diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8946 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.974690994702766,
+  "eval_steps": 500,
+  "global_step": 1272,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004708652148322542,
+      "grad_norm": 5.121918678283691,
+      "learning_rate": 7.8125e-08,
+      "loss": 0.9126,
+      "step": 1
+    },
+    {
+      "epoch": 0.009417304296645085,
+      "grad_norm": 5.174776077270508,
+      "learning_rate": 1.5625e-07,
+      "loss": 0.9156,
+      "step": 2
+    },
+    {
+      "epoch": 0.014125956444967627,
+      "grad_norm": 5.343864440917969,
+      "learning_rate": 2.3437500000000003e-07,
+      "loss": 0.9331,
+      "step": 3
+    },
+    {
+      "epoch": 0.01883460859329017,
+      "grad_norm": 5.055787563323975,
+      "learning_rate": 3.125e-07,
+      "loss": 0.9206,
+      "step": 4
+    },
+    {
+      "epoch": 0.023543260741612712,
+      "grad_norm": 5.8190765380859375,
+      "learning_rate": 3.90625e-07,
+      "loss": 1.0045,
+      "step": 5
+    },
+    {
+      "epoch": 0.028251912889935255,
+      "grad_norm": 5.174574375152588,
+      "learning_rate": 4.6875000000000006e-07,
+      "loss": 0.9213,
+      "step": 6
+    },
+    {
+      "epoch": 0.0329605650382578,
+      "grad_norm": 5.184887886047363,
+      "learning_rate": 5.468750000000001e-07,
+      "loss": 0.9327,
+      "step": 7
+    },
+    {
+      "epoch": 0.03766921718658034,
+      "grad_norm": 5.018477439880371,
+      "learning_rate": 6.25e-07,
+      "loss": 0.8872,
+      "step": 8
+    },
+    {
+      "epoch": 0.04237786933490288,
+      "grad_norm": 4.189849853515625,
+      "learning_rate": 7.03125e-07,
+      "loss": 0.8224,
+      "step": 9
+    },
+    {
+      "epoch": 0.047086521483225424,
+      "grad_norm": 4.744853973388672,
+      "learning_rate": 7.8125e-07,
+      "loss": 0.9135,
+      "step": 10
+    },
+    {
+      "epoch": 0.05179517363154797,
+      "grad_norm": 4.5790886878967285,
+      "learning_rate": 8.59375e-07,
+      "loss": 0.9163,
+      "step": 11
+    },
+    {
+      "epoch": 0.05650382577987051,
+      "grad_norm": 3.080500364303589,
+      "learning_rate": 9.375000000000001e-07,
+      "loss": 0.8982,
+      "step": 12
+    },
+    {
+      "epoch": 0.06121247792819305,
+      "grad_norm": 3.0505118370056152,
+      "learning_rate": 1.0156250000000001e-06,
+      "loss": 0.8778,
+      "step": 13
+    },
+    {
+      "epoch": 0.0659211300765156,
+      "grad_norm": 3.0311391353607178,
+      "learning_rate": 1.0937500000000001e-06,
+      "loss": 0.8865,
+      "step": 14
+    },
+    {
+      "epoch": 0.07062978222483814,
+      "grad_norm": 2.7465591430664062,
+      "learning_rate": 1.1718750000000001e-06,
+      "loss": 0.8397,
+      "step": 15
+    },
+    {
+      "epoch": 0.07533843437316068,
+      "grad_norm": 2.4795093536376953,
+      "learning_rate": 1.25e-06,
+      "loss": 0.8057,
+      "step": 16
+    },
+    {
+      "epoch": 0.08004708652148322,
+      "grad_norm": 2.3555572032928467,
+      "learning_rate": 1.328125e-06,
+      "loss": 0.7827,
+      "step": 17
+    },
+    {
+      "epoch": 0.08475573866980576,
+      "grad_norm": 2.6421377658843994,
+      "learning_rate": 1.40625e-06,
+      "loss": 0.8337,
+      "step": 18
+    },
+    {
+      "epoch": 0.0894643908181283,
+      "grad_norm": 2.405103921890259,
+      "learning_rate": 1.484375e-06,
+      "loss": 0.7903,
+      "step": 19
+    },
+    {
+      "epoch": 0.09417304296645085,
+      "grad_norm": 2.2684788703918457,
+      "learning_rate": 1.5625e-06,
+      "loss": 0.7974,
+      "step": 20
+    },
+    {
+      "epoch": 0.09888169511477339,
+      "grad_norm": 2.151175022125244,
+      "learning_rate": 1.640625e-06,
+      "loss": 0.7783,
+      "step": 21
+    },
+    {
+      "epoch": 0.10359034726309593,
+      "grad_norm": 2.6126651763916016,
+      "learning_rate": 1.71875e-06,
+      "loss": 0.7367,
+      "step": 22
+    },
+    {
+      "epoch": 0.10829899941141848,
+      "grad_norm": 2.518990993499756,
