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": 10.4072,
+      "step": 1
+    },
+    {
+      "epoch": 0.001968503937007874,
+      "grad_norm": NaN,
+      "learning_rate": 0.0,
+      "loss": 11.0865,
+      "step": 2
+    },
+    {
+      "epoch": 0.002952755905511811,
+      "grad_norm": 60.69786071777344,
+      "learning_rate": 9.940357852882705e-10,
+      "loss": 9.5114,
+      "step": 3
+    },
+    {
+      "epoch": 0.003937007874015748,
+      "grad_norm": 59.147647857666016,
+      "learning_rate": 1.988071570576541e-09,
+      "loss": 9.9584,
+      "step": 4
+    },
+    {
+      "epoch": 0.004921259842519685,
+      "grad_norm": NaN,
+      "learning_rate": 1.988071570576541e-09,
+      "loss": 10.068,
+      "step": 5
+    },
+    {
+      "epoch": 0.005905511811023622,
+      "grad_norm": 65.82320404052734,
+      "learning_rate": 2.9821073558648116e-09,
+      "loss": 11.0224,
+      "step": 6
+    },
+    {
+      "epoch": 0.006889763779527559,
+      "grad_norm": 59.096107482910156,
+      "learning_rate": 3.976143141153082e-09,
+      "loss": 9.7703,
+      "step": 7
+    },
+    {
+      "epoch": 0.007874015748031496,
+      "grad_norm": 61.43330764770508,
+      "learning_rate": 4.970178926441353e-09,
+      "loss": 10.5005,
+      "step": 8
+    },
+    {
+      "epoch": 0.008858267716535433,
+      "grad_norm": 61.25212860107422,
+      "learning_rate": 5.964214711729623e-09,
+      "loss": 10.1987,
+      "step": 9
+    },
+    {
+      "epoch": 0.00984251968503937,
+      "grad_norm": 61.6600341796875,
+      "learning_rate": 6.9582504970178946e-09,
+      "loss": 10.0277,
+      "step": 10
+    },
+    {
+      "epoch": 0.010826771653543307,
+      "grad_norm": 61.71075439453125,
+      "learning_rate": 7.952286282306164e-09,
+      "loss": 10.6965,
+      "step": 11
+    },
+    {
+      "epoch": 0.011811023622047244,
+      "grad_norm": 59.34035873413086,
+      "learning_rate": 8.946322067594435e-09,
+      "loss": 10.0609,
+      "step": 12
+    },
+    {
+      "epoch": 0.012795275590551181,
+      "grad_norm": 64.14221954345703,
+      "learning_rate": 9.940357852882705e-09,
+      "loss": 11.6214,
+      "step": 13
+    },
+    {
+      "epoch": 0.013779527559055118,
+      "grad_norm": 57.682823181152344,
+      "learning_rate": 1.0934393638170978e-08,
+      "loss": 9.4053,
+      "step": 14
+    },
+    {
+      "epoch": 0.014763779527559055,
+      "grad_norm": 62.256858825683594,
+      "learning_rate": 1.1928429423459246e-08,
+      "loss": 10.4014,
+      "step": 15
+    },
+    {
+      "epoch": 0.015748031496062992,
+      "grad_norm": 64.56417083740234,
+      "learning_rate": 1.2922465208747517e-08,
+      "loss": 10.4119,
+      "step": 16
+    },
+    {
+      "epoch": 0.01673228346456693,
+      "grad_norm": 56.612098693847656,
+      "learning_rate": 1.3916500994035789e-08,
+      "loss": 9.4658,
+      "step": 17
+    },
+    {
+      "epoch": 0.017716535433070866,
+      "grad_norm": 60.02314758300781,
+      "learning_rate": 1.4910536779324056e-08,
+      "loss": 9.2169,
+      "step": 18
+    },
+    {
+      "epoch": 0.018700787401574805,
+      "grad_norm": 66.38670349121094,
+      "learning_rate": 1.590457256461233e-08,
+      "loss": 11.2337,
+      "step": 19
+    },
+    {
+      "epoch": 0.01968503937007874,
+      "grad_norm": 64.9875259399414,
+      "learning_rate": 1.68986083499006e-08,
+      "loss": 11.0572,
+      "step": 20
+    },
+    {
+      "epoch": 0.02066929133858268,
+      "grad_norm": 59.36448287963867,
+      "learning_rate": 1.789264413518887e-08,
+      "loss": 11.0452,
+      "step": 21
+    },
+    {
+      "epoch": 0.021653543307086614,
+      "grad_norm": 59.18537902832031,
+      "learning_rate": 1.888667992047714e-08,
+      "loss": 10.31,
+      "step": 22
+    },
+    {
+      "epoch": 0.022637795275590553,
+      "grad_norm": 54.053524017333984,
+      "learning_rate": 1.988071570576541e-08,
+      "loss": 9.1395,
+      "step": 23
+    },
+    {
+      "epoch": 0.023622047244094488,
+      "grad_norm": 49.979618072509766,
+      "learning_rate": 2.087475149105368e-08,
+      "loss": 8.4201,
+      "step": 24
+    },
+    {
+      "epoch": 0.024606299212598427,
+      "grad_norm": 51.654518127441406,
+      "learning_rate": 2.1868787276341955e-08,
+      "loss": 8.6036,
+      "step": 25
+    },
+    {
+      "epoch": 0.025590551181102362,
+      "grad_norm": 63.55083465576172,
+      "learning_rate": 2.2862823061630224e-08,
+      "loss": 11.7579,
+      "step": 26
+    },
+    {
+      "epoch": 0.0265748031496063,
+      "grad_norm": 59.30263137817383,
+      "learning_rate": 2.3856858846918493e-08,
+      "loss": 10.1307,
+      "step": 27
+    },
+    {
+      "epoch": 0.027559055118110236,
+      "grad_norm": 50.75270462036133,
+      "learning_rate": 2.4850894632206765e-08,
+      "loss": 9.2915,
+      "step": 28
+    },
+    {
+      "epoch": 0.028543307086614175,
+      "grad_norm": 51.7747917175293,
+      "learning_rate": 2.5844930417495034e-08,
+      "loss": 9.0208,
+      "step": 29
+    },
+    {
+      "epoch": 0.02952755905511811,
+      "grad_norm": 46.81666564941406,
+      "learning_rate": 2.6838966202783303e-08,
+      "loss": 8.6867,
+      "step": 30
+    },
+    {
+      "epoch": 0.03051181102362205,
+      "grad_norm": 44.82905578613281,
+      "learning_rate": 2.7833001988071578e-08,
+      "loss": 8.0925,
+      "step": 31
+    },
+    {
+      "epoch": 0.031496062992125984,
+      "grad_norm": 47.148406982421875,
+      "learning_rate": 2.8827037773359847e-08,
+      "loss": 8.6617,
+      "step": 32
+    },
+    {
+      "epoch": 0.03248031496062992,
+      "grad_norm": 47.153053283691406,
+      "learning_rate": 2.982107355864811e-08,
+      "loss": 8.3374,
+      "step": 33
+    },
+    {
+      "epoch": 0.03346456692913386,
+      "grad_norm": 45.4912223815918,
+      "learning_rate": 3.081510934393639e-08,
+      "loss": 7.8566,
+      "step": 34
+    },
+    {
+      "epoch": 0.0344488188976378,
+      "grad_norm": 51.92588806152344,
+      "learning_rate": 3.180914512922466e-08,
+      "loss": 9.0698,
+      "step": 35
+    },
+    {
+      "epoch": 0.03543307086614173,
+      "grad_norm": 43.18679428100586,
+      "learning_rate": 3.280318091451293e-08,
+      "loss": 7.7727,
+      "step": 36
+    },
+    {
+      "epoch": 0.03641732283464567,
+      "grad_norm": 43.12812805175781,
+      "learning_rate": 3.37972166998012e-08,
+      "loss": 7.6128,
+      "step": 37
+    },
+    {
+      "epoch": 0.03740157480314961,
+      "grad_norm": 44.7684211730957,
+      "learning_rate": 3.479125248508947e-08,
+      "loss": 7.8762,
+      "step": 38
+    },
+    {
+      "epoch": 0.038385826771653545,
+      "grad_norm": 36.34556198120117,
+      "learning_rate": 3.578528827037774e-08,
+      "loss": 7.5191,
+      "step": 39
+    },
+    {
+      "epoch": 0.03937007874015748,
+      "grad_norm": 39.9330940246582,
+      "learning_rate": 3.6779324055666005e-08,
+      "loss": 7.5638,
+      "step": 40
+    },
+    {
+      "epoch": 0.040354330708661415,
+      "grad_norm": 39.49993133544922,
+      "learning_rate": 3.777335984095428e-08,
+      "loss": 7.1878,
+      "step": 41
+    },
+    {
+      "epoch": 0.04133858267716536,
+      "grad_norm": 34.661251068115234,
+      "learning_rate": 3.8767395626242556e-08,
+      "loss": 6.8878,
+      "step": 42
+    },
+    {
+      "epoch": 0.04232283464566929,
+      "grad_norm": 40.655540466308594,
+      "learning_rate": 3.976143141153082e-08,
+      "loss": 7.5775,
+      "step": 43
+    },
+    {
+      "epoch": 0.04330708661417323,
+      "grad_norm": 36.12670135498047,
+      "learning_rate": 4.0755467196819094e-08,
+      "loss": 7.1076,
+      "step": 44
+    },
+    {
+      "epoch": 0.04429133858267716,
+      "grad_norm": 34.178226470947266,
+      "learning_rate": 4.174950298210736e-08,
+      "loss": 6.5589,
+      "step": 45
+    },
+    {
+      "epoch": 0.045275590551181105,
+      "grad_norm": 36.6577033996582,
+      "learning_rate": 4.274353876739563e-08,
+      "loss": 7.4456,
+      "step": 46
+    },
+    {
+      "epoch": 0.04625984251968504,
+      "grad_norm": 32.548709869384766,
+      "learning_rate": 4.373757455268391e-08,
+      "loss": 6.8233,
+      "step": 47
+    },
+    {
+      "epoch": 0.047244094488188976,
+      "grad_norm": 34.000553131103516,
+      "learning_rate": 4.4731610337972176e-08,
+      "loss": 6.7633,
+      "step": 48
+    },
+    {
+      "epoch": 0.04822834645669291,
+      "grad_norm": 32.247859954833984,
+      "learning_rate": 4.572564612326045e-08,
+      "loss": 6.6024,
+      "step": 49
+    },
+    {
+      "epoch": 0.04921259842519685,
+      "grad_norm": 28.947786331176758,
+      "learning_rate": 4.6719681908548713e-08,
+      "loss": 6.2778,
+      "step": 50
+    },
+    {
+      "epoch": 0.05019685039370079,
+      "grad_norm": 30.279062271118164,
+      "learning_rate": 4.7713717693836986e-08,
+      "loss": 6.1026,
+      "step": 51
+    },
+    {
+      "epoch": 0.051181102362204724,
+      "grad_norm": 31.13785171508789,
+      "learning_rate": 4.870775347912525e-08,
+      "loss": 6.632,
+      "step": 52
+    },
+    {
+      "epoch": 0.05216535433070866,
+      "grad_norm": 29.648237228393555,
+      "learning_rate": 4.970178926441353e-08,
+      "loss": 6.6962,
+      "step": 53
+    },
+    {
+      "epoch": 0.0531496062992126,
+      "grad_norm": 28.224645614624023,
+      "learning_rate": 5.06958250497018e-08,
+      "loss": 5.8514,
+      "step": 54
+    },
+    {
+      "epoch": 0.054133858267716536,
+      "grad_norm": 28.693328857421875,
+      "learning_rate": 5.168986083499007e-08,
+      "loss": 5.9951,
+      "step": 55
+    },
+    {
+      "epoch": 0.05511811023622047,
+      "grad_norm": 24.777812957763672,
+      "learning_rate": 5.268389662027834e-08,
+      "loss": 5.4554,
+      "step": 56
+    },
+    {
+      "epoch": 0.05610236220472441,
+      "grad_norm": 26.11226463317871,
+      "learning_rate": 5.3677932405566605e-08,
+      "loss": 6.0147,
+      "step": 57
+    },
+    {
+      "epoch": 0.05708661417322835,
+      "grad_norm": 24.698375701904297,
+      "learning_rate": 5.467196819085488e-08,
+      "loss": 5.215,
+      "step": 58
+    },
+    {
+      "epoch": 0.058070866141732284,
+      "grad_norm": 26.616317749023438,
+      "learning_rate": 5.5666003976143156e-08,
+      "loss": 6.4525,
+      "step": 59
+    },
+    {
+      "epoch": 0.05905511811023622,
+      "grad_norm": 26.09321403503418,
+      "learning_rate": 5.666003976143142e-08,
+      "loss": 5.4048,
+      "step": 60
+    },
+    {
+      "epoch": 0.060039370078740155,
+      "grad_norm": 19.83713150024414,
+      "learning_rate": 5.7654075546719694e-08,
+      "loss": 5.0424,
+      "step": 61
+    },
+    {
+      "epoch": 0.0610236220472441,
+      "grad_norm": 26.56923484802246,
+      "learning_rate": 5.864811133200796e-08,
+      "loss": 6.2646,
+      "step": 62
+    },
+    {
+      "epoch": 0.06200787401574803,
+      "grad_norm": 23.580089569091797,
+      "learning_rate": 5.964214711729623e-08,
+      "loss": 5.0847,
+      "step": 63
+    },
+    {
+      "epoch": 0.06299212598425197,
+      "grad_norm": 23.453126907348633,
+      "learning_rate": 6.06361829025845e-08,
+      "loss": 5.4415,
+      "step": 64
+    },
+    {
+      "epoch": 0.0639763779527559,
+      "grad_norm": 21.229190826416016,
+      "learning_rate": 6.163021868787278e-08,
+      "loss": 5.2469,
+      "step": 65
+    },
+    {
+      "epoch": 0.06496062992125984,
+      "grad_norm": 19.477190017700195,
+      "learning_rate": 6.262425447316104e-08,
+      "loss": 5.1378,
+      "step": 66
+    },
+    {
+      "epoch": 0.06594488188976377,
+      "grad_norm": 19.40647315979004,
+      "learning_rate": 6.361829025844931e-08,
+      "loss": 5.1636,
+      "step": 67
+    },
+    {
+      "epoch": 0.06692913385826772,
+      "grad_norm": 22.20977210998535,
+      "learning_rate": 6.461232604373759e-08,
+      "loss": 5.5596,
+      "step": 68
+    },
+    {
+      "epoch": 0.06791338582677166,
+      "grad_norm": 19.186826705932617,
+      "learning_rate": 6.560636182902586e-08,
+      "loss": 4.9508,
+      "step": 69
+    },
+    {
+      "epoch": 0.0688976377952756,
+      "grad_norm": 20.190908432006836,
+      "learning_rate": 6.660039761431412e-08,
+      "loss": 5.2355,
+      "step": 70
+    },
+    {
+      "epoch": 0.06988188976377953,
+      "grad_norm": 18.122196197509766,
+      "learning_rate": 6.75944333996024e-08,
+      "loss": 4.7359,
+      "step": 71
+    },
+    {
+      "epoch": 0.07086614173228346,
+      "grad_norm": 17.524765014648438,
+      "learning_rate": 6.858846918489067e-08,
+      "loss": 4.8947,
+      "step": 72
+    },
+    {
+      "epoch": 0.0718503937007874,
+      "grad_norm": 18.821767807006836,
+      "learning_rate": 6.958250497017893e-08,
+      "loss": 4.6269,
+      "step": 73
+    },
+    {
+      "epoch": 0.07283464566929133,
+      "grad_norm": 18.19922637939453,
+      "learning_rate": 7.057654075546721e-08,
+      "loss": 4.6072,
+      "step": 74
+    },
+    {
+      "epoch": 0.07381889763779527,
+      "grad_norm": 16.908899307250977,
+      "learning_rate": 7.157057654075548e-08,
+      "loss": 4.9125,
+      "step": 75
+    },
+    {
+      "epoch": 0.07480314960629922,
+      "grad_norm": 19.90263557434082,
+      "learning_rate": 7.256461232604374e-08,
+      "loss": 4.5856,
+      "step": 76
+    },
+    {
+      "epoch": 0.07578740157480315,
+      "grad_norm": 17.92584800720215,
+      "learning_rate": 7.355864811133201e-08,
+      "loss": 4.7879,
+      "step": 77
+    },
+    {
+      "epoch": 0.07677165354330709,
+      "grad_norm": 16.29261589050293,
+      "learning_rate": 7.455268389662029e-08,
+      "loss": 4.5348,
+      "step": 78
+    },
+    {
+      "epoch": 0.07775590551181102,
+      "grad_norm": 16.3350887298584,
+      "learning_rate": 7.554671968190855e-08,
+      "loss": 4.3554,
+      "step": 79
+    },
+    {
+      "epoch": 0.07874015748031496,
+      "grad_norm": 14.408184051513672,
+      "learning_rate": 7.654075546719683e-08,
+      "loss": 4.2984,
+      "step": 80
+    },
+    {
+      "epoch": 0.0797244094488189,
+      "grad_norm": 16.71326446533203,
+      "learning_rate": 7.753479125248511e-08,
+      "loss": 4.5505,
+      "step": 81
+    },
+    {
+      "epoch": 0.08070866141732283,
+      "grad_norm": 14.62590217590332,
+      "learning_rate": 7.852882703777338e-08,
+      "loss": 4.5325,
+      "step": 82
+    },
+    {
+      "epoch": 0.08169291338582677,
+      "grad_norm": 17.189268112182617,
+      "learning_rate": 7.952286282306164e-08,
+      "loss": 4.2725,
+      "step": 83
+    },
+    {
+      "epoch": 0.08267716535433071,
+      "grad_norm": 16.960248947143555,
+      "learning_rate": 8.051689860834992e-08,
+      "loss": 4.3054,
+      "step": 84
+    },
+    {
+      "epoch": 0.08366141732283465,
+      "grad_norm": 15.114398956298828,
+      "learning_rate": 8.151093439363819e-08,
+      "loss": 4.5536,
+      "step": 85
+    },
+    {
+      "epoch": 0.08464566929133858,
+      "grad_norm": 16.153371810913086,
+      "learning_rate": 8.250497017892645e-08,
+      "loss": 4.0265,
+      "step": 86
+    },
+    {
+      "epoch": 0.08562992125984252,
+      "grad_norm": 15.731820106506348,
+      "learning_rate": 8.349900596421472e-08,
+      "loss": 4.7453,
+      "step": 87
+    },
+    {
+      "epoch": 0.08661417322834646,
+      "grad_norm": 14.69382381439209,
