diff --git "a/checkpoint-610/trainer_state.json" "b/checkpoint-610/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-610/trainer_state.json"
@@ -0,0 +1,4570 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6003937007874016,
+  "eval_steps": 153,
+  "global_step": 610,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000984251968503937,
+      "grad_norm": NaN,
+      "learning_rate": 0.0,
+      "loss": 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
+    }
+  ],
+  "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
+}