+      "learning_rate": 1.796875e-06,
+      "loss": 0.7653,
+      "step": 23
+    },
+    {
+      "epoch": 0.11300765155974102,
+      "grad_norm": 1.8703638315200806,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.7007,
+      "step": 24
+    },
+    {
+      "epoch": 0.11771630370806356,
+      "grad_norm": 2.162489652633667,
+      "learning_rate": 1.953125e-06,
+      "loss": 0.7049,
+      "step": 25
+    },
+    {
+      "epoch": 0.1224249558563861,
+      "grad_norm": 2.2163379192352295,
+      "learning_rate": 2.0312500000000002e-06,
+      "loss": 0.7253,
+      "step": 26
+    },
+    {
+      "epoch": 0.12713360800470866,
+      "grad_norm": 1.8689916133880615,
+      "learning_rate": 2.109375e-06,
+      "loss": 0.7046,
+      "step": 27
+    },
+    {
+      "epoch": 0.1318422601530312,
+      "grad_norm": 1.7359975576400757,
+      "learning_rate": 2.1875000000000002e-06,
+      "loss": 0.6888,
+      "step": 28
+    },
+    {
+      "epoch": 0.13655091230135374,
+      "grad_norm": 1.7069369554519653,
+      "learning_rate": 2.265625e-06,
+      "loss": 0.6856,
+      "step": 29
+    },
+    {
+      "epoch": 0.14125956444967627,
+      "grad_norm": 1.7160871028900146,
+      "learning_rate": 2.3437500000000002e-06,
+      "loss": 0.6859,
+      "step": 30
+    },
+    {
+      "epoch": 0.14596821659799883,
+      "grad_norm": 1.6506415605545044,
+      "learning_rate": 2.421875e-06,
+      "loss": 0.6525,
+      "step": 31
+    },
+    {
+      "epoch": 0.15067686874632136,
+      "grad_norm": 1.59455406665802,
+      "learning_rate": 2.5e-06,
+      "loss": 0.6459,
+      "step": 32
+    },
+    {
+      "epoch": 0.1553855208946439,
+      "grad_norm": 1.5882025957107544,
+      "learning_rate": 2.5781250000000004e-06,
+      "loss": 0.6538,
+      "step": 33
+    },
+    {
+      "epoch": 0.16009417304296644,
+      "grad_norm": 1.4623929262161255,
+      "learning_rate": 2.65625e-06,
+      "loss": 0.6292,
+      "step": 34
+    },
+    {
+      "epoch": 0.164802825191289,
+      "grad_norm": 1.6212574243545532,
+      "learning_rate": 2.7343750000000004e-06,
+      "loss": 0.6548,
+      "step": 35
+    },
+    {
+      "epoch": 0.16951147733961153,
+      "grad_norm": 1.4372034072875977,
+      "learning_rate": 2.8125e-06,
+      "loss": 0.6276,
+      "step": 36
+    },
+    {
+      "epoch": 0.17422012948793408,
+      "grad_norm": 1.4232583045959473,
+      "learning_rate": 2.8906250000000004e-06,
+      "loss": 0.6112,
+      "step": 37
+    },
+    {
+      "epoch": 0.1789287816362566,
+      "grad_norm": 1.3985645771026611,
+      "learning_rate": 2.96875e-06,
+      "loss": 0.6117,
+      "step": 38
+    },
+    {
+      "epoch": 0.18363743378457917,
+      "grad_norm": 1.4189324378967285,
+      "learning_rate": 3.0468750000000004e-06,
+      "loss": 0.6046,
+      "step": 39
+    },
+    {
+      "epoch": 0.1883460859329017,
+      "grad_norm": 1.5376724004745483,
+      "learning_rate": 3.125e-06,
+      "loss": 0.6116,
+      "step": 40
+    },
+    {
+      "epoch": 0.19305473808122425,
+      "grad_norm": 1.7375065088272095,
+      "learning_rate": 3.2031250000000004e-06,
+      "loss": 0.6435,
+      "step": 41
+    },
+    {
+      "epoch": 0.19776339022954678,
+      "grad_norm": 1.5353353023529053,
+      "learning_rate": 3.28125e-06,
+      "loss": 0.576,
+      "step": 42
+    },
+    {
+      "epoch": 0.20247204237786934,
+      "grad_norm": 1.706644058227539,
+      "learning_rate": 3.3593750000000003e-06,
+      "loss": 0.581,
+      "step": 43
+    },
+    {
+      "epoch": 0.20718069452619187,
+      "grad_norm": 1.9466679096221924,
+      "learning_rate": 3.4375e-06,
+      "loss": 0.622,
+      "step": 44
+    },
+    {
+      "epoch": 0.21188934667451442,
+      "grad_norm": 2.004645347595215,
+      "learning_rate": 3.5156250000000003e-06,
+      "loss": 0.6254,
+      "step": 45
+    },
+    {
+      "epoch": 0.21659799882283695,
+      "grad_norm": 2.015970468521118,
+      "learning_rate": 3.59375e-06,
+      "loss": 0.6257,
+      "step": 46
+    },
+    {