+      "learning_rate": 8.4493041749503e-08,
+      "loss": 4.071,
+      "step": 88
+    },
+    {
+      "epoch": 0.08759842519685039,
+      "grad_norm": 13.735575675964355,
+      "learning_rate": 8.548707753479126e-08,
+      "loss": 4.1582,
+      "step": 89
+    },
+    {
+      "epoch": 0.08858267716535433,
+      "grad_norm": 16.017065048217773,
+      "learning_rate": 8.648111332007953e-08,
+      "loss": 4.1131,
+      "step": 90
+    },
+    {
+      "epoch": 0.08956692913385826,
+      "grad_norm": 17.237276077270508,
+      "learning_rate": 8.747514910536782e-08,
+      "loss": 3.6582,
+      "step": 91
+    },
+    {
+      "epoch": 0.09055118110236221,
+      "grad_norm": 15.59334945678711,
+      "learning_rate": 8.846918489065609e-08,
+      "loss": 4.143,
+      "step": 92
+    },
+    {
+      "epoch": 0.09153543307086615,
+      "grad_norm": 14.918270111083984,
+      "learning_rate": 8.946322067594435e-08,
+      "loss": 4.2273,
+      "step": 93
+    },
+    {
+      "epoch": 0.09251968503937008,
+      "grad_norm": 14.899909019470215,
+      "learning_rate": 9.045725646123262e-08,
+      "loss": 3.9321,
+      "step": 94
+    },
+    {
+      "epoch": 0.09350393700787402,
+      "grad_norm": 18.112892150878906,
+      "learning_rate": 9.14512922465209e-08,
+      "loss": 4.2061,
+      "step": 95
+    },
+    {
+      "epoch": 0.09448818897637795,
+      "grad_norm": 15.854629516601562,
+      "learning_rate": 9.244532803180916e-08,
+      "loss": 4.1042,
+      "step": 96
+    },
+    {
+      "epoch": 0.09547244094488189,
+      "grad_norm": 16.44801139831543,
+      "learning_rate": 9.343936381709743e-08,
+      "loss": 3.9513,
+      "step": 97
+    },
+    {
+      "epoch": 0.09645669291338582,
+      "grad_norm": 14.854127883911133,
+      "learning_rate": 9.44333996023857e-08,
+      "loss": 3.8627,
+      "step": 98
+    },
+    {
+      "epoch": 0.09744094488188976,
+      "grad_norm": 17.02035903930664,
+      "learning_rate": 9.542743538767397e-08,
+      "loss": 4.3613,
+      "step": 99
+    },
+    {
+      "epoch": 0.0984251968503937,
+      "grad_norm": 15.354338645935059,
+      "learning_rate": 9.642147117296224e-08,
+      "loss": 3.8513,
+      "step": 100
+    },
+    {
+      "epoch": 0.09940944881889764,
+      "grad_norm": 16.35565757751465,
+      "learning_rate": 9.74155069582505e-08,
+      "loss": 3.5866,
+      "step": 101
+    },
+    {
+      "epoch": 0.10039370078740158,
+      "grad_norm": 16.66194725036621,
+      "learning_rate": 9.840954274353878e-08,
+      "loss": 3.5239,
+      "step": 102
+    },
+    {
+      "epoch": 0.10137795275590551,
+      "grad_norm": 15.875446319580078,
+      "learning_rate": 9.940357852882706e-08,
+      "loss": 3.5921,
+      "step": 103
+    },
+    {
+      "epoch": 0.10236220472440945,
+      "grad_norm": 14.344114303588867,
+      "learning_rate": 1.0039761431411533e-07,
+      "loss": 3.5962,
+      "step": 104
+    },
+    {
+      "epoch": 0.10334645669291338,
+      "grad_norm": 18.503963470458984,
+      "learning_rate": 1.013916500994036e-07,
+      "loss": 4.0001,
+      "step": 105
+    },
+    {
+      "epoch": 0.10433070866141732,
+      "grad_norm": 16.944435119628906,
+      "learning_rate": 1.0238568588469187e-07,
+      "loss": 4.1374,
+      "step": 106
+    },
+    {
+      "epoch": 0.10531496062992125,
+      "grad_norm": 16.46833038330078,
+      "learning_rate": 1.0337972166998014e-07,
+      "loss": 3.9049,
+      "step": 107
+    },
+    {
+      "epoch": 0.1062992125984252,
+      "grad_norm": 14.921700477600098,
+      "learning_rate": 1.043737574552684e-07,
+      "loss": 3.2511,
+      "step": 108
+    },
+    {
+      "epoch": 0.10728346456692914,
+      "grad_norm": 15.574972152709961,
+      "learning_rate": 1.0536779324055668e-07,
+      "loss": 3.2479,
+      "step": 109
+    },
+    {
+      "epoch": 0.10826771653543307,
+      "grad_norm": 16.810884475708008,
+      "learning_rate": 1.0636182902584495e-07,
+      "loss": 3.6414,
+      "step": 110
+    },
+    {
+      "epoch": 0.10925196850393701,
+      "grad_norm": 17.074661254882812,
+      "learning_rate": 1.0735586481113321e-07,
+      "loss": 3.6429,
+      "step": 111
+    },
+    {
+      "epoch": 0.11023622047244094,
+      "grad_norm": 18.52947235107422,
+      "learning_rate": 1.0834990059642149e-07,
+      "loss": 3.423,
+      "step": 112
+    },
+    {
+      "epoch": 0.11122047244094488,
+      "grad_norm": 18.681869506835938,
+      "learning_rate": 1.0934393638170976e-07,
+      "loss": 3.4967,
+      "step": 113
+    },
+    {
+      "epoch": 0.11220472440944881,
+      "grad_norm": 20.385292053222656,
+      "learning_rate": 1.1033797216699802e-07,
+      "loss": 3.7649,
+      "step": 114
+    },
+    {
+      "epoch": 0.11318897637795275,
+      "grad_norm": 18.912538528442383,
+      "learning_rate": 1.1133200795228631e-07,
+      "loss": 3.2845,
+      "step": 115
+    },
+    {
+      "epoch": 0.1141732283464567,
+      "grad_norm": 17.856229782104492,
+      "learning_rate": 1.1232604373757458e-07,
+      "loss": 3.356,
+      "step": 116
+    },
+    {
+      "epoch": 0.11515748031496063,
+      "grad_norm": 17.08562469482422,
+      "learning_rate": 1.1332007952286284e-07,
+      "loss": 3.2086,
+      "step": 117
+    },
+    {
+      "epoch": 0.11614173228346457,
+      "grad_norm": 17.54237937927246,
+      "learning_rate": 1.1431411530815111e-07,
+      "loss": 3.5561,
+      "step": 118
+    },
+    {
+      "epoch": 0.1171259842519685,
+      "grad_norm": 19.936498641967773,
+      "learning_rate": 1.1530815109343939e-07,
+      "loss": 3.7353,
+      "step": 119
+    },
+    {
+      "epoch": 0.11811023622047244,
+      "grad_norm": 17.135496139526367,
+      "learning_rate": 1.1630218687872765e-07,
+      "loss": 3.403,
+      "step": 120
+    },
+    {
+      "epoch": 0.11909448818897637,
+      "grad_norm": 17.260093688964844,
+      "learning_rate": 1.1729622266401592e-07,
+      "loss": 3.1009,
+      "step": 121
+    },
+    {
+      "epoch": 0.12007874015748031,
+      "grad_norm": 17.075611114501953,
+      "learning_rate": 1.182902584493042e-07,
+      "loss": 3.2139,
+      "step": 122
+    },
+    {
+      "epoch": 0.12106299212598425,
+      "grad_norm": 23.433874130249023,
+      "learning_rate": 1.1928429423459245e-07,
+      "loss": 3.3339,
+      "step": 123
+    },
+    {
+      "epoch": 0.1220472440944882,
+      "grad_norm": 18.25501251220703,
+      "learning_rate": 1.2027833001988073e-07,
+      "loss": 2.9464,
+      "step": 124
+    },
+    {
+      "epoch": 0.12303149606299213,
+      "grad_norm": 18.079578399658203,
+      "learning_rate": 1.21272365805169e-07,
+      "loss": 3.3366,
+      "step": 125
+    },
+    {
+      "epoch": 0.12401574803149606,
+      "grad_norm": 16.392736434936523,
+      "learning_rate": 1.222664015904573e-07,
+      "loss": 3.0618,
+      "step": 126
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 15.782499313354492,
+      "learning_rate": 1.2326043737574557e-07,
+      "loss": 3.0092,
+      "step": 127
+    },
+    {
+      "epoch": 0.12598425196850394,
+      "grad_norm": 14.74819278717041,
+      "learning_rate": 1.2425447316103382e-07,
+      "loss": 2.7152,
+      "step": 128
+    },
+    {
+      "epoch": 0.12696850393700787,
+      "grad_norm": 17.743946075439453,
+      "learning_rate": 1.2524850894632207e-07,
+      "loss": 2.9423,
+      "step": 129
+    },
+    {
+      "epoch": 0.1279527559055118,
+      "grad_norm": 15.759135246276855,
+      "learning_rate": 1.2624254473161035e-07,
+      "loss": 2.6569,
+      "step": 130
+    },
+    {
+      "epoch": 0.12893700787401574,
+      "grad_norm": 18.54253387451172,
+      "learning_rate": 1.2723658051689863e-07,
+      "loss": 2.8469,
+      "step": 131
+    },
+    {
+      "epoch": 0.12992125984251968,
+      "grad_norm": 18.318775177001953,
+      "learning_rate": 1.282306163021869e-07,
+      "loss": 2.9089,
+      "step": 132
+    },
+    {
+      "epoch": 0.1309055118110236,
+      "grad_norm": 17.751266479492188,
+      "learning_rate": 1.2922465208747519e-07,
+      "loss": 2.5809,
+      "step": 133
+    },
+    {
+      "epoch": 0.13188976377952755,
+      "grad_norm": 21.29975128173828,
+      "learning_rate": 1.3021868787276344e-07,
+      "loss": 2.6987,
+      "step": 134
+    },
+    {
+      "epoch": 0.1328740157480315,
+      "grad_norm": 19.9519100189209,
+      "learning_rate": 1.3121272365805172e-07,
+      "loss": 3.2518,
+      "step": 135
+    },
+    {
+      "epoch": 0.13385826771653545,
+      "grad_norm": 19.99887466430664,
+      "learning_rate": 1.3220675944333997e-07,
+      "loss": 2.9145,
+      "step": 136
+    },
+    {
+      "epoch": 0.13484251968503938,
+      "grad_norm": 17.0355167388916,
+      "learning_rate": 1.3320079522862825e-07,
+      "loss": 2.4809,
+      "step": 137
+    },
+    {
+      "epoch": 0.13582677165354332,
+      "grad_norm": 18.628467559814453,
+      "learning_rate": 1.3419483101391653e-07,
+      "loss": 2.8264,
+      "step": 138
+    },
+    {
+      "epoch": 0.13681102362204725,
+      "grad_norm": 19.511394500732422,
+      "learning_rate": 1.351888667992048e-07,
+      "loss": 2.5724,
+      "step": 139
+    },
+    {
+      "epoch": 0.1377952755905512,
+      "grad_norm": 21.264480590820312,
+      "learning_rate": 1.3618290258449306e-07,
+      "loss": 2.6949,
+      "step": 140
+    },
+    {
+      "epoch": 0.13877952755905512,
+      "grad_norm": 19.76134490966797,
+      "learning_rate": 1.3717693836978134e-07,
+      "loss": 2.6925,
+      "step": 141
+    },
+    {
+      "epoch": 0.13976377952755906,
+      "grad_norm": 20.930673599243164,
+      "learning_rate": 1.381709741550696e-07,
+      "loss": 2.9311,
+      "step": 142
+    },
+    {
+      "epoch": 0.140748031496063,
+      "grad_norm": 21.966018676757812,
+      "learning_rate": 1.3916500994035787e-07,
+      "loss": 2.5667,
+      "step": 143
+    },
+    {
+      "epoch": 0.14173228346456693,
+      "grad_norm": 21.916505813598633,
+      "learning_rate": 1.4015904572564615e-07,
+      "loss": 3.2471,
+      "step": 144
+    },
+    {
+      "epoch": 0.14271653543307086,
+      "grad_norm": 20.081771850585938,
+      "learning_rate": 1.4115308151093443e-07,
+      "loss": 2.2441,
+      "step": 145
+    },
+    {
+      "epoch": 0.1437007874015748,
+      "grad_norm": 22.893489837646484,
+      "learning_rate": 1.421471172962227e-07,
+      "loss": 2.75,
+      "step": 146
+    },
+    {
+      "epoch": 0.14468503937007873,
+      "grad_norm": 23.95358657836914,
+      "learning_rate": 1.4314115308151096e-07,
+      "loss": 2.9669,
+      "step": 147
+    },
+    {
+      "epoch": 0.14566929133858267,
+      "grad_norm": 21.101062774658203,
+      "learning_rate": 1.4413518886679924e-07,
+      "loss": 2.736,
+      "step": 148
+    },
+    {
+      "epoch": 0.1466535433070866,
+      "grad_norm": 25.240341186523438,
+      "learning_rate": 1.451292246520875e-07,
+      "loss": 3.104,
+      "step": 149
+    },
+    {
+      "epoch": 0.14763779527559054,
+      "grad_norm": 18.358688354492188,
+      "learning_rate": 1.4612326043737577e-07,
+      "loss": 2.2175,
+      "step": 150
+    },
+    {
+      "epoch": 0.1486220472440945,
+      "grad_norm": 21.986661911010742,
+      "learning_rate": 1.4711729622266402e-07,
+      "loss": 2.7415,
+      "step": 151
+    },
+    {
+      "epoch": 0.14960629921259844,
+      "grad_norm": 20.64093017578125,
+      "learning_rate": 1.4811133200795232e-07,
+      "loss": 1.8707,
+      "step": 152
+    },
+    {
+      "epoch": 0.15059055118110237,
+      "grad_norm": 20.602142333984375,
+      "learning_rate": 1.4910536779324058e-07,
+      "loss": 2.5961,
+      "step": 153
+    },
+    {
+      "epoch": 0.15059055118110237,
+      "eval_Qnli-dev_cosine_accuracy": 0.634765625,
+      "eval_Qnli-dev_cosine_accuracy_threshold": 0.8508153557777405,
+      "eval_Qnli-dev_cosine_ap": 0.6461335447626624,
+      "eval_Qnli-dev_cosine_f1": 0.6505636070853462,
+      "eval_Qnli-dev_cosine_f1_threshold": 0.7770615816116333,
+      "eval_Qnli-dev_cosine_precision": 0.5246753246753246,
+      "eval_Qnli-dev_cosine_recall": 0.8559322033898306,
+      "eval_Qnli-dev_dot_accuracy": 0.634765625,
+      "eval_Qnli-dev_dot_accuracy_threshold": 653.7443237304688,
+      "eval_Qnli-dev_dot_ap": 0.6461682282377894,
+      "eval_Qnli-dev_dot_f1": 0.6505636070853462,
+      "eval_Qnli-dev_dot_f1_threshold": 597.0731811523438,
+      "eval_Qnli-dev_dot_precision": 0.5246753246753246,
+      "eval_Qnli-dev_dot_recall": 0.8559322033898306,
+      "eval_Qnli-dev_euclidean_accuracy": 0.634765625,
+      "eval_Qnli-dev_euclidean_accuracy_threshold": 15.141305923461914,
+      "eval_Qnli-dev_euclidean_ap": 0.6461382925406688,
+      "eval_Qnli-dev_euclidean_f1": 0.6505636070853462,
+      "eval_Qnli-dev_euclidean_f1_threshold": 18.50943946838379,
+      "eval_Qnli-dev_euclidean_precision": 0.5246753246753246,
+      "eval_Qnli-dev_euclidean_recall": 0.8559322033898306,
+      "eval_Qnli-dev_manhattan_accuracy": 0.6328125,
+      "eval_Qnli-dev_manhattan_accuracy_threshold": 331.46282958984375,
+      "eval_Qnli-dev_manhattan_ap": 0.6431949026371255,
+      "eval_Qnli-dev_manhattan_f1": 0.6501650165016502,
+      "eval_Qnli-dev_manhattan_f1_threshold": 404.6050109863281,
+      "eval_Qnli-dev_manhattan_precision": 0.5324324324324324,
+      "eval_Qnli-dev_manhattan_recall": 0.8347457627118644,
+      "eval_Qnli-dev_max_accuracy": 0.634765625,
+      "eval_Qnli-dev_max_accuracy_threshold": 653.7443237304688,
+      "eval_Qnli-dev_max_ap": 0.6461682282377894,
+      "eval_Qnli-dev_max_f1": 0.6505636070853462,
+      "eval_Qnli-dev_max_f1_threshold": 597.0731811523438,
+      "eval_Qnli-dev_max_precision": 0.5324324324324324,
+      "eval_Qnli-dev_max_recall": 0.8559322033898306,
+      "eval_allNLI-dev_cosine_accuracy": 0.67578125,
+      "eval_allNLI-dev_cosine_accuracy_threshold": 0.9452645182609558,
+      "eval_allNLI-dev_cosine_ap": 0.4264736612515921,
+      "eval_allNLI-dev_cosine_f1": 0.512,
+      "eval_allNLI-dev_cosine_f1_threshold": 0.8565204739570618,
+      "eval_allNLI-dev_cosine_precision": 0.39143730886850153,
+      "eval_allNLI-dev_cosine_recall": 0.7398843930635838,
+      "eval_allNLI-dev_dot_accuracy": 0.67578125,
+      "eval_allNLI-dev_dot_accuracy_threshold": 726.30615234375,
+      "eval_allNLI-dev_dot_ap": 0.42647535250956575,
+      "eval_allNLI-dev_dot_f1": 0.512,
+      "eval_allNLI-dev_dot_f1_threshold": 658.1103515625,
+      "eval_allNLI-dev_dot_precision": 0.39143730886850153,
+      "eval_allNLI-dev_dot_recall": 0.7398843930635838,
+      "eval_allNLI-dev_euclidean_accuracy": 0.67578125,
+      "eval_allNLI-dev_euclidean_accuracy_threshold": 9.171283721923828,
+      "eval_allNLI-dev_euclidean_ap": 0.4264736612515921,
+      "eval_allNLI-dev_euclidean_f1": 0.512,
+      "eval_allNLI-dev_euclidean_f1_threshold": 14.84876823425293,
+      "eval_allNLI-dev_euclidean_precision": 0.39143730886850153,
+      "eval_allNLI-dev_euclidean_recall": 0.7398843930635838,
+      "eval_allNLI-dev_manhattan_accuracy": 0.67578125,
+      "eval_allNLI-dev_manhattan_accuracy_threshold": 201.49061584472656,
+      "eval_allNLI-dev_manhattan_ap": 0.4252213828672732,
+      "eval_allNLI-dev_manhattan_f1": 0.5107692307692308,
+      "eval_allNLI-dev_manhattan_f1_threshold": 417.52728271484375,
+      "eval_allNLI-dev_manhattan_precision": 0.3480083857442348,
+      "eval_allNLI-dev_manhattan_recall": 0.9595375722543352,
