_rate": 9.744597157501957e-07, + "loss": 1.8377, + "step": 1220 + } + ], + "logging_steps": 1, + "max_steps": 3048, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 305, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 32, + "trial_name": null, + "trial_params": null +}