+      "epoch": 0.2213066509711595,
+      "grad_norm": 1.8542283773422241,
+      "learning_rate": 3.6718750000000003e-06,
+      "loss": 0.5857,
+      "step": 47
+    },
+    {
+      "epoch": 0.22601530311948204,
+      "grad_norm": 1.8879379034042358,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.5901,
+      "step": 48
+    },
+    {
+      "epoch": 0.2307239552678046,
+      "grad_norm": 2.0544469356536865,
+      "learning_rate": 3.828125000000001e-06,
+      "loss": 0.6142,
+      "step": 49
+    },
+    {
+      "epoch": 0.23543260741612712,
+      "grad_norm": 1.9305391311645508,
+      "learning_rate": 3.90625e-06,
+      "loss": 0.5927,
+      "step": 50
+    },
+    {
+      "epoch": 0.24014125956444968,
+      "grad_norm": 2.0514976978302,
+      "learning_rate": 3.984375e-06,
+      "loss": 0.6225,
+      "step": 51
+    },
+    {
+      "epoch": 0.2448499117127722,
+      "grad_norm": 1.771240472793579,
+      "learning_rate": 4.0625000000000005e-06,
+      "loss": 0.5577,
+      "step": 52
+    },
+    {
+      "epoch": 0.24955856386109476,
+      "grad_norm": 1.7672276496887207,
+      "learning_rate": 4.140625000000001e-06,
+      "loss": 0.5625,
+      "step": 53
+    },
+    {
+      "epoch": 0.2542672160094173,
+      "grad_norm": 1.8797359466552734,
+      "learning_rate": 4.21875e-06,
+      "loss": 0.5839,
+      "step": 54
+    },
+    {
+      "epoch": 0.2589758681577399,
+      "grad_norm": 1.905333399772644,
+      "learning_rate": 4.296875e-06,
+      "loss": 0.5837,
+      "step": 55
+    },
+    {
+      "epoch": 0.2636845203060624,
+      "grad_norm": 2.143460273742676,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.6485,
+      "step": 56
+    },
+    {
+      "epoch": 0.26839317245438493,
+      "grad_norm": 1.9987819194793701,
+      "learning_rate": 4.453125000000001e-06,
+      "loss": 0.591,
+      "step": 57
+    },
+    {
+      "epoch": 0.2731018246027075,
+      "grad_norm": 2.0481247901916504,
+      "learning_rate": 4.53125e-06,
+      "loss": 0.5949,
+      "step": 58
+    },
+    {
+      "epoch": 0.27781047675103004,
+      "grad_norm": 1.8108389377593994,
+      "learning_rate": 4.609375e-06,
+      "loss": 0.5264,
+      "step": 59
+    },
+    {
+      "epoch": 0.28251912889935255,
+      "grad_norm": 1.840555191040039,
+      "learning_rate": 4.6875000000000004e-06,
+      "loss": 0.5586,
+      "step": 60
+    },
+    {
+      "epoch": 0.2872277810476751,
+      "grad_norm": 1.7868009805679321,
+      "learning_rate": 4.765625000000001e-06,
+      "loss": 0.5445,
+      "step": 61
+    },
+    {
+      "epoch": 0.29193643319599766,
+      "grad_norm": 1.8135056495666504,
+      "learning_rate": 4.84375e-06,
+      "loss": 0.5522,
+      "step": 62
+    },
+    {
+      "epoch": 0.2966450853443202,
+      "grad_norm": 1.8280186653137207,
+      "learning_rate": 4.921875e-06,
+      "loss": 0.5493,
+      "step": 63
+    },
+    {
+      "epoch": 0.3013537374926427,
+      "grad_norm": 1.9932650327682495,
+      "learning_rate": 5e-06,
+      "loss": 0.5866,
+      "step": 64
+    },
+    {
+      "epoch": 0.30606238964096527,
+      "grad_norm": 1.926773190498352,
+      "learning_rate": 4.9999915457391215e-06,
+      "loss": 0.5705,
+      "step": 65
+    },
+    {
+      "epoch": 0.3107710417892878,
+      "grad_norm": 1.8135457038879395,
+      "learning_rate": 4.999966183013663e-06,
+      "loss": 0.5475,
+      "step": 66
+    },
+    {
+      "epoch": 0.3154796939376104,
+      "grad_norm": 1.7810639142990112,
+      "learning_rate": 4.999923911995162e-06,
+      "loss": 0.5442,
+      "step": 67
+    },
+    {
+      "epoch": 0.3201883460859329,
+      "grad_norm": 1.7810635566711426,
+      "learning_rate": 4.999864732969518e-06,
+      "loss": 0.5468,
+      "step": 68
+    },
+    {
+      "epoch": 0.32489699823425544,
+      "grad_norm": 1.7482165098190308,
+      "learning_rate": 4.9997886463369815e-06,
+      "loss": 0.5292,
+      "step": 69
+    },
+    {
+      "epoch": 0.329605650382578,
+      "grad_norm": 1.8129892349243164,