+      "eval_allNLI-dev_max_accuracy": 0.67578125,
+      "eval_allNLI-dev_max_accuracy_threshold": 726.30615234375,
+      "eval_allNLI-dev_max_ap": 0.42647535250956575,
+      "eval_allNLI-dev_max_f1": 0.512,
+      "eval_allNLI-dev_max_f1_threshold": 658.1103515625,
+      "eval_allNLI-dev_max_precision": 0.39143730886850153,
+      "eval_allNLI-dev_max_recall": 0.9595375722543352,
+      "eval_loss": 2.2652623653411865,
+      "eval_runtime": 50.7627,
+      "eval_samples_per_second": 32.78,
+      "eval_sequential_score": 0.6461682282377894,
+      "eval_steps_per_second": 0.138,
+      "eval_sts-test_pearson_cosine": 0.2749904272806095,
+      "eval_sts-test_pearson_dot": 0.27496363262371837,
+      "eval_sts-test_pearson_euclidean": 0.2934483033082174,
+      "eval_sts-test_pearson_manhattan": 0.2923996087310511,
+      "eval_sts-test_pearson_max": 0.2934483033082174,
+      "eval_sts-test_spearman_cosine": 0.31159390381099095,
+      "eval_sts-test_spearman_dot": 0.31138581044552094,
+      "eval_sts-test_spearman_euclidean": 0.3115817314678925,
+      "eval_sts-test_spearman_manhattan": 0.3095556181083969,
+      "eval_sts-test_spearman_max": 0.31159390381099095,
+      "step": 153
+    },
+    {
+      "epoch": 0.1515748031496063,
+      "grad_norm": 22.330442428588867,
+      "learning_rate": 1.5009940357852886e-07,
+      "loss": 3.1149,
+      "step": 154
+    },
+    {
+      "epoch": 0.15255905511811024,
+      "grad_norm": 23.656953811645508,
+      "learning_rate": 1.510934393638171e-07,
+      "loss": 2.2976,
+      "step": 155
+    },
+    {
+      "epoch": 0.15354330708661418,
+      "grad_norm": 20.271608352661133,
+      "learning_rate": 1.5208747514910539e-07,
+      "loss": 2.4436,
+      "step": 156
+    },
+    {
+      "epoch": 0.1545275590551181,
+      "grad_norm": 25.410293579101562,
+      "learning_rate": 1.5308151093439367e-07,
+      "loss": 2.8826,
+      "step": 157
+    },
+    {
+      "epoch": 0.15551181102362205,
+      "grad_norm": 23.772783279418945,
+      "learning_rate": 1.5407554671968192e-07,
+      "loss": 2.3664,
+      "step": 158
+    },
+    {
+      "epoch": 0.15649606299212598,
+      "grad_norm": 23.44937515258789,
+      "learning_rate": 1.5506958250497022e-07,
+      "loss": 2.2485,
+      "step": 159
+    },
+    {
+      "epoch": 0.15748031496062992,
+      "grad_norm": 23.024261474609375,
+      "learning_rate": 1.5606361829025848e-07,
+      "loss": 2.5167,
+      "step": 160
+    },
+    {
+      "epoch": 0.15846456692913385,
+      "grad_norm": 20.63090705871582,
+      "learning_rate": 1.5705765407554675e-07,
+      "loss": 1.7183,
+      "step": 161
+    },
+    {
+      "epoch": 0.1594488188976378,
+      "grad_norm": 21.573190689086914,
+      "learning_rate": 1.58051689860835e-07,
+      "loss": 2.1003,
+      "step": 162
+    },
+    {
+      "epoch": 0.16043307086614172,
+      "grad_norm": 23.774974822998047,
+      "learning_rate": 1.5904572564612329e-07,
+      "loss": 2.5785,
+      "step": 163
+    },
+    {
+      "epoch": 0.16141732283464566,
+      "grad_norm": 27.151123046875,
+      "learning_rate": 1.6003976143141154e-07,
+      "loss": 2.8789,
+      "step": 164
+    },
+    {
+      "epoch": 0.1624015748031496,
+      "grad_norm": 23.50958824157715,
+      "learning_rate": 1.6103379721669984e-07,
+      "loss": 2.3425,
+      "step": 165
+    },
+    {
+      "epoch": 0.16338582677165353,
+      "grad_norm": 23.77661895751953,
+      "learning_rate": 1.620278330019881e-07,
+      "loss": 2.0966,
+      "step": 166
+    },
+    {
+      "epoch": 0.1643700787401575,
+      "grad_norm": 22.070526123046875,
+      "learning_rate": 1.6302186878727637e-07,
+      "loss": 2.1126,
+      "step": 167
+    },
+    {
+      "epoch": 0.16535433070866143,
+      "grad_norm": 22.653602600097656,
+      "learning_rate": 1.6401590457256465e-07,
+      "loss": 2.1824,
+      "step": 168
+    },
+    {
+      "epoch": 0.16633858267716536,
+      "grad_norm": 21.470808029174805,
+      "learning_rate": 1.650099403578529e-07,
+      "loss": 2.2009,
+      "step": 169
+    },
+    {
+      "epoch": 0.1673228346456693,
+      "grad_norm": 25.822694778442383,
+      "learning_rate": 1.6600397614314118e-07,
+      "loss": 2.3796,
+      "step": 170
+    },
+    {
+      "epoch": 0.16830708661417323,
+      "grad_norm": 22.609458923339844,
+      "learning_rate": 1.6699801192842944e-07,
+      "loss": 2.3096,
+      "step": 171
+    },
+    {
+      "epoch": 0.16929133858267717,
+      "grad_norm": 24.10075569152832,
+      "learning_rate": 1.6799204771371774e-07,
+      "loss": 2.7897,
+      "step": 172
+    },
+    {
+      "epoch": 0.1702755905511811,
+      "grad_norm": 22.21641731262207,
+      "learning_rate": 1.68986083499006e-07,
+      "loss": 2.2097,
+      "step": 173
+    },
+    {
+      "epoch": 0.17125984251968504,
+      "grad_norm": 17.717933654785156,
+      "learning_rate": 1.6998011928429427e-07,
+      "loss": 1.7508,
+      "step": 174
+    },
+    {
+      "epoch": 0.17224409448818898,
+      "grad_norm": 22.352798461914062,
+      "learning_rate": 1.7097415506958253e-07,
+      "loss": 2.353,
+      "step": 175
+    },
+    {
+      "epoch": 0.1732283464566929,
+      "grad_norm": 23.421472549438477,
+      "learning_rate": 1.719681908548708e-07,
+      "loss": 2.4276,
+      "step": 176
+    },
+    {
+      "epoch": 0.17421259842519685,
+      "grad_norm": 20.41706657409668,
+      "learning_rate": 1.7296222664015906e-07,
+      "loss": 2.1016,
+      "step": 177
+    },
+    {
+      "epoch": 0.17519685039370078,
+      "grad_norm": 19.39253807067871,
+      "learning_rate": 1.7395626242544734e-07,
+      "loss": 1.8461,
+      "step": 178
+    },
+    {
+      "epoch": 0.17618110236220472,
+      "grad_norm": 19.994935989379883,
+      "learning_rate": 1.7495029821073564e-07,
+      "loss": 1.8104,
+      "step": 179
+    },
+    {
+      "epoch": 0.17716535433070865,
+      "grad_norm": 24.119571685791016,
+      "learning_rate": 1.759443339960239e-07,
+      "loss": 2.6023,
+      "step": 180
+    },
+    {
+      "epoch": 0.1781496062992126,
+      "grad_norm": 28.111419677734375,
+      "learning_rate": 1.7693836978131217e-07,
+      "loss": 2.5261,
+      "step": 181
+    },
+    {
+      "epoch": 0.17913385826771652,
+      "grad_norm": 26.244142532348633,
+      "learning_rate": 1.7793240556660042e-07,
+      "loss": 2.1053,
+      "step": 182
+    },
+    {
+      "epoch": 0.18011811023622049,
+      "grad_norm": 24.29507827758789,
+      "learning_rate": 1.789264413518887e-07,
+      "loss": 1.9712,
+      "step": 183
+    },
+    {
+      "epoch": 0.18110236220472442,
+      "grad_norm": 26.753236770629883,
+      "learning_rate": 1.7992047713717695e-07,
+      "loss": 2.4693,
+      "step": 184
+    },
+    {
+      "epoch": 0.18208661417322836,
+      "grad_norm": 23.031953811645508,
+      "learning_rate": 1.8091451292246523e-07,
+      "loss": 2.1119,
+      "step": 185
+    },
+    {
+      "epoch": 0.1830708661417323,
+      "grad_norm": 24.926044464111328,
+      "learning_rate": 1.819085487077535e-07,
+      "loss": 2.4797,
+      "step": 186
+    },
+    {
+      "epoch": 0.18405511811023623,
+      "grad_norm": 19.047605514526367,
+      "learning_rate": 1.829025844930418e-07,
+      "loss": 2.1587,
+      "step": 187
+    },
+    {
+      "epoch": 0.18503937007874016,
+      "grad_norm": 26.785459518432617,
+      "learning_rate": 1.8389662027833004e-07,
+      "loss": 1.9578,
+      "step": 188
+    },
+    {
+      "epoch": 0.1860236220472441,
+      "grad_norm": 22.257556915283203,
+      "learning_rate": 1.8489065606361832e-07,
+      "loss": 2.1368,
+      "step": 189
+    },
+    {
+      "epoch": 0.18700787401574803,
+      "grad_norm": 24.006427764892578,
+      "learning_rate": 1.8588469184890657e-07,
+      "loss": 2.4212,
+      "step": 190
+    },
+    {
+      "epoch": 0.18799212598425197,
+      "grad_norm": 22.14805793762207,
+      "learning_rate": 1.8687872763419485e-07,
+      "loss": 1.9591,
+      "step": 191
+    },
+    {
+      "epoch": 0.1889763779527559,
+      "grad_norm": 19.438581466674805,
+      "learning_rate": 1.8787276341948313e-07,
+      "loss": 1.5816,
+      "step": 192
+    },
+    {
+      "epoch": 0.18996062992125984,
+      "grad_norm": 19.473068237304688,
+      "learning_rate": 1.888667992047714e-07,
+      "loss": 1.4029,
+      "step": 193
+    },
+    {
+      "epoch": 0.19094488188976377,
+      "grad_norm": 22.895261764526367,
+      "learning_rate": 1.898608349900597e-07,
+      "loss": 1.9385,
+      "step": 194
+    },
+    {
+      "epoch": 0.1919291338582677,
+      "grad_norm": 22.117504119873047,
+      "learning_rate": 1.9085487077534794e-07,
+      "loss": 1.5596,
+      "step": 195
+    },
+    {
+      "epoch": 0.19291338582677164,
+      "grad_norm": 25.059682846069336,
+      "learning_rate": 1.9184890656063622e-07,
+      "loss": 1.6663,
+      "step": 196
+    },
+    {
+      "epoch": 0.19389763779527558,
+      "grad_norm": 31.338993072509766,
+      "learning_rate": 1.9284294234592447e-07,
+      "loss": 2.0026,
+      "step": 197
+    },
+    {
+      "epoch": 0.19488188976377951,
+      "grad_norm": 25.27398681640625,
+      "learning_rate": 1.9383697813121275e-07,
+      "loss": 2.0046,
+      "step": 198
+    },
+    {
+      "epoch": 0.19586614173228348,
+      "grad_norm": 22.62946128845215,
+      "learning_rate": 1.94831013916501e-07,
+      "loss": 1.5016,
+      "step": 199
+    },
+    {
+      "epoch": 0.1968503937007874,
+      "grad_norm": 26.856422424316406,
+      "learning_rate": 1.958250497017893e-07,
+      "loss": 2.184,
+      "step": 200
+    },
+    {
+      "epoch": 0.19783464566929135,
+      "grad_norm": 27.561872482299805,
+      "learning_rate": 1.9681908548707756e-07,
+      "loss": 2.3442,
+      "step": 201
+    },
+    {
+      "epoch": 0.19881889763779528,
+      "grad_norm": 29.985654830932617,
+      "learning_rate": 1.9781312127236584e-07,
+      "loss": 2.6981,
+      "step": 202
+    },
+    {
+      "epoch": 0.19980314960629922,
+      "grad_norm": 22.171070098876953,
+      "learning_rate": 1.9880715705765412e-07,
+      "loss": 2.5481,
+      "step": 203
+    },
+    {
+      "epoch": 0.20078740157480315,
+      "grad_norm": 26.214073181152344,
+      "learning_rate": 1.9980119284294237e-07,
+      "loss": 2.9798,
+      "step": 204
+    },
+    {
+      "epoch": 0.2017716535433071,
+      "grad_norm": 22.690643310546875,
+      "learning_rate": 2.0079522862823065e-07,
+      "loss": 2.287,
+      "step": 205
+    },
+    {
+      "epoch": 0.20275590551181102,
+      "grad_norm": 21.528181076049805,
+      "learning_rate": 2.017892644135189e-07,
+      "loss": 1.9393,
+      "step": 206
+    },
+    {
+      "epoch": 0.20374015748031496,
+      "grad_norm": 24.61495018005371,
+      "learning_rate": 2.027833001988072e-07,
+      "loss": 2.892,
+      "step": 207
+    },
+    {
+      "epoch": 0.2047244094488189,
+      "grad_norm": 21.661781311035156,
+      "learning_rate": 2.0377733598409546e-07,
+      "loss": 2.26,
+      "step": 208
+    },
+    {
+      "epoch": 0.20570866141732283,
+      "grad_norm": 21.17841339111328,
+      "learning_rate": 2.0477137176938374e-07,
+      "loss": 2.5911,
+      "step": 209
+    },
+    {
+      "epoch": 0.20669291338582677,
+      "grad_norm": 22.297616958618164,
+      "learning_rate": 2.05765407554672e-07,
+      "loss": 2.1239,
+      "step": 210
+    },
+    {
+      "epoch": 0.2076771653543307,
+      "grad_norm": 18.447376251220703,
+      "learning_rate": 2.0675944333996027e-07,
+      "loss": 2.0683,
+      "step": 211
+    },
+    {
+      "epoch": 0.20866141732283464,
+      "grad_norm": 19.692792892456055,
+      "learning_rate": 2.0775347912524852e-07,
+      "loss": 1.768,
+      "step": 212
+    },
+    {
+      "epoch": 0.20964566929133857,
+      "grad_norm": 24.793012619018555,
+      "learning_rate": 2.087475149105368e-07,
+      "loss": 2.5468,
+      "step": 213
+    },
+    {
+      "epoch": 0.2106299212598425,
+      "grad_norm": 21.60909652709961,
+      "learning_rate": 2.097415506958251e-07,
+      "loss": 1.8956,
+      "step": 214
+    },
+    {
+      "epoch": 0.21161417322834647,
+      "grad_norm": 22.114286422729492,
+      "learning_rate": 2.1073558648111336e-07,
+      "loss": 2.044,
+      "step": 215
+    },
+    {
+      "epoch": 0.2125984251968504,
+      "grad_norm": 19.02837562561035,
+      "learning_rate": 2.1172962226640164e-07,
+      "loss": 1.5721,
+      "step": 216
+    },
+    {
+      "epoch": 0.21358267716535434,
+      "grad_norm": 19.785751342773438,
+      "learning_rate": 2.127236580516899e-07,
+      "loss": 1.6278,
+      "step": 217
+    },
+    {
+      "epoch": 0.21456692913385828,
+      "grad_norm": 21.3282470703125,
+      "learning_rate": 2.1371769383697817e-07,
+      "loss": 1.7754,
+      "step": 218
+    },
+    {
+      "epoch": 0.2155511811023622,
+      "grad_norm": 25.80916404724121,
+      "learning_rate": 2.1471172962226642e-07,
+      "loss": 1.8594,
+      "step": 219
+    },
+    {
+      "epoch": 0.21653543307086615,
+      "grad_norm": 21.912315368652344,
+      "learning_rate": 2.1570576540755473e-07,
+      "loss": 1.8309,
+      "step": 220
+    },
+    {
+      "epoch": 0.21751968503937008,
+      "grad_norm": 24.051366806030273,
+      "learning_rate": 2.1669980119284298e-07,
+      "loss": 2.0619,
+      "step": 221
+    },
+    {
+      "epoch": 0.21850393700787402,
+      "grad_norm": 24.29237174987793,
+      "learning_rate": 2.1769383697813126e-07,
+      "loss": 2.3335,
+      "step": 222
+    },
+    {
+      "epoch": 0.21948818897637795,
+      "grad_norm": 25.850160598754883,
+      "learning_rate": 2.186878727634195e-07,
+      "loss": 2.023,
+      "step": 223
+    },
+    {
+      "epoch": 0.2204724409448819,
+      "grad_norm": 27.208112716674805,
+      "learning_rate": 2.196819085487078e-07,
+      "loss": 2.1975,
+      "step": 224
+    },
+    {
+      "epoch": 0.22145669291338582,
+      "grad_norm": 22.276878356933594,
+      "learning_rate": 2.2067594433399604e-07,
+      "loss": 1.9228,
+      "step": 225
+    },
+    {
+      "epoch": 0.22244094488188976,
+      "grad_norm": 30.213895797729492,
+      "learning_rate": 2.2166998011928432e-07,
+      "loss": 2.3565,
+      "step": 226
+    },
+    {
+      "epoch": 0.2234251968503937,
+      "grad_norm": 22.16749382019043,
+      "learning_rate": 2.2266401590457263e-07,
+      "loss": 1.896,
+      "step": 227
+    },
+    {
+      "epoch": 0.22440944881889763,
+      "grad_norm": 21.131744384765625,
+      "learning_rate": 2.2365805168986088e-07,
+      "loss": 2.0912,
+      "step": 228
+    },
+    {
+      "epoch": 0.22539370078740156,
+      "grad_norm": 25.1036376953125,
+      "learning_rate": 2.2465208747514916e-07,
+      "loss": 2.7703,
+      "step": 229
+    },
+    {
+      "epoch": 0.2263779527559055,
+      "grad_norm": 25.316814422607422,
+      "learning_rate": 2.256461232604374e-07,
+      "loss": 1.6988,
+      "step": 230
+    },
+    {
+      "epoch": 0.22736220472440946,
+      "grad_norm": 25.363996505737305,
+      "learning_rate": 2.266401590457257e-07,
+      "loss": 2.0406,
+      "step": 231
+    },
+    {
+      "epoch": 0.2283464566929134,
+      "grad_norm": 21.906835556030273,
+      "learning_rate": 2.2763419483101394e-07,
+      "loss": 1.9288,
+      "step": 232
+    },
+    {
+      "epoch": 0.22933070866141733,
+      "grad_norm": 21.407150268554688,