+      "learning_rate": 4.999695652612156e-06,
+      "loss": 0.5645,
+      "step": 70
+    },
+    {
+      "epoch": 0.33431430253090055,
+      "grad_norm": 1.8132847547531128,
+      "learning_rate": 4.999585752423999e-06,
+      "loss": 0.5458,
+      "step": 71
+    },
+    {
+      "epoch": 0.33902295467922305,
+      "grad_norm": 1.8230698108673096,
+      "learning_rate": 4.999458946515808e-06,
+      "loss": 0.5455,
+      "step": 72
+    },
+    {
+      "epoch": 0.3437316068275456,
+      "grad_norm": 1.8087327480316162,
+      "learning_rate": 4.999315235745224e-06,
+      "loss": 0.5467,
+      "step": 73
+    },
+    {
+      "epoch": 0.34844025897586817,
+      "grad_norm": 1.8371257781982422,
+      "learning_rate": 4.999154621084221e-06,
+      "loss": 0.559,
+      "step": 74
+    },
+    {
+      "epoch": 0.3531489111241907,
+      "grad_norm": 1.7116756439208984,
+      "learning_rate": 4.998977103619103e-06,
+      "loss": 0.531,
+      "step": 75
+    },
+    {
+      "epoch": 0.3578575632725132,
+      "grad_norm": 1.8413268327713013,
+      "learning_rate": 4.9987826845504916e-06,
+      "loss": 0.5471,
+      "step": 76
+    },
+    {
+      "epoch": 0.3625662154208358,
+      "grad_norm": 1.7932018041610718,
+      "learning_rate": 4.9985713651933235e-06,
+      "loss": 0.5405,
+      "step": 77
+    },
+    {
+      "epoch": 0.36727486756915834,
+      "grad_norm": 1.739492654800415,
+      "learning_rate": 4.998343146976837e-06,
+      "loss": 0.5233,
+      "step": 78
+    },
+    {
+      "epoch": 0.3719835197174809,
+      "grad_norm": 1.6671689748764038,
+      "learning_rate": 4.998098031444567e-06,
+      "loss": 0.5129,
+      "step": 79
+    },
+    {
+      "epoch": 0.3766921718658034,
+      "grad_norm": 1.7220134735107422,
+      "learning_rate": 4.997836020254328e-06,
+      "loss": 0.5167,
+      "step": 80
+    },
+    {
+      "epoch": 0.38140082401412595,
+      "grad_norm": 1.743154764175415,
+      "learning_rate": 4.99755711517821e-06,
+      "loss": 0.5116,
+      "step": 81
+    },
+    {
+      "epoch": 0.3861094761624485,
+      "grad_norm": 1.5877171754837036,
+      "learning_rate": 4.9972613181025605e-06,
+      "loss": 0.5058,
+      "step": 82
+    },
+    {
+      "epoch": 0.39081812831077106,
+      "grad_norm": 1.6139500141143799,
+      "learning_rate": 4.996948631027978e-06,
+      "loss": 0.523,
+      "step": 83
+    },
+    {
+      "epoch": 0.39552678045909356,
+      "grad_norm": 1.6250971555709839,
+      "learning_rate": 4.996619056069292e-06,
+      "loss": 0.5277,
+      "step": 84
+    },
+    {
+      "epoch": 0.4002354326074161,
+      "grad_norm": 1.4910732507705688,
+      "learning_rate": 4.996272595455553e-06,
+      "loss": 0.4913,
+      "step": 85
+    },
+    {
+      "epoch": 0.4049440847557387,
+      "grad_norm": 1.590549349784851,
+      "learning_rate": 4.995909251530014e-06,
+      "loss": 0.5074,
+      "step": 86
+    },
+    {
+      "epoch": 0.40965273690406123,
+      "grad_norm": 1.728352665901184,
+      "learning_rate": 4.995529026750121e-06,
+      "loss": 0.5258,
+      "step": 87
+    },
+    {
+      "epoch": 0.41436138905238373,
+      "grad_norm": 1.5192519426345825,
+      "learning_rate": 4.995131923687488e-06,
+      "loss": 0.5165,
+      "step": 88
+    },
+    {
+      "epoch": 0.4190700412007063,
+      "grad_norm": 1.5756913423538208,
+      "learning_rate": 4.994717945027886e-06,
+      "loss": 0.5299,
+      "step": 89
+    },
+    {
+      "epoch": 0.42377869334902885,
+      "grad_norm": 1.5243970155715942,
+      "learning_rate": 4.9942870935712215e-06,
+      "loss": 0.5292,
+      "step": 90
+    },
+    {
+      "epoch": 0.4284873454973514,
+      "grad_norm": 1.583005428314209,
+      "learning_rate": 4.993839372231519e-06,
+      "loss": 0.5074,
+      "step": 91
+    },
+    {
+      "epoch": 0.4331959976456739,
+      "grad_norm": 1.5291098356246948,
+      "learning_rate": 4.993374784036902e-06,
+      "loss": 0.523,
+      "step": 92
+    },
+    {
+      "epoch": 0.43790464979399646,
+      "grad_norm": 1.4610439538955688,
+      "learning_rate": 4.992893332129568e-06,
+      "loss": 0.4949,
+      "step": 93
+    },
+    {
+      "epoch": 0.442613301942319,
+      "grad_norm": 1.5078333616256714,
+      "learning_rate": 4.992395019765776e-06,
+      "loss": 0.5043,
+      "step": 94
+    },
+    {
+      "epoch": 0.44732195409064157,
+      "grad_norm": 1.5027759075164795,
+      "learning_rate": 4.991879850315813e-06,
+      "loss": 0.5022,
+      "step": 95
+    },
+    {
+      "epoch": 0.45203060623896407,
+      "grad_norm": 1.3885493278503418,
+      "learning_rate": 4.991347827263983e-06,
+      "loss": 0.4761,
+      "step": 96
+    },
+    {
+      "epoch": 0.45673925838728663,
+      "grad_norm": 1.3093187808990479,
+      "learning_rate": 4.990798954208574e-06,
+      "loss": 0.4884,
+      "step": 97
+    },
+    {
+      "epoch": 0.4614479105356092,
+      "grad_norm": 1.391524314880371,
+      "learning_rate": 4.99023323486184e-06,
+      "loss": 0.5107,
+      "step": 98
+    },
+    {
+      "epoch": 0.46615656268393174,
+      "grad_norm": 1.3108298778533936,
+      "learning_rate": 4.989650673049971e-06,
+      "loss": 0.5018,
+      "step": 99
+    },
+    {
+      "epoch": 0.47086521483225424,
+      "grad_norm": 1.3510005474090576,
+      "learning_rate": 4.98905127271307e-06,
+      "loss": 0.5027,
+      "step": 100
+    },
+    {
+      "epoch": 0.4755738669805768,
+      "grad_norm": 1.1885933876037598,
+      "learning_rate": 4.988435037905129e-06,
+      "loss": 0.4626,
+      "step": 101
+    },
+    {
+      "epoch": 0.48028251912889935,
+      "grad_norm": 1.1308561563491821,
+      "learning_rate": 4.987801972793993e-06,
+      "loss": 0.4846,
+      "step": 102
+    },
+    {
+      "epoch": 0.4849911712772219,
+      "grad_norm": 1.134296178817749,
+      "learning_rate": 4.987152081661343e-06,
+      "loss": 0.5007,
+      "step": 103
+    },
+    {
+      "epoch": 0.4896998234255444,
+      "grad_norm": 1.1893833875656128,
+      "learning_rate": 4.986485368902656e-06,
+      "loss": 0.5055,
+      "step": 104
+    },
+    {
+      "epoch": 0.49440847557386697,
+      "grad_norm": 1.007875919342041,
+      "learning_rate": 4.985801839027183e-06,
+      "loss": 0.4775,
+      "step": 105
+    },
+    {
+      "epoch": 0.4991171277221895,
+      "grad_norm": 1.2207539081573486,
+      "learning_rate": 4.985101496657918e-06,
+      "loss": 0.5006,
+      "step": 106
+    },
+    {
+      "epoch": 0.503825779870512,
+      "grad_norm": 1.033278465270996,
+      "learning_rate": 4.9843843465315604e-06,
+      "loss": 0.4911,
+      "step": 107
+    },
+    {
+      "epoch": 0.5085344320188346,
+      "grad_norm": 1.0263997316360474,
+      "learning_rate": 4.98365039349849e-06,
+      "loss": 0.4772,
+      "step": 108
+    },
+    {
+      "epoch": 0.5132430841671571,
+      "grad_norm": 0.9356820583343506,
+      "learning_rate": 4.982899642522732e-06,
+      "loss": 0.4779,
+      "step": 109
+    },
+    {
+      "epoch": 0.5179517363154797,
+      "grad_norm": 0.8354411721229553,
+      "learning_rate": 4.982132098681923e-06,
+      "loss": 0.4846,
+      "step": 110
+    },
+    {
+      "epoch": 0.5226603884638023,
+      "grad_norm": 0.8493514657020569,
+      "learning_rate": 4.981347767167273e-06,
+      "loss": 0.5135,
+      "step": 111
+    },
+    {
+      "epoch": 0.5273690406121248,
+      "grad_norm": 0.980951726436615,
+      "learning_rate": 4.980546653283538e-06,
+      "loss": 0.4827,
+      "step": 112
+    },
+    {
+      "epoch": 0.5320776927604474,
+      "grad_norm": 0.6815721392631531,
+      "learning_rate": 4.979728762448979e-06,
+      "loss": 0.4508,
+      "step": 113
+    },
+    {
+      "epoch": 0.5367863449087699,
+      "grad_norm": 0.7471316456794739,
+      "learning_rate": 4.978894100195325e-06,
+      "loss": 0.5024,
+      "step": 114
+    },
+    {
+      "epoch": 0.5414949970570924,
+      "grad_norm": 0.6778037548065186,
+      "learning_rate": 4.978042672167739e-06,
+      "loss": 0.4927,
+      "step": 115
+    },
+    {