+      "learning_rate": 2.2862823061630222e-07,
+      "loss": 2.0457,
+      "step": 233
+    },
+    {
+      "epoch": 0.23031496062992127,
+      "grad_norm": 21.2374210357666,
+      "learning_rate": 2.296222664015905e-07,
+      "loss": 1.7061,
+      "step": 234
+    },
+    {
+      "epoch": 0.2312992125984252,
+      "grad_norm": 20.94179344177246,
+      "learning_rate": 2.3061630218687878e-07,
+      "loss": 1.6244,
+      "step": 235
+    },
+    {
+      "epoch": 0.23228346456692914,
+      "grad_norm": 21.845712661743164,
+      "learning_rate": 2.3161033797216703e-07,
+      "loss": 2.0241,
+      "step": 236
+    },
+    {
+      "epoch": 0.23326771653543307,
+      "grad_norm": 19.496191024780273,
+      "learning_rate": 2.326043737574553e-07,
+      "loss": 1.567,
+      "step": 237
+    },
+    {
+      "epoch": 0.234251968503937,
+      "grad_norm": 21.819353103637695,
+      "learning_rate": 2.335984095427436e-07,
+      "loss": 1.8084,
+      "step": 238
+    },
+    {
+      "epoch": 0.23523622047244094,
+      "grad_norm": 27.17051124572754,
+      "learning_rate": 2.3459244532803184e-07,
+      "loss": 2.4363,
+      "step": 239
+    },
+    {
+      "epoch": 0.23622047244094488,
+      "grad_norm": 24.850723266601562,
+      "learning_rate": 2.3558648111332012e-07,
+      "loss": 1.7532,
+      "step": 240
+    },
+    {
+      "epoch": 0.2372047244094488,
+      "grad_norm": 24.120052337646484,
+      "learning_rate": 2.365805168986084e-07,
+      "loss": 2.0797,
+      "step": 241
+    },
+    {
+      "epoch": 0.23818897637795275,
+      "grad_norm": 23.708179473876953,
+      "learning_rate": 2.3757455268389668e-07,
+      "loss": 1.9562,
+      "step": 242
+    },
+    {
+      "epoch": 0.23917322834645668,
+      "grad_norm": 20.58576774597168,
+      "learning_rate": 2.385685884691849e-07,
+      "loss": 1.6751,
+      "step": 243
+    },
+    {
+      "epoch": 0.24015748031496062,
+      "grad_norm": 25.161970138549805,
+      "learning_rate": 2.395626242544732e-07,
+      "loss": 2.0265,
+      "step": 244
+    },
+    {
+      "epoch": 0.24114173228346455,
+      "grad_norm": 22.079387664794922,
+      "learning_rate": 2.4055666003976146e-07,
+      "loss": 1.6065,
+      "step": 245
+    },
+    {
+      "epoch": 0.2421259842519685,
+      "grad_norm": 20.125717163085938,
+      "learning_rate": 2.4155069582504976e-07,
+      "loss": 1.7439,
+      "step": 246
+    },
+    {
+      "epoch": 0.24311023622047245,
+      "grad_norm": 22.855445861816406,
+      "learning_rate": 2.42544731610338e-07,
+      "loss": 2.0237,
+      "step": 247
+    },
+    {
+      "epoch": 0.2440944881889764,
+      "grad_norm": 22.16312026977539,
+      "learning_rate": 2.4353876739562627e-07,
+      "loss": 1.6128,
+      "step": 248
+    },
+    {
+      "epoch": 0.24507874015748032,
+      "grad_norm": 21.942440032958984,
+      "learning_rate": 2.445328031809146e-07,
+      "loss": 1.6581,
+      "step": 249
+    },
+    {
+      "epoch": 0.24606299212598426,
+      "grad_norm": 24.027729034423828,
+      "learning_rate": 2.4552683896620283e-07,
+      "loss": 2.1538,
+      "step": 250
+    },
+    {
+      "epoch": 0.2470472440944882,
+      "grad_norm": 23.1455078125,
+      "learning_rate": 2.4652087475149113e-07,
+      "loss": 2.049,
+      "step": 251
+    },
+    {
+      "epoch": 0.24803149606299213,
+      "grad_norm": 22.017520904541016,
+      "learning_rate": 2.475149105367794e-07,
+      "loss": 1.2573,
+      "step": 252
+    },
+    {
+      "epoch": 0.24901574803149606,
+      "grad_norm": 18.747554779052734,
+      "learning_rate": 2.4850894632206764e-07,
+      "loss": 1.5619,
+      "step": 253
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.277063369750977,
+      "learning_rate": 2.495029821073559e-07,
+      "loss": 1.2611,
+      "step": 254
+    },
+    {
+      "epoch": 0.25098425196850394,
+      "grad_norm": 19.781038284301758,
+      "learning_rate": 2.5049701789264414e-07,
+      "loss": 1.3443,
+      "step": 255
+    },
+    {
+      "epoch": 0.25196850393700787,
+      "grad_norm": 21.605199813842773,
+      "learning_rate": 2.5149105367793245e-07,
+      "loss": 1.3436,
+      "step": 256
+    },
+    {
+      "epoch": 0.2529527559055118,
+      "grad_norm": 29.64748764038086,
+      "learning_rate": 2.524850894632207e-07,
+      "loss": 2.8117,
+      "step": 257
+    },
+    {
+      "epoch": 0.25393700787401574,
+      "grad_norm": 19.245962142944336,
+      "learning_rate": 2.53479125248509e-07,
+      "loss": 1.7563,
+      "step": 258
+    },
+    {
+      "epoch": 0.2549212598425197,
+      "grad_norm": 18.752405166625977,
+      "learning_rate": 2.5447316103379726e-07,
+      "loss": 1.3148,
+      "step": 259
+    },
+    {
+      "epoch": 0.2559055118110236,
+      "grad_norm": 24.133914947509766,
+      "learning_rate": 2.554671968190855e-07,
+      "loss": 2.0278,
+      "step": 260
+    },
+    {
+      "epoch": 0.25688976377952755,
+      "grad_norm": 19.490673065185547,
+      "learning_rate": 2.564612326043738e-07,
+      "loss": 1.2403,
+      "step": 261
+    },
+    {
+      "epoch": 0.2578740157480315,
+      "grad_norm": 20.14169692993164,
+      "learning_rate": 2.5745526838966207e-07,
+      "loss": 1.588,
+      "step": 262
+    },
+    {
+      "epoch": 0.2588582677165354,
+      "grad_norm": 26.948959350585938,
+      "learning_rate": 2.5844930417495037e-07,
+      "loss": 2.0071,
+      "step": 263
+    },
+    {
+      "epoch": 0.25984251968503935,
+      "grad_norm": 22.330224990844727,
+      "learning_rate": 2.5944333996023857e-07,
+      "loss": 1.5312,
+      "step": 264
+    },
+    {
+      "epoch": 0.2608267716535433,
+      "grad_norm": 21.379629135131836,
+      "learning_rate": 2.604373757455269e-07,
+      "loss": 1.8641,
+      "step": 265
+    },
+    {
+      "epoch": 0.2618110236220472,
+      "grad_norm": 20.703140258789062,
+      "learning_rate": 2.614314115308152e-07,
+      "loss": 1.2933,
+      "step": 266
+    },
+    {
+      "epoch": 0.26279527559055116,
+      "grad_norm": 22.9117488861084,
+      "learning_rate": 2.6242544731610343e-07,
+      "loss": 1.6262,
+      "step": 267
+    },
+    {
+      "epoch": 0.2637795275590551,
+      "grad_norm": 23.842002868652344,
+      "learning_rate": 2.634194831013917e-07,
+      "loss": 1.721,
+      "step": 268
+    },
+    {
+      "epoch": 0.26476377952755903,
+      "grad_norm": 20.449384689331055,
+      "learning_rate": 2.6441351888667994e-07,
+      "loss": 1.4713,
+      "step": 269
+    },
+    {
+      "epoch": 0.265748031496063,
+      "grad_norm": 25.885969161987305,
+      "learning_rate": 2.6540755467196824e-07,
+      "loss": 1.4625,
+      "step": 270
+    },
+    {
+      "epoch": 0.26673228346456695,
+      "grad_norm": 24.52666473388672,
+      "learning_rate": 2.664015904572565e-07,
+      "loss": 1.7254,
+      "step": 271
+    },
+    {
+      "epoch": 0.2677165354330709,
+      "grad_norm": 23.4957275390625,
+      "learning_rate": 2.6739562624254475e-07,
+      "loss": 1.5108,
+      "step": 272
+    },
+    {
+      "epoch": 0.2687007874015748,
+      "grad_norm": 23.828855514526367,
+      "learning_rate": 2.6838966202783305e-07,
+      "loss": 2.1126,
+      "step": 273
+    },
+    {
+      "epoch": 0.26968503937007876,
+      "grad_norm": 21.05967903137207,
+      "learning_rate": 2.693836978131213e-07,
+      "loss": 1.3967,
+      "step": 274
+    },
+    {
+      "epoch": 0.2706692913385827,
+      "grad_norm": 24.555776596069336,
+      "learning_rate": 2.703777335984096e-07,
+      "loss": 1.7067,
+      "step": 275
+    },
+    {
+      "epoch": 0.27165354330708663,
+      "grad_norm": 22.135860443115234,
+      "learning_rate": 2.7137176938369786e-07,
+      "loss": 1.4847,
+      "step": 276
+    },
+    {
+      "epoch": 0.27263779527559057,
+      "grad_norm": 22.105913162231445,
+      "learning_rate": 2.723658051689861e-07,
+      "loss": 1.6515,
+      "step": 277
+    },
+    {
+      "epoch": 0.2736220472440945,
+      "grad_norm": 16.617225646972656,
+      "learning_rate": 2.7335984095427437e-07,
+      "loss": 0.9367,
+      "step": 278
+    },
+    {
+      "epoch": 0.27460629921259844,
+      "grad_norm": 26.01727867126465,
+      "learning_rate": 2.743538767395627e-07,
+      "loss": 2.0267,
+      "step": 279
+    },
+    {
+      "epoch": 0.2755905511811024,
+      "grad_norm": 22.522462844848633,
+      "learning_rate": 2.75347912524851e-07,
+      "loss": 1.5023,
+      "step": 280
+    },
+    {
+      "epoch": 0.2765748031496063,
+      "grad_norm": 20.646358489990234,
+      "learning_rate": 2.763419483101392e-07,
+      "loss": 1.1248,
+      "step": 281
+    },
+    {
+      "epoch": 0.27755905511811024,
+      "grad_norm": 23.3087158203125,
+      "learning_rate": 2.773359840954275e-07,
+      "loss": 1.6224,
+      "step": 282
+    },
+    {
+      "epoch": 0.2785433070866142,
+      "grad_norm": 24.115968704223633,
+      "learning_rate": 2.7833001988071574e-07,
+      "loss": 1.7969,
+      "step": 283
+    },
+    {
+      "epoch": 0.2795275590551181,
+      "grad_norm": 27.229188919067383,
+      "learning_rate": 2.7932405566600404e-07,
+      "loss": 2.2498,
+      "step": 284
+    },
+    {
+      "epoch": 0.28051181102362205,
+      "grad_norm": 25.797216415405273,
+      "learning_rate": 2.803180914512923e-07,
+      "loss": 1.7477,
+      "step": 285
+    },
+    {
+      "epoch": 0.281496062992126,
+      "grad_norm": 23.666858673095703,
+      "learning_rate": 2.8131212723658055e-07,
+      "loss": 1.6261,
+      "step": 286
+    },
+    {
+      "epoch": 0.2824803149606299,
+      "grad_norm": 26.826387405395508,
+      "learning_rate": 2.8230616302186885e-07,
+      "loss": 2.0911,
+      "step": 287
+    },
+    {
+      "epoch": 0.28346456692913385,
+      "grad_norm": 23.288511276245117,
+      "learning_rate": 2.833001988071571e-07,
+      "loss": 1.9519,
+      "step": 288
+    },
+    {
+      "epoch": 0.2844488188976378,
+      "grad_norm": 19.76810073852539,
+      "learning_rate": 2.842942345924454e-07,
+      "loss": 1.3132,
+      "step": 289
+    },
+    {
+      "epoch": 0.2854330708661417,
+      "grad_norm": 24.65778923034668,
+      "learning_rate": 2.852882703777336e-07,
+      "loss": 2.3292,
+      "step": 290
+    },
+    {
+      "epoch": 0.28641732283464566,
+      "grad_norm": 19.230430603027344,
+      "learning_rate": 2.862823061630219e-07,
+      "loss": 1.3781,
+      "step": 291
+    },
+    {
+      "epoch": 0.2874015748031496,
+      "grad_norm": 20.665143966674805,
+      "learning_rate": 2.8727634194831017e-07,
+      "loss": 1.5753,
+      "step": 292
+    },
+    {
+      "epoch": 0.28838582677165353,
+      "grad_norm": 18.584993362426758,
+      "learning_rate": 2.8827037773359847e-07,
+      "loss": 1.4158,
+      "step": 293
+    },
+    {
+      "epoch": 0.28937007874015747,
+      "grad_norm": 20.731081008911133,
+      "learning_rate": 2.892644135188867e-07,
+      "loss": 2.1661,
+      "step": 294
+    },
+    {
+      "epoch": 0.2903543307086614,
+      "grad_norm": 20.006755828857422,
+      "learning_rate": 2.90258449304175e-07,
+      "loss": 1.4928,
+      "step": 295
+    },
+    {
+      "epoch": 0.29133858267716534,
+      "grad_norm": 20.01453971862793,
+      "learning_rate": 2.912524850894633e-07,
+      "loss": 2.2825,
+      "step": 296
+    },
+    {
+      "epoch": 0.29232283464566927,
+      "grad_norm": 18.137697219848633,
+      "learning_rate": 2.9224652087475153e-07,
+      "loss": 1.7261,
+      "step": 297
+    },
+    {
+      "epoch": 0.2933070866141732,
+      "grad_norm": 20.046411514282227,
+      "learning_rate": 2.9324055666003984e-07,
+      "loss": 1.8635,
+      "step": 298
+    },
+    {
+      "epoch": 0.29429133858267714,
+      "grad_norm": 15.859444618225098,
+      "learning_rate": 2.9423459244532804e-07,
+      "loss": 0.974,
+      "step": 299
+    },
+    {
+      "epoch": 0.2952755905511811,
+      "grad_norm": 17.976015090942383,
+      "learning_rate": 2.9522862823061634e-07,
+      "loss": 1.53,
+      "step": 300
+    },
+    {
+      "epoch": 0.296259842519685,
+      "grad_norm": 20.52704429626465,
+      "learning_rate": 2.9622266401590465e-07,
+      "loss": 1.5985,
+      "step": 301
+    },
+    {
+      "epoch": 0.297244094488189,
+      "grad_norm": 19.45134925842285,
+      "learning_rate": 2.972166998011929e-07,
+      "loss": 1.2169,
+      "step": 302
+    },
+    {
+      "epoch": 0.29822834645669294,
+      "grad_norm": 22.01141357421875,
+      "learning_rate": 2.9821073558648115e-07,
+      "loss": 1.771,
+      "step": 303
+    },
+    {
+      "epoch": 0.2992125984251969,
+      "grad_norm": 17.591617584228516,
+      "learning_rate": 2.992047713717694e-07,
+      "loss": 1.4506,
+      "step": 304
+    },
+    {
+      "epoch": 0.3001968503937008,
+      "grad_norm": 25.382244110107422,
+      "learning_rate": 3.001988071570577e-07,
+      "loss": 1.9496,
+      "step": 305
+    },
+    {
+      "epoch": 0.30118110236220474,
+      "grad_norm": 21.026479721069336,
+      "learning_rate": 3.01192842942346e-07,
+      "loss": 1.2436,
+      "step": 306
+    },
+    {
+      "epoch": 0.30118110236220474,
+      "eval_Qnli-dev_cosine_accuracy": 0.671875,
+      "eval_Qnli-dev_cosine_accuracy_threshold": 0.822678804397583,
+      "eval_Qnli-dev_cosine_ap": 0.6991833375897509,
+      "eval_Qnli-dev_cosine_f1": 0.6688,
+      "eval_Qnli-dev_cosine_f1_threshold": 0.7263889312744141,
+      "eval_Qnli-dev_cosine_precision": 0.5372750642673522,
+      "eval_Qnli-dev_cosine_recall": 0.885593220338983,
+      "eval_Qnli-dev_dot_accuracy": 0.671875,
+      "eval_Qnli-dev_dot_accuracy_threshold": 632.18896484375,
+      "eval_Qnli-dev_dot_ap": 0.6993450442740953,
+      "eval_Qnli-dev_dot_f1": 0.6688,
+      "eval_Qnli-dev_dot_f1_threshold": 558.1864013671875,
+      "eval_Qnli-dev_dot_precision": 0.5372750642673522,
+      "eval_Qnli-dev_dot_recall": 0.885593220338983,
+      "eval_Qnli-dev_euclidean_accuracy": 0.671875,
+      "eval_Qnli-dev_euclidean_accuracy_threshold": 16.508346557617188,
+      "eval_Qnli-dev_euclidean_ap": 0.6991985293193381,
+      "eval_Qnli-dev_euclidean_f1": 0.6688,
+      "eval_Qnli-dev_euclidean_f1_threshold": 20.50628089904785,
+      "eval_Qnli-dev_euclidean_precision": 0.5372750642673522,
+      "eval_Qnli-dev_euclidean_recall": 0.885593220338983,
+      "eval_Qnli-dev_manhattan_accuracy": 0.66796875,
+      "eval_Qnli-dev_manhattan_accuracy_threshold": 362.6116638183594,
+      "eval_Qnli-dev_manhattan_ap": 0.6988100851084365,
+      "eval_Qnli-dev_manhattan_f1": 0.669826224328594,
+      "eval_Qnli-dev_manhattan_f1_threshold": 456.92926025390625,
+      "eval_Qnli-dev_manhattan_precision": 0.5340050377833753,
+      "eval_Qnli-dev_manhattan_recall": 0.8983050847457628,
+      "eval_Qnli-dev_max_accuracy": 0.671875,
+      "eval_Qnli-dev_max_accuracy_threshold": 632.18896484375,
+      "eval_Qnli-dev_max_ap": 0.6993450442740953,
+      "eval_Qnli-dev_max_f1": 0.669826224328594,
+      "eval_Qnli-dev_max_f1_threshold": 558.1864013671875,
+      "eval_Qnli-dev_max_precision": 0.5372750642673522,
+      "eval_Qnli-dev_max_recall": 0.8983050847457628,
+      "eval_allNLI-dev_cosine_accuracy": 0.68359375,
+      "eval_allNLI-dev_cosine_accuracy_threshold": 0.9261605739593506,
+      "eval_allNLI-dev_cosine_ap": 0.47961880692471753,
+      "eval_allNLI-dev_cosine_f1": 0.5101351351351351,
+      "eval_allNLI-dev_cosine_f1_threshold": 0.7954977750778198,