+      "epoch": 0.546203649205415,
+      "grad_norm": 0.8497841954231262,
+      "learning_rate": 4.977174484124776e-06,
+      "loss": 0.4908,
+      "step": 116
+    },
+    {
+      "epoch": 0.5509123013537375,
+      "grad_norm": 0.6059902310371399,
+      "learning_rate": 4.9762895419383475e-06,
+      "loss": 0.4703,
+      "step": 117
+    },
+    {
+      "epoch": 0.5556209535020601,
+      "grad_norm": 0.6304749250411987,
+      "learning_rate": 4.975387851593677e-06,
+      "loss": 0.5208,
+      "step": 118
+    },
+    {
+      "epoch": 0.5603296056503826,
+      "grad_norm": 0.6655656099319458,
+      "learning_rate": 4.9744694191892675e-06,
+      "loss": 0.5069,
+      "step": 119
+    },
+    {
+      "epoch": 0.5650382577987051,
+      "grad_norm": 0.6231979131698608,
+      "learning_rate": 4.973534250936851e-06,
+      "loss": 0.4722,
+      "step": 120
+    },
+    {
+      "epoch": 0.5697469099470277,
+      "grad_norm": 0.5612891912460327,
+      "learning_rate": 4.972582353161354e-06,
+      "loss": 0.4688,
+      "step": 121
+    },
+    {
+      "epoch": 0.5744555620953502,
+      "grad_norm": 0.4817934036254883,
+      "learning_rate": 4.9716137323008485e-06,
+      "loss": 0.457,
+      "step": 122
+    },
+    {
+      "epoch": 0.5791642142436727,
+      "grad_norm": 0.4845065176486969,
+      "learning_rate": 4.970628394906514e-06,
+      "loss": 0.4719,
+      "step": 123
+    },
+    {
+      "epoch": 0.5838728663919953,
+      "grad_norm": 0.5664409399032593,
+      "learning_rate": 4.96962634764259e-06,
+      "loss": 0.4985,
+      "step": 124
+    },
+    {
+      "epoch": 0.5885815185403178,
+      "grad_norm": 0.4690743386745453,
+      "learning_rate": 4.968607597286332e-06,
+      "loss": 0.4759,
+      "step": 125
+    },
+    {
+      "epoch": 0.5932901706886404,
+      "grad_norm": 0.4690413773059845,
+      "learning_rate": 4.967572150727965e-06,
+      "loss": 0.4837,
+      "step": 126
+    },
+    {
+      "epoch": 0.5979988228369629,
+      "grad_norm": 0.4350973069667816,
+      "learning_rate": 4.966520014970637e-06,
+      "loss": 0.4679,
+      "step": 127
+    },
+    {
+      "epoch": 0.6027074749852854,
+      "grad_norm": 0.4283631145954132,
+      "learning_rate": 4.965451197130373e-06,
+      "loss": 0.4495,
+      "step": 128
+    },
+    {
+      "epoch": 0.607416127133608,
+      "grad_norm": 0.5446098446846008,
+      "learning_rate": 4.964365704436023e-06,
+      "loss": 0.4868,
+      "step": 129
+    },
+    {
+      "epoch": 0.6121247792819305,
+      "grad_norm": 0.4670584797859192,
+      "learning_rate": 4.963263544229219e-06,
+      "loss": 0.4717,
+      "step": 130
+    },
+    {
+      "epoch": 0.616833431430253,
+      "grad_norm": 0.4510785639286041,
+      "learning_rate": 4.962144723964322e-06,
+      "loss": 0.4777,
+      "step": 131
+    },
+    {
+      "epoch": 0.6215420835785757,
+      "grad_norm": 0.42210468649864197,
+      "learning_rate": 4.961009251208368e-06,
+      "loss": 0.452,
+      "step": 132
+    },
+    {
+      "epoch": 0.6262507357268982,
+      "grad_norm": 0.42281368374824524,
+      "learning_rate": 4.959857133641025e-06,
+      "loss": 0.4547,
+      "step": 133
+    },
+    {
+      "epoch": 0.6309593878752208,
+      "grad_norm": 0.5155296325683594,
+      "learning_rate": 4.9586883790545345e-06,
+      "loss": 0.4736,
+      "step": 134
+    },
+    {
+      "epoch": 0.6356680400235433,
+      "grad_norm": 0.453692227602005,
+      "learning_rate": 4.957502995353663e-06,
+      "loss": 0.4761,
+      "step": 135
+    },
+    {
+      "epoch": 0.6403766921718658,
+      "grad_norm": 0.4068131744861603,
+      "learning_rate": 4.956300990555643e-06,
+      "loss": 0.4593,
+      "step": 136
+    },
+    {
+      "epoch": 0.6450853443201884,
+      "grad_norm": 0.43510329723358154,
+      "learning_rate": 4.955082372790124e-06,
+      "loss": 0.4581,
+      "step": 137
+    },
+    {
+      "epoch": 0.6497939964685109,
+      "grad_norm": 0.44176122546195984,
+      "learning_rate": 4.953847150299119e-06,
+      "loss": 0.4584,
+      "step": 138
+    },
+    {
+      "epoch": 0.6545026486168334,
+      "grad_norm": 0.47031915187835693,
+      "learning_rate": 4.952595331436939e-06,
+      "loss": 0.4646,
+      "step": 139
+    },
+    {
+      "epoch": 0.659211300765156,
+      "grad_norm": 0.4248315989971161,
+      "learning_rate": 4.951326924670148e-06,
+      "loss": 0.4636,
+      "step": 140
+    },
+    {
+      "epoch": 0.6639199529134785,
+      "grad_norm": 0.40988391637802124,
+      "learning_rate": 4.9500419385775e-06,
+      "loss": 0.4442,
+      "step": 141
+    },
+    {
+      "epoch": 0.6686286050618011,
+      "grad_norm": 0.471097856760025,
+      "learning_rate": 4.948740381849879e-06,
+      "loss": 0.4693,
+      "step": 142
+    },
+    {
+      "epoch": 0.6733372572101236,
+      "grad_norm": 0.46013182401657104,
+      "learning_rate": 4.947422263290249e-06,
+      "loss": 0.4683,
+      "step": 143
+    },
+    {
+      "epoch": 0.6780459093584461,
+      "grad_norm": 0.45335468649864197,
+      "learning_rate": 4.946087591813581e-06,
+      "loss": 0.4739,
+      "step": 144
+    },
+    {
+      "epoch": 0.6827545615067687,
+      "grad_norm": 0.4198642671108246,
+      "learning_rate": 4.944736376446804e-06,
+      "loss": 0.4516,
+      "step": 145
+    },
+    {
+      "epoch": 0.6874632136550912,
+      "grad_norm": 0.5052395462989807,
+      "learning_rate": 4.943368626328741e-06,
+      "loss": 0.465,
+      "step": 146
+    },
+    {
+      "epoch": 0.6921718658034137,
+      "grad_norm": 0.462447851896286,
+      "learning_rate": 4.941984350710045e-06,
+      "loss": 0.479,
+      "step": 147
+    },
+    {
+      "epoch": 0.6968805179517363,
+      "grad_norm": 0.4185675084590912,
+      "learning_rate": 4.940583558953138e-06,
+      "loss": 0.4458,
+      "step": 148
+    },
+    {
+      "epoch": 0.7015891701000588,
+      "grad_norm": 0.4506431519985199,
+      "learning_rate": 4.939166260532145e-06,
+      "loss": 0.4866,
+      "step": 149
+    },
+    {
+      "epoch": 0.7062978222483814,
+      "grad_norm": 0.41306138038635254,
+      "learning_rate": 4.937732465032838e-06,
+      "loss": 0.4834,
+      "step": 150
+    },
+    {
+      "epoch": 0.711006474396704,
+      "grad_norm": 0.42707082629203796,
+      "learning_rate": 4.936282182152559e-06,
+      "loss": 0.4582,
+      "step": 151
+    },
+    {
+      "epoch": 0.7157151265450264,
+      "grad_norm": 0.5058028101921082,
+      "learning_rate": 4.934815421700166e-06,
+      "loss": 0.4883,
+      "step": 152
+    },
+    {
+      "epoch": 0.7204237786933491,
+      "grad_norm": 0.476229727268219,
+      "learning_rate": 4.933332193595956e-06,
+      "loss": 0.4913,
+      "step": 153
+    },
+    {
+      "epoch": 0.7251324308416716,
+      "grad_norm": 0.40548497438430786,
+      "learning_rate": 4.931832507871611e-06,
+      "loss": 0.4497,
+      "step": 154
+    },
+    {
+      "epoch": 0.7298410829899941,
+      "grad_norm": 0.401818186044693,
+      "learning_rate": 4.930316374670115e-06,
+      "loss": 0.4444,
+      "step": 155
+    },
+    {
+      "epoch": 0.7345497351383167,
+      "grad_norm": 0.3804225027561188,
+      "learning_rate": 4.9287838042457e-06,
+      "loss": 0.4499,
+      "step": 156
+    },
+    {
+      "epoch": 0.7392583872866392,
+      "grad_norm": 0.5111300349235535,
+      "learning_rate": 4.927234806963763e-06,
+      "loss": 0.4635,
+      "step": 157
+    },
+    {
+      "epoch": 0.7439670394349618,
+      "grad_norm": 0.4021705389022827,
+      "learning_rate": 4.925669393300808e-06,
+      "loss": 0.4597,
+      "step": 158
+    },
+    {
+      "epoch": 0.7486756915832843,
+      "grad_norm": 0.39262616634368896,
+      "learning_rate": 4.924087573844365e-06,
+      "loss": 0.4607,
+      "step": 159
+    },
+    {
+      "epoch": 0.7533843437316068,
+      "grad_norm": 0.3751242160797119,
+      "learning_rate": 4.922489359292928e-06,
+      "loss": 0.4376,
+      "step": 160
+    },
+    {
+      "epoch": 0.7580929958799294,
+      "grad_norm": 0.880349338054657,