+      "eval_allNLI-dev_cosine_precision": 0.360381861575179,
+      "eval_allNLI-dev_cosine_recall": 0.8728323699421965,
+      "eval_allNLI-dev_dot_accuracy": 0.68359375,
+      "eval_allNLI-dev_dot_accuracy_threshold": 711.72412109375,
+      "eval_allNLI-dev_dot_ap": 0.4796043382357695,
+      "eval_allNLI-dev_dot_f1": 0.5101351351351351,
+      "eval_allNLI-dev_dot_f1_threshold": 611.3052978515625,
+      "eval_allNLI-dev_dot_precision": 0.360381861575179,
+      "eval_allNLI-dev_dot_recall": 0.8728323699421965,
+      "eval_allNLI-dev_euclidean_accuracy": 0.68359375,
+      "eval_allNLI-dev_euclidean_accuracy_threshold": 10.652993202209473,
+      "eval_allNLI-dev_euclidean_ap": 0.47952222164725533,
+      "eval_allNLI-dev_euclidean_f1": 0.5101351351351351,
+      "eval_allNLI-dev_euclidean_f1_threshold": 17.728565216064453,
+      "eval_allNLI-dev_euclidean_precision": 0.360381861575179,
+      "eval_allNLI-dev_euclidean_recall": 0.8728323699421965,
+      "eval_allNLI-dev_manhattan_accuracy": 0.6796875,
+      "eval_allNLI-dev_manhattan_accuracy_threshold": 236.265380859375,
+      "eval_allNLI-dev_manhattan_ap": 0.4807847983836923,
+      "eval_allNLI-dev_manhattan_f1": 0.5103668261562998,
+      "eval_allNLI-dev_manhattan_f1_threshold": 417.0115966796875,
+      "eval_allNLI-dev_manhattan_precision": 0.3524229074889868,
+      "eval_allNLI-dev_manhattan_recall": 0.9248554913294798,
+      "eval_allNLI-dev_max_accuracy": 0.68359375,
+      "eval_allNLI-dev_max_accuracy_threshold": 711.72412109375,
+      "eval_allNLI-dev_max_ap": 0.4807847983836923,
+      "eval_allNLI-dev_max_f1": 0.5103668261562998,
+      "eval_allNLI-dev_max_f1_threshold": 611.3052978515625,
+      "eval_allNLI-dev_max_precision": 0.360381861575179,
+      "eval_allNLI-dev_max_recall": 0.9248554913294798,
+      "eval_loss": 1.5213453769683838,
+      "eval_runtime": 50.241,
+      "eval_samples_per_second": 33.12,
+      "eval_sequential_score": 0.6993450442740953,
+      "eval_steps_per_second": 0.139,
+      "eval_sts-test_pearson_cosine": 0.4516880351405699,
+      "eval_sts-test_pearson_dot": 0.4516761779594392,
+      "eval_sts-test_pearson_euclidean": 0.4615962658267084,
+      "eval_sts-test_pearson_manhattan": 0.46201541629939036,
+      "eval_sts-test_pearson_max": 0.46201541629939036,
+      "eval_sts-test_spearman_cosine": 0.46725208205491164,
+      "eval_sts-test_spearman_dot": 0.4672651046719766,
+      "eval_sts-test_spearman_euclidean": 0.46725722844997175,
+      "eval_sts-test_spearman_manhattan": 0.4677601878767496,
+      "eval_sts-test_spearman_max": 0.4677601878767496,
+      "step": 306
+    },
+    {
+      "epoch": 0.3021653543307087,
+      "grad_norm": 27.924283981323242,
+      "learning_rate": 3.021868787276342e-07,
+      "loss": 2.2057,
+      "step": 307
+    },
+    {
+      "epoch": 0.3031496062992126,
+      "grad_norm": 19.933040618896484,
+      "learning_rate": 3.031809145129225e-07,
+      "loss": 1.6786,
+      "step": 308
+    },
+    {
+      "epoch": 0.30413385826771655,
+      "grad_norm": 20.79754638671875,
+      "learning_rate": 3.0417495029821077e-07,
+      "loss": 1.748,
+      "step": 309
+    },
+    {
+      "epoch": 0.3051181102362205,
+      "grad_norm": 21.448503494262695,
+      "learning_rate": 3.051689860834991e-07,
+      "loss": 1.5541,
+      "step": 310
+    },
+    {
+      "epoch": 0.3061023622047244,
+      "grad_norm": 23.68476676940918,
+      "learning_rate": 3.0616302186878733e-07,
+      "loss": 2.2968,
+      "step": 311
+    },
+    {
+      "epoch": 0.30708661417322836,
+      "grad_norm": 19.669784545898438,
+      "learning_rate": 3.071570576540756e-07,
+      "loss": 1.585,
+      "step": 312
+    },
+    {
+      "epoch": 0.3080708661417323,
+      "grad_norm": 21.109407424926758,
+      "learning_rate": 3.0815109343936384e-07,
+      "loss": 1.8371,
+      "step": 313
+    },
+    {
+      "epoch": 0.3090551181102362,
+      "grad_norm": 17.614957809448242,
+      "learning_rate": 3.0914512922465214e-07,
+      "loss": 1.1129,
+      "step": 314
+    },
+    {
+      "epoch": 0.31003937007874016,
+      "grad_norm": 18.3952693939209,
+      "learning_rate": 3.1013916500994045e-07,
+      "loss": 1.5495,
+      "step": 315
+    },
+    {
+      "epoch": 0.3110236220472441,
+      "grad_norm": 17.686391830444336,
+      "learning_rate": 3.1113320079522865e-07,
+      "loss": 1.4327,
+      "step": 316
+    },
+    {
+      "epoch": 0.31200787401574803,
+      "grad_norm": 21.255008697509766,
+      "learning_rate": 3.1212723658051695e-07,
+      "loss": 1.4801,
+      "step": 317
+    },
+    {
+      "epoch": 0.31299212598425197,
+      "grad_norm": 19.738698959350586,
+      "learning_rate": 3.131212723658052e-07,
+      "loss": 1.7096,
+      "step": 318
+    },
+    {
+      "epoch": 0.3139763779527559,
+      "grad_norm": 21.54611587524414,
+      "learning_rate": 3.141153081510935e-07,
+      "loss": 1.6717,
+      "step": 319
+    },
+    {
+      "epoch": 0.31496062992125984,
+      "grad_norm": 21.207399368286133,
+      "learning_rate": 3.151093439363818e-07,
+      "loss": 1.7151,
+      "step": 320
+    },
+    {
+      "epoch": 0.3159448818897638,
+      "grad_norm": 23.244253158569336,
+      "learning_rate": 3.1610337972167e-07,
+      "loss": 1.7081,
+      "step": 321
+    },
+    {
+      "epoch": 0.3169291338582677,
+      "grad_norm": 19.656099319458008,
+      "learning_rate": 3.170974155069583e-07,
+      "loss": 1.431,
+      "step": 322
+    },
+    {
+      "epoch": 0.31791338582677164,
+      "grad_norm": 22.234607696533203,
+      "learning_rate": 3.1809145129224657e-07,
+      "loss": 1.5734,
+      "step": 323
+    },
+    {
+      "epoch": 0.3188976377952756,
+      "grad_norm": 23.064062118530273,
+      "learning_rate": 3.190854870775349e-07,
+      "loss": 1.7307,
+      "step": 324
+    },
+    {
+      "epoch": 0.3198818897637795,
+      "grad_norm": 17.646806716918945,
+      "learning_rate": 3.200795228628231e-07,
+      "loss": 1.0644,
+      "step": 325
+    },
+    {
+      "epoch": 0.32086614173228345,
+      "grad_norm": 18.901668548583984,
+      "learning_rate": 3.210735586481114e-07,
+      "loss": 1.0651,
+      "step": 326
+    },
+    {
+      "epoch": 0.3218503937007874,
+      "grad_norm": 22.49507713317871,
+      "learning_rate": 3.220675944333997e-07,
+      "loss": 1.4805,
+      "step": 327
+    },
+    {
+      "epoch": 0.3228346456692913,
+      "grad_norm": 17.89360809326172,
+      "learning_rate": 3.2306163021868794e-07,
+      "loss": 0.839,
+      "step": 328
+    },
+    {
+      "epoch": 0.32381889763779526,
+      "grad_norm": 20.79975128173828,
+      "learning_rate": 3.240556660039762e-07,
+      "loss": 1.1801,
+      "step": 329
+    },
+    {
+      "epoch": 0.3248031496062992,
+      "grad_norm": 18.659631729125977,
+      "learning_rate": 3.2504970178926444e-07,
+      "loss": 1.36,
+      "step": 330
+    },
+    {
+      "epoch": 0.3257874015748031,
+      "grad_norm": 21.92584800720215,
+      "learning_rate": 3.2604373757455275e-07,
+      "loss": 1.3371,
+      "step": 331
+    },
+    {
+      "epoch": 0.32677165354330706,
+      "grad_norm": 17.514934539794922,
+      "learning_rate": 3.27037773359841e-07,
+      "loss": 1.1707,
+      "step": 332
+    },
+    {
+      "epoch": 0.327755905511811,
+      "grad_norm": 18.580942153930664,
+      "learning_rate": 3.280318091451293e-07,
+      "loss": 1.2572,
+      "step": 333
+    },
+    {
+      "epoch": 0.328740157480315,
+      "grad_norm": 22.246044158935547,
+      "learning_rate": 3.2902584493041756e-07,
+      "loss": 1.3537,
+      "step": 334
+    },
+    {
+      "epoch": 0.3297244094488189,
+      "grad_norm": 26.39656639099121,
+      "learning_rate": 3.300198807157058e-07,
+      "loss": 1.7096,
+      "step": 335
+    },
+    {
+      "epoch": 0.33070866141732286,
+      "grad_norm": 24.07045555114746,
+      "learning_rate": 3.310139165009941e-07,
+      "loss": 1.5137,
+      "step": 336
+    },
+    {
+      "epoch": 0.3316929133858268,
+      "grad_norm": 18.82029914855957,
+      "learning_rate": 3.3200795228628237e-07,
+      "loss": 1.1989,
+      "step": 337
+    },
+    {
+      "epoch": 0.33267716535433073,
+      "grad_norm": 19.72603416442871,
+      "learning_rate": 3.330019880715706e-07,
+      "loss": 1.3559,
+      "step": 338
+    },
+    {
+      "epoch": 0.33366141732283466,
+      "grad_norm": 19.72848129272461,
+      "learning_rate": 3.3399602385685887e-07,
+      "loss": 1.3643,
+      "step": 339
+    },
+    {
+      "epoch": 0.3346456692913386,
+      "grad_norm": 19.329681396484375,
+      "learning_rate": 3.349900596421472e-07,
+      "loss": 1.2283,
+      "step": 340
+    },
+    {
+      "epoch": 0.33562992125984253,
+      "grad_norm": 25.018512725830078,
+      "learning_rate": 3.359840954274355e-07,
+      "loss": 1.5829,
+      "step": 341
+    },
+    {
+      "epoch": 0.33661417322834647,
+      "grad_norm": 18.037017822265625,
+      "learning_rate": 3.369781312127237e-07,
+      "loss": 1.1866,
+      "step": 342
+    },
+    {
+      "epoch": 0.3375984251968504,
+      "grad_norm": 21.839269638061523,
+      "learning_rate": 3.37972166998012e-07,
+      "loss": 1.531,
+      "step": 343
+    },
+    {
+      "epoch": 0.33858267716535434,
+      "grad_norm": 21.1947021484375,
+      "learning_rate": 3.3896620278330024e-07,
+      "loss": 1.5581,
+      "step": 344
+    },
+    {
+      "epoch": 0.3395669291338583,
+      "grad_norm": 22.119661331176758,
+      "learning_rate": 3.3996023856858855e-07,
+      "loss": 1.5587,
+      "step": 345
+    },
+    {
+      "epoch": 0.3405511811023622,
+      "grad_norm": 19.932144165039062,
+      "learning_rate": 3.409542743538768e-07,
+      "loss": 1.1403,
+      "step": 346
+    },
+    {
+      "epoch": 0.34153543307086615,
+      "grad_norm": 26.49030303955078,
+      "learning_rate": 3.4194831013916505e-07,
+      "loss": 1.9728,
+      "step": 347
+    },
+    {
+      "epoch": 0.3425196850393701,
+      "grad_norm": 18.05373191833496,
+      "learning_rate": 3.4294234592445336e-07,
+      "loss": 1.0818,
+      "step": 348
+    },
+    {
+      "epoch": 0.343503937007874,
+      "grad_norm": 19.15143585205078,
+      "learning_rate": 3.439363817097416e-07,
+      "loss": 1.2993,
+      "step": 349
+    },
+    {
+      "epoch": 0.34448818897637795,
+      "grad_norm": 21.032453536987305,
+      "learning_rate": 3.449304174950299e-07,
+      "loss": 1.7779,
+      "step": 350
+    },
+    {
+      "epoch": 0.3454724409448819,
+      "grad_norm": 20.775541305541992,
+      "learning_rate": 3.459244532803181e-07,
+      "loss": 1.319,
+      "step": 351
+    },
+    {
+      "epoch": 0.3464566929133858,
+      "grad_norm": 24.86508560180664,
+      "learning_rate": 3.469184890656064e-07,
+      "loss": 1.9236,
+      "step": 352
+    },
+    {
+      "epoch": 0.34744094488188976,
+      "grad_norm": 19.496232986450195,
+      "learning_rate": 3.4791252485089467e-07,
+      "loss": 1.3085,
+      "step": 353
+    },
+    {
+      "epoch": 0.3484251968503937,
+      "grad_norm": 24.63651466369629,
+      "learning_rate": 3.48906560636183e-07,
+      "loss": 2.2049,
+      "step": 354
+    },
+    {
+      "epoch": 0.3494094488188976,
+      "grad_norm": 16.268550872802734,
+      "learning_rate": 3.499005964214713e-07,
+      "loss": 1.3697,
+      "step": 355
+    },
+    {
+      "epoch": 0.35039370078740156,
+      "grad_norm": 17.859365463256836,
+      "learning_rate": 3.508946322067595e-07,
+      "loss": 1.5367,
+      "step": 356
+    },
+    {
+      "epoch": 0.3513779527559055,
+      "grad_norm": 22.554317474365234,
+      "learning_rate": 3.518886679920478e-07,
+      "loss": 1.2516,
+      "step": 357
+    },
+    {
+      "epoch": 0.35236220472440943,
+      "grad_norm": 20.86847496032715,
+      "learning_rate": 3.5288270377733604e-07,
+      "loss": 1.6497,
+      "step": 358
+    },
+    {
+      "epoch": 0.35334645669291337,
+      "grad_norm": 17.967374801635742,
+      "learning_rate": 3.5387673956262434e-07,
+      "loss": 1.2457,
+      "step": 359
+    },
+    {
+      "epoch": 0.3543307086614173,
+      "grad_norm": 20.069971084594727,
+      "learning_rate": 3.5487077534791254e-07,
+      "loss": 1.2733,
+      "step": 360
+    },
+    {
+      "epoch": 0.35531496062992124,
+      "grad_norm": 20.93694496154785,
+      "learning_rate": 3.5586481113320085e-07,
+      "loss": 1.4768,
+      "step": 361
+    },
+    {
+      "epoch": 0.3562992125984252,
+      "grad_norm": 19.617618560791016,
+      "learning_rate": 3.5685884691848915e-07,
+      "loss": 1.1363,
+      "step": 362
+    },
+    {
+      "epoch": 0.3572834645669291,
+      "grad_norm": 23.839365005493164,
+      "learning_rate": 3.578528827037774e-07,
+      "loss": 1.5731,
+      "step": 363
+    },
+    {
+      "epoch": 0.35826771653543305,
+      "grad_norm": 17.460094451904297,
+      "learning_rate": 3.5884691848906566e-07,
+      "loss": 1.0821,
+      "step": 364
+    },
+    {
+      "epoch": 0.359251968503937,
+      "grad_norm": 19.736909866333008,
+      "learning_rate": 3.598409542743539e-07,
+      "loss": 1.1563,
+      "step": 365
+    },
+    {
+      "epoch": 0.36023622047244097,
+      "grad_norm": 23.919334411621094,
+      "learning_rate": 3.608349900596422e-07,
+      "loss": 1.8843,
+      "step": 366
+    },
+    {
+      "epoch": 0.3612204724409449,
+      "grad_norm": 20.10567283630371,
+      "learning_rate": 3.6182902584493047e-07,
+      "loss": 1.2239,
+      "step": 367
+    },
+    {
+      "epoch": 0.36220472440944884,
+      "grad_norm": 20.869524002075195,
+      "learning_rate": 3.6282306163021877e-07,
+      "loss": 1.4411,
+      "step": 368
+    },
+    {
+      "epoch": 0.3631889763779528,
+      "grad_norm": 23.02889633178711,
+      "learning_rate": 3.63817097415507e-07,
+      "loss": 2.1003,
+      "step": 369
+    },
+    {
+      "epoch": 0.3641732283464567,
+      "grad_norm": 19.797988891601562,
+      "learning_rate": 3.648111332007953e-07,
+      "loss": 1.6558,
+      "step": 370
+    },
+    {
+      "epoch": 0.36515748031496065,
+      "grad_norm": 20.25324058532715,
+      "learning_rate": 3.658051689860836e-07,
+      "loss": 1.6502,
+      "step": 371
+    },
+    {
+      "epoch": 0.3661417322834646,
+      "grad_norm": 19.865921020507812,
+      "learning_rate": 3.6679920477137184e-07,
+      "loss": 1.7204,
+      "step": 372
+    },
+    {
+      "epoch": 0.3671259842519685,
+      "grad_norm": 20.396156311035156,
+      "learning_rate": 3.677932405566601e-07,
+      "loss": 1.7422,
+      "step": 373
+    },
+    {
+      "epoch": 0.36811023622047245,
+      "grad_norm": 20.016502380371094,
+      "learning_rate": 3.6878727634194834e-07,
+      "loss": 1.3859,
+      "step": 374
+    },
+    {
+      "epoch": 0.3690944881889764,
+      "grad_norm": 14.640628814697266,
+      "learning_rate": 3.6978131212723665e-07,
+      "loss": 0.8876,
+      "step": 375
+    },
+    {
+      "epoch": 0.3700787401574803,
+      "grad_norm": 20.027318954467773,
+      "learning_rate": 3.7077534791252495e-07,
+      "loss": 1.2399,
+      "step": 376
+    },
+    {
+      "epoch": 0.37106299212598426,
+      "grad_norm": 17.005905151367188,
+      "learning_rate": 3.7176938369781315e-07,
+      "loss": 1.1039,
+      "step": 377
+    },
+    {
+      "epoch": 0.3720472440944882,
+      "grad_norm": 21.551816940307617,
+      "learning_rate": 3.7276341948310145e-07,
+      "loss": 1.733,
+      "step": 378
+    },
+    {
+      "epoch": 0.37303149606299213,
+      "grad_norm": 23.17444610595703,
+      "learning_rate": 3.737574552683897e-07,
+      "loss": 1.6897,
+      "step": 379
+    },
+    {
+      "epoch": 0.37401574803149606,
+      "grad_norm": 22.414926528930664,
+      "learning_rate": 3.74751491053678e-07,
+      "loss": 2.0532,
+      "step": 380
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 17.610034942626953,
+      "learning_rate": 3.7574552683896626e-07,
+      "loss": 1.0156,
+      "step": 381
+    },
+    {
+      "epoch": 0.37598425196850394,
+      "grad_norm": 15.841965675354004,
+      "learning_rate": 3.767395626242545e-07,
+      "loss": 0.8888,
+      "step": 382
+    },
+    {
+      "epoch": 0.37696850393700787,
+      "grad_norm": 20.21027946472168,
+      "learning_rate": 3.777335984095428e-07,
+      "loss": 1.322,
+      "step": 383
+    },
+    {
+      "epoch": 0.3779527559055118,
+      "grad_norm": 23.444580078125,
+      "learning_rate": 3.787276341948311e-07,
+      "loss": 1.6828,
+      "step": 384
+    },
+    {
+      "epoch": 0.37893700787401574,
+      "grad_norm": 18.69243812561035,
+      "learning_rate": 3.797216699801194e-07,
+      "loss": 1.1567,
+      "step": 385
+    },
+    {
+      "epoch": 0.3799212598425197,
+      "grad_norm": 20.93277931213379,
+      "learning_rate": 3.807157057654076e-07,
+      "loss": 1.6117,
+      "step": 386
+    },
+    {
+      "epoch": 0.3809055118110236,
+      "grad_norm": 17.97561264038086,
+      "learning_rate": 3.817097415506959e-07,
+      "loss": 1.1776,
+      "step": 387
+    },
+    {
+      "epoch": 0.38188976377952755,
+      "grad_norm": 23.98857307434082,
+      "learning_rate": 3.8270377733598414e-07,
+      "loss": 1.641,
+      "step": 388
+    },
+    {
+      "epoch": 0.3828740157480315,
+      "grad_norm": 18.457351684570312,
+      "learning_rate": 3.8369781312127244e-07,
+      "loss": 1.3454,
+      "step": 389
+    },
+    {
+      "epoch": 0.3838582677165354,
+      "grad_norm": 22.10936737060547,
+      "learning_rate": 3.8469184890656075e-07,
+      "loss": 1.4292,
+      "step": 390
+    },
+    {
+      "epoch": 0.38484251968503935,
+      "grad_norm": 18.214380264282227,
+      "learning_rate": 3.8568588469184895e-07,
+      "loss": 1.2256,
+      "step": 391
+    },
+    {
+      "epoch": 0.3858267716535433,
+      "grad_norm": 19.54240608215332,
+      "learning_rate": 3.8667992047713725e-07,
+      "loss": 1.08,
+      "step": 392
+    },
+    {
+      "epoch": 0.3868110236220472,
+      "grad_norm": 15.17556095123291,
+      "learning_rate": 3.876739562624255e-07,
+      "loss": 0.7436,
+      "step": 393
+    },
+    {
+      "epoch": 0.38779527559055116,
+      "grad_norm": 17.79245948791504,
+      "learning_rate": 3.886679920477138e-07,
+      "loss": 1.4112,
+      "step": 394
+    },
+    {
+      "epoch": 0.3887795275590551,
+      "grad_norm": 15.915721893310547,
+      "learning_rate": 3.89662027833002e-07,
+      "loss": 0.8917,
+      "step": 395
+    },
+    {
+      "epoch": 0.38976377952755903,
+      "grad_norm": 17.685527801513672,
+      "learning_rate": 3.906560636182903e-07,
+      "loss": 0.9955,
+      "step": 396
+    },
+    {
+      "epoch": 0.390748031496063,
+      "grad_norm": 18.575515747070312,
+      "learning_rate": 3.916500994035786e-07,
+      "loss": 1.2867,
+      "step": 397
+    },
+    {
+      "epoch": 0.39173228346456695,
+      "grad_norm": 19.448511123657227,
+      "learning_rate": 3.9264413518886687e-07,
+      "loss": 1.0683,
+      "step": 398
+    },
+    {
+      "epoch": 0.3927165354330709,
+      "grad_norm": 18.82157325744629,
+      "learning_rate": 3.936381709741551e-07,
+      "loss": 0.9355,
+      "step": 399
+    },
+    {
+      "epoch": 0.3937007874015748,
+      "grad_norm": 17.697452545166016,
+      "learning_rate": 3.946322067594434e-07,
+      "loss": 1.1153,
+      "step": 400
+    },
+    {
+      "epoch": 0.39468503937007876,
+      "grad_norm": 18.46881866455078,
+      "learning_rate": 3.956262425447317e-07,
+      "loss": 1.1724,
+      "step": 401
+    },
+    {
+      "epoch": 0.3956692913385827,
+      "grad_norm": 24.634397506713867,
+      "learning_rate": 3.9662027833001993e-07,
+      "loss": 1.4069,
+      "step": 402
+    },
+    {
+      "epoch": 0.39665354330708663,
+      "grad_norm": 20.34600257873535,
+      "learning_rate": 3.9761431411530824e-07,
+      "loss": 1.2546,
+      "step": 403
+    },
+    {
+      "epoch": 0.39763779527559057,
+      "grad_norm": 26.437278747558594,
+      "learning_rate": 3.986083499005965e-07,
+      "loss": 2.2862,
+      "step": 404
+    },
+    {
+      "epoch": 0.3986220472440945,
+      "grad_norm": 17.79900360107422,
+      "learning_rate": 3.9960238568588474e-07,
+      "loss": 1.2316,
+      "step": 405
+    },
+    {
+      "epoch": 0.39960629921259844,
+      "grad_norm": 24.866077423095703,
+      "learning_rate": 4.0059642147117305e-07,
+      "loss": 1.7876,
+      "step": 406
+    },
+    {
+      "epoch": 0.4005905511811024,
+      "grad_norm": 14.236503601074219,
+      "learning_rate": 4.015904572564613e-07,
+      "loss": 0.6936,
+      "step": 407
+    },
+    {
+      "epoch": 0.4015748031496063,
+      "grad_norm": 21.97986602783203,
+      "learning_rate": 4.0258449304174955e-07,
+      "loss": 1.3852,
+      "step": 408
+    },
+    {
+      "epoch": 0.40255905511811024,
+      "grad_norm": 19.722715377807617,
+      "learning_rate": 4.035785288270378e-07,
+      "loss": 1.9046,
+      "step": 409
+    },
+    {
+      "epoch": 0.4035433070866142,
+      "grad_norm": 18.966148376464844,
+      "learning_rate": 4.045725646123261e-07,
+      "loss": 1.4972,
+      "step": 410
+    },
+    {
+      "epoch": 0.4045275590551181,
+      "grad_norm": 13.10688304901123,
+      "learning_rate": 4.055666003976144e-07,
+      "loss": 0.5531,
+      "step": 411
+    },
+    {
+      "epoch": 0.40551181102362205,
+      "grad_norm": 16.64056968688965,
+      "learning_rate": 4.065606361829026e-07,
+      "loss": 1.3685,
+      "step": 412
+    },
+    {
+      "epoch": 0.406496062992126,
+      "grad_norm": 17.951778411865234,
+      "learning_rate": 4.075546719681909e-07,
+      "loss": 1.1367,
+      "step": 413
+    },
+    {
+      "epoch": 0.4074803149606299,
+      "grad_norm": 18.79546356201172,
+      "learning_rate": 4.085487077534792e-07,
+      "loss": 1.1304,
+      "step": 414
+    },
+    {
+      "epoch": 0.40846456692913385,
+      "grad_norm": 23.208480834960938,
+      "learning_rate": 4.095427435387675e-07,
+      "loss": 1.5953,
+      "step": 415
+    },
+    {
+      "epoch": 0.4094488188976378,
+      "grad_norm": 24.188060760498047,
+      "learning_rate": 4.1053677932405573e-07,
+      "loss": 2.0308,
+      "step": 416
+    },
+    {
+      "epoch": 0.4104330708661417,
+      "grad_norm": 24.551694869995117,
+      "learning_rate": 4.11530815109344e-07,
+      "loss": 1.7275,
+      "step": 417
+    },
+    {
+      "epoch": 0.41141732283464566,
+      "grad_norm": 17.98610496520996,
+      "learning_rate": 4.125248508946323e-07,
+      "loss": 0.9921,
+      "step": 418
+    },
+    {
+      "epoch": 0.4124015748031496,
+      "grad_norm": 18.42479705810547,
+      "learning_rate": 4.1351888667992054e-07,
+      "loss": 1.3418,
+      "step": 419
+    },
+    {
+      "epoch": 0.41338582677165353,
+      "grad_norm": 18.93707275390625,
+      "learning_rate": 4.1451292246520885e-07,
+      "loss": 1.108,
+      "step": 420
+    },
+    {
+      "epoch": 0.41437007874015747,
+      "grad_norm": 17.49553680419922,
+      "learning_rate": 4.1550695825049705e-07,
+      "loss": 1.4359,
+      "step": 421
+    },
+    {
+      "epoch": 0.4153543307086614,
+      "grad_norm": 18.783565521240234,
+      "learning_rate": 4.1650099403578535e-07,
+      "loss": 1.4537,
+      "step": 422
+    },
+    {
+      "epoch": 0.41633858267716534,
+      "grad_norm": 16.676990509033203,
+      "learning_rate": 4.174950298210736e-07,
+      "loss": 0.8416,
+      "step": 423
+    },
+    {
+      "epoch": 0.41732283464566927,
+      "grad_norm": 15.969378471374512,
+      "learning_rate": 4.184890656063619e-07,
+      "loss": 0.8904,
+      "step": 424
+    },
+    {
+      "epoch": 0.4183070866141732,
+      "grad_norm": 15.156847953796387,
+      "learning_rate": 4.194831013916502e-07,
+      "loss": 0.7937,
+      "step": 425
+    },
+    {
+      "epoch": 0.41929133858267714,
+      "grad_norm": 16.689002990722656,
+      "learning_rate": 4.204771371769384e-07,
+      "loss": 0.9105,
+      "step": 426
+    },
+    {
+      "epoch": 0.4202755905511811,
+      "grad_norm": 18.2681884765625,
+      "learning_rate": 4.214711729622267e-07,
+      "loss": 1.1661,
+      "step": 427
+    },
+    {
+      "epoch": 0.421259842519685,
+      "grad_norm": 17.457895278930664,
+      "learning_rate": 4.2246520874751497e-07,
+      "loss": 0.7751,
+      "step": 428
+    },
+    {
+      "epoch": 0.422244094488189,
+      "grad_norm": 16.101627349853516,
+      "learning_rate": 4.234592445328033e-07,
+      "loss": 0.9039,
+      "step": 429
+    },
+    {
+      "epoch": 0.42322834645669294,
+      "grad_norm": 21.5301513671875,
+      "learning_rate": 4.244532803180915e-07,
+      "loss": 1.2651,
+      "step": 430
+    },
+    {
+      "epoch": 0.4242125984251969,
+      "grad_norm": 24.272584915161133,
+      "learning_rate": 4.254473161033798e-07,
+      "loss": 1.44,
+      "step": 431
+    },
+    {
+      "epoch": 0.4251968503937008,
+      "grad_norm": 18.749507904052734,
+      "learning_rate": 4.264413518886681e-07,
+      "loss": 0.9795,
+      "step": 432
+    },
+    {
+      "epoch": 0.42618110236220474,
+      "grad_norm": 29.54834747314453,
+      "learning_rate": 4.2743538767395634e-07,
+      "loss": 2.1892,
+      "step": 433
+    },
+    {
+      "epoch": 0.4271653543307087,
+      "grad_norm": 21.851760864257812,
+      "learning_rate": 4.284294234592446e-07,
+      "loss": 1.214,
+      "step": 434
+    },
+    {
+      "epoch": 0.4281496062992126,
+      "grad_norm": 21.678668975830078,
+      "learning_rate": 4.2942345924453284e-07,
+      "loss": 1.185,
+      "step": 435
+    },
+    {
+      "epoch": 0.42913385826771655,
+      "grad_norm": 20.97591781616211,
+      "learning_rate": 4.3041749502982115e-07,
+      "loss": 1.2501,
+      "step": 436
+    },
+    {
+      "epoch": 0.4301181102362205,
+      "grad_norm": 20.558670043945312,
+      "learning_rate": 4.3141153081510945e-07,
+      "loss": 1.6432,
+      "step": 437
+    },
+    {
+      "epoch": 0.4311023622047244,
+      "grad_norm": 18.290483474731445,
+      "learning_rate": 4.324055666003977e-07,
+      "loss": 1.0203,
+      "step": 438
+    },
+    {
+      "epoch": 0.43208661417322836,
+      "grad_norm": 21.71299171447754,
+      "learning_rate": 4.3339960238568596e-07,
+      "loss": 1.5179,
+      "step": 439
+    },
+    {
+      "epoch": 0.4330708661417323,
+      "grad_norm": 19.34381866455078,
+      "learning_rate": 4.343936381709742e-07,
+      "loss": 1.1445,
+      "step": 440
+    },
+    {
+      "epoch": 0.4340551181102362,
+      "grad_norm": 17.65093994140625,
+      "learning_rate": 4.353876739562625e-07,
+      "loss": 1.3099,
+      "step": 441
+    },
+    {
+      "epoch": 0.43503937007874016,
+      "grad_norm": 16.550668716430664,
+      "learning_rate": 4.3638170974155077e-07,
+      "loss": 0.8856,
+      "step": 442
+    },
+    {
+      "epoch": 0.4360236220472441,
+      "grad_norm": 11.908206939697266,
+      "learning_rate": 4.37375745526839e-07,
+      "loss": 0.5869,
+      "step": 443
+    },
+    {
+      "epoch": 0.43700787401574803,
+      "grad_norm": 23.472492218017578,
+      "learning_rate": 4.383697813121273e-07,
+      "loss": 1.6335,
+      "step": 444
+    },
+    {
+      "epoch": 0.43799212598425197,
+      "grad_norm": 21.598403930664062,
+      "learning_rate": 4.393638170974156e-07,
+      "loss": 1.4134,
+      "step": 445
+    },
+    {
+      "epoch": 0.4389763779527559,
+      "grad_norm": 19.63410186767578,
+      "learning_rate": 4.403578528827039e-07,
+      "loss": 1.0244,
+      "step": 446
+    },
+    {
+      "epoch": 0.43996062992125984,
+      "grad_norm": 18.266443252563477,
+      "learning_rate": 4.413518886679921e-07,
+      "loss": 1.103,
+      "step": 447
+    },
+    {
+      "epoch": 0.4409448818897638,
+      "grad_norm": 21.08526611328125,
+      "learning_rate": 4.423459244532804e-07,
+      "loss": 0.9848,
+      "step": 448
+    },
+    {
+      "epoch": 0.4419291338582677,
+      "grad_norm": 24.54269027709961,
+      "learning_rate": 4.4333996023856864e-07,
+      "loss": 1.5089,
+      "step": 449
+    },
+    {
+      "epoch": 0.44291338582677164,
+      "grad_norm": 19.617361068725586,
+      "learning_rate": 4.4433399602385695e-07,
+      "loss": 1.0422,
+      "step": 450
+    },
+    {
+      "epoch": 0.4438976377952756,
+      "grad_norm": 20.60758399963379,
+      "learning_rate": 4.4532803180914525e-07,
+      "loss": 1.0462,
+      "step": 451
+    },
+    {
+      "epoch": 0.4448818897637795,
+      "grad_norm": 20.939077377319336,
+      "learning_rate": 4.4632206759443345e-07,
+      "loss": 1.2857,
+      "step": 452
+    },
+    {
+      "epoch": 0.44586614173228345,
+      "grad_norm": 19.295501708984375,
+      "learning_rate": 4.4731610337972176e-07,
+      "loss": 1.4132,
+      "step": 453
+    },
+    {
+      "epoch": 0.4468503937007874,
+      "grad_norm": 22.07086753845215,
+      "learning_rate": 4.4831013916501e-07,
+      "loss": 1.3061,
+      "step": 454
+    },
+    {
+      "epoch": 0.4478346456692913,
+      "grad_norm": 23.644939422607422,
+      "learning_rate": 4.493041749502983e-07,
+      "loss": 1.3977,
+      "step": 455
+    },
+    {
+      "epoch": 0.44881889763779526,
+      "grad_norm": 21.088239669799805,
+      "learning_rate": 4.502982107355865e-07,
+      "loss": 1.3557,
+      "step": 456
+    },
+    {
+      "epoch": 0.4498031496062992,
+      "grad_norm": 23.44856834411621,
+      "learning_rate": 4.512922465208748e-07,
+      "loss": 1.3595,
+      "step": 457
+    },
+    {
+      "epoch": 0.4507874015748031,
+      "grad_norm": 16.646135330200195,
+      "learning_rate": 4.522862823061631e-07,
+      "loss": 0.8647,
+      "step": 458
+    },
+    {
+      "epoch": 0.45177165354330706,
+      "grad_norm": 21.29861831665039,
+      "learning_rate": 4.532803180914514e-07,
+      "loss": 1.3905,
+      "step": 459
+    },
+    {
+      "epoch": 0.45177165354330706,
+      "eval_Qnli-dev_cosine_accuracy": 0.677734375,
+      "eval_Qnli-dev_cosine_accuracy_threshold": 0.7905216217041016,
+      "eval_Qnli-dev_cosine_ap": 0.7092754004450792,
+      "eval_Qnli-dev_cosine_f1": 0.7005444646098005,
+      "eval_Qnli-dev_cosine_f1_threshold": 0.7738929986953735,
+      "eval_Qnli-dev_cosine_precision": 0.6126984126984127,
+      "eval_Qnli-dev_cosine_recall": 0.8177966101694916,
+      "eval_Qnli-dev_dot_accuracy": 0.677734375,
+      "eval_Qnli-dev_dot_accuracy_threshold": 607.5902099609375,
+      "eval_Qnli-dev_dot_ap": 0.7093564593718997,
+      "eval_Qnli-dev_dot_f1": 0.7005444646098005,
+      "eval_Qnli-dev_dot_f1_threshold": 594.82958984375,
+      "eval_Qnli-dev_dot_precision": 0.6126984126984127,
+      "eval_Qnli-dev_dot_recall": 0.8177966101694916,
+      "eval_Qnli-dev_euclidean_accuracy": 0.677734375,
+      "eval_Qnli-dev_euclidean_accuracy_threshold": 17.944761276245117,
+      "eval_Qnli-dev_euclidean_ap": 0.7092822066603943,
+      "eval_Qnli-dev_euclidean_f1": 0.7005444646098005,
+      "eval_Qnli-dev_euclidean_f1_threshold": 18.64351463317871,
+      "eval_Qnli-dev_euclidean_precision": 0.6126984126984127,
+      "eval_Qnli-dev_euclidean_recall": 0.8177966101694916,
+      "eval_Qnli-dev_manhattan_accuracy": 0.677734375,
+      "eval_Qnli-dev_manhattan_accuracy_threshold": 403.15380859375,
+      "eval_Qnli-dev_manhattan_ap": 0.7081773921924046,
+      "eval_Qnli-dev_manhattan_f1": 0.696113074204947,
+      "eval_Qnli-dev_manhattan_f1_threshold": 417.12298583984375,
+      "eval_Qnli-dev_manhattan_precision": 0.5969696969696969,
+      "eval_Qnli-dev_manhattan_recall": 0.8347457627118644,
+      "eval_Qnli-dev_max_accuracy": 0.677734375,
+      "eval_Qnli-dev_max_accuracy_threshold": 607.5902099609375,
+      "eval_Qnli-dev_max_ap": 0.7093564593718997,
+      "eval_Qnli-dev_max_f1": 0.7005444646098005,
+      "eval_Qnli-dev_max_f1_threshold": 594.82958984375,
+      "eval_Qnli-dev_max_precision": 0.6126984126984127,
+      "eval_Qnli-dev_max_recall": 0.8347457627118644,
+      "eval_allNLI-dev_cosine_accuracy": 0.689453125,
+      "eval_allNLI-dev_cosine_accuracy_threshold": 0.9186867475509644,
+      "eval_allNLI-dev_cosine_ap": 0.493528404484353,
+      "eval_allNLI-dev_cosine_f1": 0.5283018867924528,
+      "eval_allNLI-dev_cosine_f1_threshold": 0.8205500245094299,
+      "eval_allNLI-dev_cosine_precision": 0.4144736842105263,
+      "eval_allNLI-dev_cosine_recall": 0.7283236994219653,
+      "eval_allNLI-dev_dot_accuracy": 0.689453125,
+      "eval_allNLI-dev_dot_accuracy_threshold": 706.1145629882812,
+      "eval_allNLI-dev_dot_ap": 0.49366530598882574,
+      "eval_allNLI-dev_dot_f1": 0.5283018867924528,
+      "eval_allNLI-dev_dot_f1_threshold": 630.7012939453125,
+      "eval_allNLI-dev_dot_precision": 0.4144736842105263,
+      "eval_allNLI-dev_dot_recall": 0.7283236994219653,
+      "eval_allNLI-dev_euclidean_accuracy": 0.689453125,
+      "eval_allNLI-dev_euclidean_accuracy_threshold": 11.180255889892578,
+      "eval_allNLI-dev_euclidean_ap": 0.49351830756144216,
+      "eval_allNLI-dev_euclidean_f1": 0.5283018867924528,
+      "eval_allNLI-dev_euclidean_f1_threshold": 16.60896873474121,
+      "eval_allNLI-dev_euclidean_precision": 0.4144736842105263,
+      "eval_allNLI-dev_euclidean_recall": 0.7283236994219653,
+      "eval_allNLI-dev_manhattan_accuracy": 0.69140625,
+      "eval_allNLI-dev_manhattan_accuracy_threshold": 251.75421142578125,
+      "eval_allNLI-dev_manhattan_ap": 0.4930222958157673,
+      "eval_allNLI-dev_manhattan_f1": 0.5285412262156448,
+      "eval_allNLI-dev_manhattan_f1_threshold": 364.7342529296875,
+      "eval_allNLI-dev_manhattan_precision": 0.4166666666666667,
+      "eval_allNLI-dev_manhattan_recall": 0.7225433526011561,
+      "eval_allNLI-dev_max_accuracy": 0.69140625,
+      "eval_allNLI-dev_max_accuracy_threshold": 706.1145629882812,
+      "eval_allNLI-dev_max_ap": 0.49366530598882574,
+      "eval_allNLI-dev_max_f1": 0.5285412262156448,
+      "eval_allNLI-dev_max_f1_threshold": 630.7012939453125,
+      "eval_allNLI-dev_max_precision": 0.4166666666666667,
+      "eval_allNLI-dev_max_recall": 0.7283236994219653,
+      "eval_loss": 1.2969162464141846,
+      "eval_runtime": 50.0648,
+      "eval_samples_per_second": 33.237,
+      "eval_sequential_score": 0.7093564593718997,
+      "eval_steps_per_second": 0.14,
+      "eval_sts-test_pearson_cosine": 0.5415735956899457,
+      "eval_sts-test_pearson_dot": 0.5415314946904102,
+      "eval_sts-test_pearson_euclidean": 0.5463232738462487,
+      "eval_sts-test_pearson_manhattan": 0.5451734598739164,
+      "eval_sts-test_pearson_max": 0.5463232738462487,
+      "eval_sts-test_spearman_cosine": 0.5432558343516091,
+      "eval_sts-test_spearman_dot": 0.543190094748625,
+      "eval_sts-test_spearman_euclidean": 0.5433125789510533,
+      "eval_sts-test_spearman_manhattan": 0.5424313371117337,
+      "eval_sts-test_spearman_max": 0.5433125789510533,
+      "step": 459
+    },
+    {
+      "epoch": 0.452755905511811,
+      "grad_norm": 19.235441207885742,
+      "learning_rate": 4.542743538767397e-07,
+      "loss": 0.9467,
+      "step": 460
+    },
+    {
+      "epoch": 0.453740157480315,
+      "grad_norm": 26.713695526123047,
+      "learning_rate": 4.552683896620279e-07,
+      "loss": 1.9372,
+      "step": 461
+    },
+    {
+      "epoch": 0.4547244094488189,
+      "grad_norm": 13.988560676574707,
+      "learning_rate": 4.562624254473162e-07,
+      "loss": 0.871,
+      "step": 462
+    },
+    {
+      "epoch": 0.45570866141732286,
+      "grad_norm": 20.346981048583984,
+      "learning_rate": 4.5725646123260444e-07,
+      "loss": 1.2282,
+      "step": 463
+    },
+    {
+      "epoch": 0.4566929133858268,
+      "grad_norm": 18.093189239501953,
+      "learning_rate": 4.5825049701789274e-07,
+      "loss": 1.3845,
+      "step": 464
+    },
+    {
+      "epoch": 0.45767716535433073,
+      "grad_norm": 20.26774024963379,
+      "learning_rate": 4.59244532803181e-07,
+      "loss": 1.2571,
+      "step": 465
+    },
+    {
+      "epoch": 0.45866141732283466,
+      "grad_norm": 22.511838912963867,
+      "learning_rate": 4.6023856858846925e-07,
+      "loss": 1.2288,
+      "step": 466
+    },
+    {
+      "epoch": 0.4596456692913386,
+      "grad_norm": 15.89570140838623,
+      "learning_rate": 4.6123260437375755e-07,
+      "loss": 1.1165,
+      "step": 467
+    },
+    {
+      "epoch": 0.46062992125984253,
+      "grad_norm": 21.215049743652344,
+      "learning_rate": 4.622266401590458e-07,
+      "loss": 1.8463,
+      "step": 468
+    },
+    {
+      "epoch": 0.46161417322834647,
+      "grad_norm": 18.364500045776367,
+      "learning_rate": 4.6322067594433406e-07,
+      "loss": 0.9158,
+      "step": 469
+    },
+    {
+      "epoch": 0.4625984251968504,
+      "grad_norm": 18.762876510620117,
+      "learning_rate": 4.642147117296223e-07,
+      "loss": 0.8711,
+      "step": 470
+    },
+    {
+      "epoch": 0.46358267716535434,
+      "grad_norm": 19.44731330871582,
+      "learning_rate": 4.652087475149106e-07,
+      "loss": 1.4741,
+      "step": 471
+    },
+    {
+      "epoch": 0.4645669291338583,
+      "grad_norm": 15.906646728515625,
+      "learning_rate": 4.662027833001989e-07,
+      "loss": 0.914,
+      "step": 472
+    },
+    {
+      "epoch": 0.4655511811023622,
+      "grad_norm": 15.942811012268066,
+      "learning_rate": 4.671968190854872e-07,
+      "loss": 0.9435,
+      "step": 473
+    },
+    {
+      "epoch": 0.46653543307086615,
+      "grad_norm": 17.3239803314209,
+      "learning_rate": 4.681908548707754e-07,
+      "loss": 1.0876,
+      "step": 474
+    },
+    {
+      "epoch": 0.4675196850393701,
+      "grad_norm": 22.522132873535156,
+      "learning_rate": 4.691848906560637e-07,
+      "loss": 1.2365,
+      "step": 475
+    },
+    {
+      "epoch": 0.468503937007874,
+      "grad_norm": 18.67418098449707,
+      "learning_rate": 4.70178926441352e-07,
+      "loss": 1.1237,
+      "step": 476
+    },
+    {
+      "epoch": 0.46948818897637795,
+      "grad_norm": 22.97576332092285,
+      "learning_rate": 4.7117296222664024e-07,
+      "loss": 1.0097,
+      "step": 477
+    },
+    {
+      "epoch": 0.4704724409448819,
+      "grad_norm": 18.241655349731445,
+      "learning_rate": 4.721669980119285e-07,
+      "loss": 1.1548,
+      "step": 478
+    },
+    {
+      "epoch": 0.4714566929133858,
+      "grad_norm": 23.206424713134766,
+      "learning_rate": 4.731610337972168e-07,
+      "loss": 1.3203,
+      "step": 479
+    },
+    {
+      "epoch": 0.47244094488188976,
+      "grad_norm": 20.233009338378906,
+      "learning_rate": 4.7415506958250505e-07,
+      "loss": 1.2533,
+      "step": 480
+    },
+    {
+      "epoch": 0.4734251968503937,
+      "grad_norm": 20.482505798339844,
+      "learning_rate": 4.7514910536779335e-07,
+      "loss": 1.093,
+      "step": 481
+    },
+    {
+      "epoch": 0.4744094488188976,
+      "grad_norm": 19.419912338256836,
+      "learning_rate": 4.7614314115308155e-07,
+      "loss": 1.2591,
+      "step": 482
+    },
+    {
+      "epoch": 0.47539370078740156,
+      "grad_norm": 15.261035919189453,
+      "learning_rate": 4.771371769383698e-07,
+      "loss": 0.6764,
+      "step": 483
+    },
+    {
+      "epoch": 0.4763779527559055,
+      "grad_norm": 15.785579681396484,
+      "learning_rate": 4.781312127236581e-07,
+      "loss": 0.8922,
+      "step": 484
+    },
+    {
+      "epoch": 0.47736220472440943,
+      "grad_norm": 15.987274169921875,
+      "learning_rate": 4.791252485089464e-07,
+      "loss": 0.8524,
+      "step": 485
+    },
+    {
+      "epoch": 0.47834645669291337,
+      "grad_norm": 21.142431259155273,
+      "learning_rate": 4.801192842942347e-07,
+      "loss": 1.2777,
+      "step": 486
+    },
+    {
+      "epoch": 0.4793307086614173,
+      "grad_norm": 18.106870651245117,
+      "learning_rate": 4.811133200795229e-07,
+      "loss": 1.1682,
+      "step": 487
+    },
+    {
+      "epoch": 0.48031496062992124,
+      "grad_norm": 19.057527542114258,
+      "learning_rate": 4.821073558648112e-07,
+      "loss": 0.8617,
+      "step": 488
+    },
+    {
+      "epoch": 0.4812992125984252,
+      "grad_norm": 18.410564422607422,
+      "learning_rate": 4.831013916500995e-07,
+      "loss": 1.0303,
+      "step": 489
+    },
+    {
+      "epoch": 0.4822834645669291,
+      "grad_norm": 20.72304916381836,
+      "learning_rate": 4.840954274353877e-07,
+      "loss": 0.9843,
+      "step": 490
+    },
+    {
+      "epoch": 0.48326771653543305,
+      "grad_norm": 20.666017532348633,
+      "learning_rate": 4.85089463220676e-07,
+      "loss": 1.2951,
+      "step": 491
+    },
+    {
+      "epoch": 0.484251968503937,
+      "grad_norm": 23.45387077331543,
+      "learning_rate": 4.860834990059642e-07,
+      "loss": 1.7889,
+      "step": 492
+    },
+    {
+      "epoch": 0.48523622047244097,
+      "grad_norm": 18.90536117553711,
+      "learning_rate": 4.870775347912525e-07,
+      "loss": 1.118,
+      "step": 493
+    },
+    {
+      "epoch": 0.4862204724409449,
+      "grad_norm": 14.932930946350098,
+      "learning_rate": 4.880715705765408e-07,
+      "loss": 0.6772,
+      "step": 494
+    },
+    {
+      "epoch": 0.48720472440944884,
+      "grad_norm": 21.98814582824707,
+      "learning_rate": 4.890656063618291e-07,
+      "loss": 1.5058,
+      "step": 495
+    },
+    {
+      "epoch": 0.4881889763779528,
+      "grad_norm": 15.60319995880127,
+      "learning_rate": 4.900596421471173e-07,
+      "loss": 1.0068,
+      "step": 496
+    },
+    {
+      "epoch": 0.4891732283464567,
+      "grad_norm": 15.000826835632324,
+      "learning_rate": 4.910536779324057e-07,
+      "loss": 0.9024,
+      "step": 497
+    },
+    {
+      "epoch": 0.49015748031496065,
+      "grad_norm": 20.488412857055664,
+      "learning_rate": 4.92047713717694e-07,
+      "loss": 1.4816,
+      "step": 498
+    },
+    {
+      "epoch": 0.4911417322834646,
+      "grad_norm": 16.296037673950195,
+      "learning_rate": 4.930417495029823e-07,
+      "loss": 0.894,
+      "step": 499
+    },
+    {
+      "epoch": 0.4921259842519685,
+      "grad_norm": 16.42416000366211,
+      "learning_rate": 4.940357852882705e-07,
+      "loss": 1.1582,
+      "step": 500
+    },
+    {
+      "epoch": 0.49311023622047245,
+      "grad_norm": 19.995149612426758,
+      "learning_rate": 4.950298210735588e-07,
+      "loss": 1.4804,
+      "step": 501
+    },
+    {
+      "epoch": 0.4940944881889764,
+      "grad_norm": 16.41611671447754,
+      "learning_rate": 4.96023856858847e-07,
+      "loss": 1.2636,
+      "step": 502
+    },
+    {
+      "epoch": 0.4950787401574803,
+      "grad_norm": 16.084943771362305,
+      "learning_rate": 4.970178926441353e-07,
+      "loss": 1.0094,
+      "step": 503
+    },
+    {
+      "epoch": 0.49606299212598426,
+      "grad_norm": 14.031723022460938,
+      "learning_rate": 4.980119284294235e-07,
+      "loss": 0.7594,
+      "step": 504
+    },
+    {
+      "epoch": 0.4970472440944882,
+      "grad_norm": 18.90411949157715,
+      "learning_rate": 4.990059642147118e-07,
+      "loss": 1.2898,
+      "step": 505
+    },
+    {
+      "epoch": 0.49803149606299213,
+      "grad_norm": 19.5909423828125,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.3565,
+      "step": 506
+    },
+    {
+      "epoch": 0.49901574803149606,
+      "grad_norm": 16.011884689331055,
+      "learning_rate": 5.009940357852883e-07,
+      "loss": 1.0325,
+      "step": 507
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 16.250404357910156,
+      "learning_rate": 5.019880715705767e-07,
+      "loss": 1.0519,
+      "step": 508
+    },
+    {
+      "epoch": 0.5009842519685039,
+      "grad_norm": 16.943971633911133,
+      "learning_rate": 5.029821073558649e-07,
+      "loss": 0.9802,
+      "step": 509
+    },
+    {
+      "epoch": 0.5019685039370079,
+      "grad_norm": 17.802446365356445,
+      "learning_rate": 5.039761431411532e-07,
+      "loss": 1.1117,
+      "step": 510
+    },
+    {
+      "epoch": 0.5029527559055118,
+      "grad_norm": 18.555269241333008,
+      "learning_rate": 5.049701789264414e-07,
+      "loss": 1.3585,
+      "step": 511
+    },
+    {
+      "epoch": 0.5039370078740157,
+      "grad_norm": 15.964585304260254,
+      "learning_rate": 5.059642147117297e-07,
+      "loss": 1.0381,
+      "step": 512
+    },
+    {
+      "epoch": 0.5049212598425197,
+      "grad_norm": 14.011932373046875,
+      "learning_rate": 5.06958250497018e-07,
+      "loss": 1.0171,
+      "step": 513
+    },
+    {
+      "epoch": 0.5059055118110236,
+      "grad_norm": 12.088032722473145,
+      "learning_rate": 5.079522862823062e-07,
+      "loss": 0.5678,
+      "step": 514
+    },
+    {
+      "epoch": 0.5068897637795275,
+      "grad_norm": 17.39454460144043,
+      "learning_rate": 5.089463220675945e-07,
+      "loss": 0.9347,
+      "step": 515
+    },
+    {
+      "epoch": 0.5078740157480315,
+      "grad_norm": 14.170563697814941,
+      "learning_rate": 5.099403578528828e-07,
+      "loss": 0.6305,
+      "step": 516
+    },
+    {
+      "epoch": 0.5088582677165354,
+      "grad_norm": 12.88972282409668,
+      "learning_rate": 5.10934393638171e-07,
+      "loss": 0.7072,
+      "step": 517
+    },
+    {
+      "epoch": 0.5098425196850394,
+      "grad_norm": 16.725210189819336,
+      "learning_rate": 5.119284294234593e-07,
+      "loss": 0.9746,
+      "step": 518
+    },
+    {
+      "epoch": 0.5108267716535433,
+      "grad_norm": 19.107908248901367,
+      "learning_rate": 5.129224652087476e-07,
+      "loss": 1.1782,
+      "step": 519
+    },
+    {
+      "epoch": 0.5118110236220472,
+      "grad_norm": 19.989774703979492,
+      "learning_rate": 5.139165009940358e-07,
+      "loss": 1.1354,
+      "step": 520
+    },
+    {
+      "epoch": 0.5127952755905512,
+      "grad_norm": 21.850637435913086,
+      "learning_rate": 5.149105367793241e-07,
+      "loss": 1.5752,
+      "step": 521
+    },
+    {
+      "epoch": 0.5137795275590551,
+      "grad_norm": 14.340449333190918,
+      "learning_rate": 5.159045725646124e-07,
+      "loss": 0.5952,
+      "step": 522
+    },
+    {
+      "epoch": 0.514763779527559,
+      "grad_norm": 18.608619689941406,
+      "learning_rate": 5.168986083499007e-07,
+      "loss": 1.1171,
+      "step": 523
+    },
+    {
+      "epoch": 0.515748031496063,
+      "grad_norm": 18.41301918029785,
+      "learning_rate": 5.178926441351889e-07,
+      "loss": 0.8234,
+      "step": 524
+    },
+    {
+      "epoch": 0.5167322834645669,
+      "grad_norm": 25.120445251464844,
+      "learning_rate": 5.188866799204771e-07,
+      "loss": 1.6701,
+      "step": 525
+    },
+    {
+      "epoch": 0.5177165354330708,
+      "grad_norm": 18.166357040405273,
+      "learning_rate": 5.198807157057656e-07,
+      "loss": 1.2111,
+      "step": 526
+    },
+    {
+      "epoch": 0.5187007874015748,
+      "grad_norm": 16.0269718170166,
+      "learning_rate": 5.208747514910538e-07,
+      "loss": 0.8299,
+      "step": 527
+    },
+    {
+      "epoch": 0.5196850393700787,
+      "grad_norm": 22.35710334777832,
+      "learning_rate": 5.21868787276342e-07,
+      "loss": 1.5734,
+      "step": 528
+    },
+    {
+      "epoch": 0.5206692913385826,
+      "grad_norm": 14.45018196105957,
+      "learning_rate": 5.228628230616304e-07,
+      "loss": 0.9172,
+      "step": 529
+    },
+    {
+      "epoch": 0.5216535433070866,
+      "grad_norm": 15.022900581359863,
+      "learning_rate": 5.238568588469186e-07,
+      "loss": 0.8025,
+      "step": 530
+    },
+    {
+      "epoch": 0.5226377952755905,
+      "grad_norm": 18.771406173706055,
+      "learning_rate": 5.248508946322069e-07,
+      "loss": 1.1499,
+      "step": 531
+    },
+    {
+      "epoch": 0.5236220472440944,
+      "grad_norm": 19.911235809326172,
+      "learning_rate": 5.258449304174951e-07,
+      "loss": 1.0328,
+      "step": 532
+    },
+    {
+      "epoch": 0.5246062992125984,
+      "grad_norm": 19.31245231628418,
+      "learning_rate": 5.268389662027834e-07,
+      "loss": 1.1305,
+      "step": 533
+    },
+    {
+      "epoch": 0.5255905511811023,
+      "grad_norm": 14.231760025024414,
+      "learning_rate": 5.278330019880717e-07,
+      "loss": 0.6715,
+      "step": 534
+    },
+    {
+      "epoch": 0.5265748031496063,
+      "grad_norm": 16.54633331298828,
+      "learning_rate": 5.288270377733599e-07,
+      "loss": 1.1361,
+      "step": 535
+    },
+    {
+      "epoch": 0.5275590551181102,
+      "grad_norm": 19.317935943603516,
+      "learning_rate": 5.298210735586483e-07,
+      "loss": 0.9132,
+      "step": 536
+    },
+    {
+      "epoch": 0.5285433070866141,
+      "grad_norm": 20.749347686767578,
+      "learning_rate": 5.308151093439365e-07,
+      "loss": 1.2195,
+      "step": 537
+    },
+    {
+      "epoch": 0.5295275590551181,
+      "grad_norm": 12.56061840057373,
+      "learning_rate": 5.318091451292247e-07,
+      "loss": 0.3731,
+      "step": 538
+    },
+    {
+      "epoch": 0.530511811023622,
+      "grad_norm": 17.94748878479004,
+      "learning_rate": 5.32803180914513e-07,
+      "loss": 1.0005,
+      "step": 539
+    },
+    {
+      "epoch": 0.531496062992126,
+      "grad_norm": 15.667882919311523,
+      "learning_rate": 5.337972166998013e-07,
+      "loss": 0.5519,
+      "step": 540
+    },
+    {
+      "epoch": 0.53248031496063,
+      "grad_norm": 18.331357955932617,
+      "learning_rate": 5.347912524850895e-07,
+      "loss": 0.7529,
+      "step": 541
+    },
+    {
+      "epoch": 0.5334645669291339,
+      "grad_norm": 28.6558895111084,
+      "learning_rate": 5.357852882703778e-07,
+      "loss": 1.7004,
+      "step": 542
+    },
+    {
+      "epoch": 0.5344488188976378,
+      "grad_norm": 25.073156356811523,
+      "learning_rate": 5.367793240556661e-07,
+      "loss": 1.4667,
+      "step": 543
+    },
+    {
+      "epoch": 0.5354330708661418,
+      "grad_norm": 19.546232223510742,
+      "learning_rate": 5.377733598409544e-07,
+      "loss": 0.8349,
+      "step": 544
+    },
+    {
+      "epoch": 0.5364173228346457,
+      "grad_norm": 24.613113403320312,
+      "learning_rate": 5.387673956262426e-07,
+      "loss": 1.5575,
+      "step": 545
+    },
+    {
+      "epoch": 0.5374015748031497,
+      "grad_norm": 23.96712875366211,
+      "learning_rate": 5.397614314115308e-07,
+      "loss": 1.1703,
+      "step": 546
+    },
+    {
+      "epoch": 0.5383858267716536,
+      "grad_norm": 18.8927001953125,
+      "learning_rate": 5.407554671968192e-07,
+      "loss": 1.01,
+      "step": 547
+    },
+    {
+      "epoch": 0.5393700787401575,
+      "grad_norm": 19.96955108642578,
+      "learning_rate": 5.417495029821074e-07,
+      "loss": 1.1114,
+      "step": 548
+    },
+    {
+      "epoch": 0.5403543307086615,
+      "grad_norm": 11.816988945007324,
+      "learning_rate": 5.427435387673957e-07,
+      "loss": 0.516,
+      "step": 549
+    },
+    {
+      "epoch": 0.5413385826771654,
+      "grad_norm": 18.070484161376953,
+      "learning_rate": 5.43737574552684e-07,
+      "loss": 1.0422,
+      "step": 550
+    },
+    {
+      "epoch": 0.5423228346456693,
+      "grad_norm": 16.6149845123291,
+      "learning_rate": 5.447316103379722e-07,
+      "loss": 1.078,
+      "step": 551
+    },
+    {
+      "epoch": 0.5433070866141733,
+      "grad_norm": 17.799449920654297,
+      "learning_rate": 5.457256461232605e-07,
+      "loss": 1.0573,
+      "step": 552
+    },
+    {
+      "epoch": 0.5442913385826772,
+      "grad_norm": 19.17674446105957,
+      "learning_rate": 5.467196819085487e-07,
+      "loss": 0.9754,
+      "step": 553
+    },
+    {
+      "epoch": 0.5452755905511811,
+      "grad_norm": 19.11102867126465,
+      "learning_rate": 5.477137176938371e-07,
+      "loss": 0.9227,
+      "step": 554
+    },
+    {
+      "epoch": 0.5462598425196851,
+      "grad_norm": 24.212202072143555,
+      "learning_rate": 5.487077534791253e-07,
+      "loss": 1.5012,
+      "step": 555
+    },
+    {
+      "epoch": 0.547244094488189,
+      "grad_norm": 16.03239631652832,
+      "learning_rate": 5.497017892644135e-07,
+      "loss": 1.0697,
+      "step": 556
+    },
+    {
+      "epoch": 0.5482283464566929,
+      "grad_norm": 21.717912673950195,
+      "learning_rate": 5.50695825049702e-07,
+      "loss": 1.4437,
+      "step": 557
+    },
+    {
+      "epoch": 0.5492125984251969,
+      "grad_norm": 18.13492202758789,
+      "learning_rate": 5.516898608349902e-07,
+      "loss": 1.0697,
+      "step": 558
+    },
+    {
+      "epoch": 0.5501968503937008,
+      "grad_norm": 15.775596618652344,
+      "learning_rate": 5.526838966202784e-07,
+      "loss": 0.8346,
+      "step": 559
+    },
+    {
+      "epoch": 0.5511811023622047,
+      "grad_norm": 16.541046142578125,
+      "learning_rate": 5.536779324055667e-07,
+      "loss": 0.6421,
+      "step": 560
+    },
+    {
+      "epoch": 0.5521653543307087,
+      "grad_norm": 16.008865356445312,
+      "learning_rate": 5.54671968190855e-07,
+      "loss": 0.6687,
+      "step": 561
+    },
+    {
+      "epoch": 0.5531496062992126,
+      "grad_norm": 18.417842864990234,
+      "learning_rate": 5.556660039761433e-07,
+      "loss": 0.982,
+      "step": 562
+    },
+    {
+      "epoch": 0.5541338582677166,
+      "grad_norm": 17.81726837158203,
+      "learning_rate": 5.566600397614315e-07,
+      "loss": 0.9299,
+      "step": 563
+    },
+    {
+      "epoch": 0.5551181102362205,
+      "grad_norm": 20.61170768737793,
+      "learning_rate": 5.576540755467198e-07,
+      "loss": 1.5852,
+      "step": 564
+    },
+    {
+      "epoch": 0.5561023622047244,
+      "grad_norm": 18.6957950592041,
+      "learning_rate": 5.586481113320081e-07,
+      "loss": 1.2132,
+      "step": 565
+    },
+    {
+      "epoch": 0.5570866141732284,
+      "grad_norm": 15.698697090148926,
+      "learning_rate": 5.596421471172963e-07,
+      "loss": 0.8426,
+      "step": 566
+    },
+    {
+      "epoch": 0.5580708661417323,
+      "grad_norm": 17.419189453125,
+      "learning_rate": 5.606361829025846e-07,
+      "loss": 1.0496,
+      "step": 567
+    },
+    {
+      "epoch": 0.5590551181102362,
+      "grad_norm": 15.717952728271484,
+      "learning_rate": 5.616302186878729e-07,
+      "loss": 1.0436,
+      "step": 568
+    },
+    {
+      "epoch": 0.5600393700787402,
+      "grad_norm": 13.559417724609375,
+      "learning_rate": 5.626242544731611e-07,
+      "loss": 0.806,
+      "step": 569
+    },
+    {
+      "epoch": 0.5610236220472441,
+      "grad_norm": 12.456329345703125,
+      "learning_rate": 5.636182902584494e-07,
+      "loss": 0.6396,
+      "step": 570
+    },
+    {
+      "epoch": 0.562007874015748,
+      "grad_norm": 21.304899215698242,
+      "learning_rate": 5.646123260437377e-07,
+      "loss": 1.6315,
+      "step": 571
+    },
+    {
+      "epoch": 0.562992125984252,
+      "grad_norm": 16.898534774780273,
+      "learning_rate": 5.656063618290259e-07,
+      "loss": 1.3286,
+      "step": 572
+    },
+    {
+      "epoch": 0.5639763779527559,
+      "grad_norm": 14.442305564880371,
+      "learning_rate": 5.666003976143142e-07,
+      "loss": 0.7682,
+      "step": 573
+    },
+    {
+      "epoch": 0.5649606299212598,
+      "grad_norm": 14.211771011352539,
+      "learning_rate": 5.675944333996024e-07,
+      "loss": 0.7861,
+      "step": 574
+    },
+    {
+      "epoch": 0.5659448818897638,
+      "grad_norm": 16.277984619140625,
+      "learning_rate": 5.685884691848908e-07,
+      "loss": 1.0368,
+      "step": 575
+    },
+    {
+      "epoch": 0.5669291338582677,
+      "grad_norm": 18.00242805480957,
+      "learning_rate": 5.69582504970179e-07,
+      "loss": 1.1497,
+      "step": 576
+    },
+    {
+      "epoch": 0.5679133858267716,
+      "grad_norm": 17.694448471069336,
+      "learning_rate": 5.705765407554672e-07,
+      "loss": 0.9691,
+      "step": 577
+    },
+    {
+      "epoch": 0.5688976377952756,
+      "grad_norm": 15.09469223022461,
+      "learning_rate": 5.715705765407556e-07,
+      "loss": 0.7447,
+      "step": 578
+    },
+    {
+      "epoch": 0.5698818897637795,
+      "grad_norm": 21.568681716918945,
+      "learning_rate": 5.725646123260438e-07,
+      "loss": 1.3933,
+      "step": 579
+    },
+    {
+      "epoch": 0.5708661417322834,
+      "grad_norm": 17.880390167236328,
+      "learning_rate": 5.735586481113321e-07,
+      "loss": 1.0668,
+      "step": 580
+    },
+    {
+      "epoch": 0.5718503937007874,
+      "grad_norm": 14.031826972961426,
+      "learning_rate": 5.745526838966203e-07,
+      "loss": 0.6065,
+      "step": 581
+    },
+    {
+      "epoch": 0.5728346456692913,
+      "grad_norm": 16.969873428344727,
+      "learning_rate": 5.755467196819086e-07,
+      "loss": 0.9566,
+      "step": 582
+    },
+    {
+      "epoch": 0.5738188976377953,
+      "grad_norm": 15.315662384033203,
+      "learning_rate": 5.765407554671969e-07,
+      "loss": 0.7957,
+      "step": 583
+    },
+    {
+      "epoch": 0.5748031496062992,
+      "grad_norm": 18.925649642944336,
+      "learning_rate": 5.775347912524851e-07,
+      "loss": 1.0232,
+      "step": 584
+    },
+    {
+      "epoch": 0.5757874015748031,
+      "grad_norm": 19.93893814086914,
+      "learning_rate": 5.785288270377734e-07,
+      "loss": 1.4559,
+      "step": 585
+    },
+    {
+      "epoch": 0.5767716535433071,
+      "grad_norm": 14.928529739379883,
+      "learning_rate": 5.795228628230618e-07,
+      "loss": 0.8003,
+      "step": 586
+    },
+    {
+      "epoch": 0.577755905511811,
+      "grad_norm": 18.931825637817383,
+      "learning_rate": 5.8051689860835e-07,
+      "loss": 0.9504,
+      "step": 587
+    },
+    {
+      "epoch": 0.5787401574803149,
+      "grad_norm": 21.6677188873291,
+      "learning_rate": 5.815109343936383e-07,
+      "loss": 1.5257,
+      "step": 588
+    },
+    {
+      "epoch": 0.5797244094488189,
+      "grad_norm": 13.008130073547363,
+      "learning_rate": 5.825049701789266e-07,
+      "loss": 0.5798,
+      "step": 589
+    },
+    {
+      "epoch": 0.5807086614173228,
+      "grad_norm": 17.05565071105957,
+      "learning_rate": 5.834990059642148e-07,
+      "loss": 0.8169,
+      "step": 590
+    },
+    {
+      "epoch": 0.5816929133858267,
+      "grad_norm": 17.504405975341797,
+      "learning_rate": 5.844930417495031e-07,
+      "loss": 1.1131,
+      "step": 591
+    },
+    {
+      "epoch": 0.5826771653543307,
+      "grad_norm": 19.08473014831543,
+      "learning_rate": 5.854870775347914e-07,
+      "loss": 1.2498,
+      "step": 592
+    },
+    {
+      "epoch": 0.5836614173228346,
+      "grad_norm": 16.938589096069336,
+      "learning_rate": 5.864811133200797e-07,
+      "loss": 0.8541,
+      "step": 593
+    },
+    {
+      "epoch": 0.5846456692913385,
+      "grad_norm": 18.441364288330078,
+      "learning_rate": 5.874751491053679e-07,
+      "loss": 1.0848,
+      "step": 594
+    },
+    {
+      "epoch": 0.5856299212598425,
+      "grad_norm": 16.72596549987793,
+      "learning_rate": 5.884691848906561e-07,
+      "loss": 0.8909,
+      "step": 595
+    },
+    {
+      "epoch": 0.5866141732283464,
+      "grad_norm": 15.525846481323242,
+      "learning_rate": 5.894632206759445e-07,
+      "loss": 0.7572,
+      "step": 596
+    },
+    {
+      "epoch": 0.5875984251968503,
+      "grad_norm": 22.291608810424805,
+      "learning_rate": 5.904572564612327e-07,
+      "loss": 1.3636,
+      "step": 597
+    },
+    {
+      "epoch": 0.5885826771653543,
+      "grad_norm": 14.620648384094238,
+      "learning_rate": 5.914512922465209e-07,
+      "loss": 0.8493,
+      "step": 598
+    },
+    {
+      "epoch": 0.5895669291338582,
+      "grad_norm": 15.47690486907959,
+      "learning_rate": 5.924453280318093e-07,
+      "loss": 0.9594,
+      "step": 599
+    },
+    {
+      "epoch": 0.5905511811023622,
+      "grad_norm": 21.12307357788086,
+      "learning_rate": 5.934393638170975e-07,
+      "loss": 1.1143,
+      "step": 600
+    },
+    {
+      "epoch": 0.5915354330708661,
+      "grad_norm": 14.08637809753418,
+      "learning_rate": 5.944333996023858e-07,
+      "loss": 0.7093,
+      "step": 601
+    },
+    {
+      "epoch": 0.59251968503937,
+      "grad_norm": 15.423602104187012,
+      "learning_rate": 5.95427435387674e-07,
+      "loss": 1.0542,
+      "step": 602
+    },
+    {
+      "epoch": 0.593503937007874,
+      "grad_norm": 17.396760940551758,
+      "learning_rate": 5.964214711729623e-07,
+      "loss": 1.0621,
+      "step": 603
+    },
+    {
+      "epoch": 0.594488188976378,
+      "grad_norm": 12.562100410461426,
+      "learning_rate": 5.974155069582506e-07,
+      "loss": 0.6916,
+      "step": 604
+    },
+    {
+      "epoch": 0.5954724409448819,
+      "grad_norm": 20.98981285095215,
+      "learning_rate": 5.984095427435388e-07,
+      "loss": 1.0125,
+      "step": 605
+    },
+    {
+      "epoch": 0.5964566929133859,
+      "grad_norm": 16.202804565429688,
+      "learning_rate": 5.994035785288272e-07,
+      "loss": 0.8425,
+      "step": 606
+    },
+    {
+      "epoch": 0.5974409448818898,
+      "grad_norm": 18.76996421813965,
+      "learning_rate": 6.003976143141154e-07,
+      "loss": 1.2868,
+      "step": 607
+    },
+    {
+      "epoch": 0.5984251968503937,
+      "grad_norm": 17.652177810668945,
+      "learning_rate": 6.013916500994036e-07,
+      "loss": 1.3505,
+      "step": 608
+    },
+    {
+      "epoch": 0.5994094488188977,
+      "grad_norm": 18.34078598022461,
+      "learning_rate": 6.02385685884692e-07,
+      "loss": 1.2699,
+      "step": 609
+    },
+    {
+      "epoch": 0.6003937007874016,
+      "grad_norm": 18.553659439086914,
+      "learning_rate": 6.033797216699802e-07,
+      "loss": 1.1798,
+      "step": 610
+    }
+  ],
+  "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
+}