+      "learning_rate": 4.920874760455874e-06,
+      "loss": 0.4372,
+      "step": 161
+    },
+    {
+      "epoch": 0.7628016480282519,
+      "grad_norm": 0.4332122206687927,
+      "learning_rate": 4.919243788253394e-06,
+      "loss": 0.4577,
+      "step": 162
+    },
+    {
+      "epoch": 0.7675103001765744,
+      "grad_norm": 0.40380653738975525,
+      "learning_rate": 4.917596453716421e-06,
+      "loss": 0.4561,
+      "step": 163
+    },
+    {
+      "epoch": 0.772218952324897,
+      "grad_norm": 0.40455198287963867,
+      "learning_rate": 4.915932767986552e-06,
+      "loss": 0.4469,
+      "step": 164
+    },
+    {
+      "epoch": 0.7769276044732195,
+      "grad_norm": 0.3832710087299347,
+      "learning_rate": 4.914252742315972e-06,
+      "loss": 0.4463,
+      "step": 165
+    },
+    {
+      "epoch": 0.7816362566215421,
+      "grad_norm": 0.46339336037635803,
+      "learning_rate": 4.9125563880673814e-06,
+      "loss": 0.4591,
+      "step": 166
+    },
+    {
+      "epoch": 0.7863449087698646,
+      "grad_norm": 0.4066482186317444,
+      "learning_rate": 4.9108437167139186e-06,
+      "loss": 0.447,
+      "step": 167
+    },
+    {
+      "epoch": 0.7910535609181871,
+      "grad_norm": 0.38544756174087524,
+      "learning_rate": 4.90911473983908e-06,
+      "loss": 0.4556,
+      "step": 168
+    },
+    {
+      "epoch": 0.7957622130665097,
+      "grad_norm": 0.39471447467803955,
+      "learning_rate": 4.90736946913664e-06,
+      "loss": 0.443,
+      "step": 169
+    },
+    {
+      "epoch": 0.8004708652148322,
+      "grad_norm": 0.42424091696739197,
+      "learning_rate": 4.905607916410581e-06,
+      "loss": 0.4561,
+      "step": 170
+    },
+    {
+      "epoch": 0.8051795173631549,
+      "grad_norm": 0.4095331132411957,
+      "learning_rate": 4.9038300935750026e-06,
+      "loss": 0.4622,
+      "step": 171
+    },
+    {
+      "epoch": 0.8098881695114774,
+      "grad_norm": 0.41514596343040466,
+      "learning_rate": 4.902036012654048e-06,
+      "loss": 0.4547,
+      "step": 172
+    },
+    {
+      "epoch": 0.8145968216597999,
+      "grad_norm": 0.4074968099594116,
+      "learning_rate": 4.900225685781819e-06,
+      "loss": 0.4555,
+      "step": 173
+    },
+    {
+      "epoch": 0.8193054738081225,
+      "grad_norm": 0.40082669258117676,
+      "learning_rate": 4.898399125202296e-06,
+      "loss": 0.4473,
+      "step": 174
+    },
+    {
+      "epoch": 0.824014125956445,
+      "grad_norm": 0.49021676182746887,
+      "learning_rate": 4.896556343269254e-06,
+      "loss": 0.4404,
+      "step": 175
+    },
+    {
+      "epoch": 0.8287227781047675,
+      "grad_norm": 0.3969082236289978,
+      "learning_rate": 4.894697352446182e-06,
+      "loss": 0.4535,
+      "step": 176
+    },
+    {
+      "epoch": 0.8334314302530901,
+      "grad_norm": 0.39450210332870483,
+      "learning_rate": 4.892822165306195e-06,
+      "loss": 0.4313,
+      "step": 177
+    },
+    {
+      "epoch": 0.8381400824014126,
+      "grad_norm": 0.41130906343460083,
+      "learning_rate": 4.8909307945319474e-06,
+      "loss": 0.4725,
+      "step": 178
+    },
+    {
+      "epoch": 0.8428487345497352,
+      "grad_norm": 0.3974241614341736,
+      "learning_rate": 4.8890232529155556e-06,
+      "loss": 0.4665,
+      "step": 179
+    },
+    {
+      "epoch": 0.8475573866980577,
+      "grad_norm": 0.46764135360717773,
+      "learning_rate": 4.887099553358502e-06,
+      "loss": 0.4699,
+      "step": 180
+    },
+    {
+      "epoch": 0.8522660388463802,
+      "grad_norm": 0.41044411063194275,
+      "learning_rate": 4.885159708871552e-06,
+      "loss": 0.4685,
+      "step": 181
+    },
+    {
+      "epoch": 0.8569746909947028,
+      "grad_norm": 0.4245174825191498,
+      "learning_rate": 4.883203732574668e-06,
+      "loss": 0.4465,
+      "step": 182
+    },
+    {
+      "epoch": 0.8616833431430253,
+      "grad_norm": 0.41129034757614136,
+      "learning_rate": 4.881231637696918e-06,
+      "loss": 0.4824,
+      "step": 183
+    },
+    {
+      "epoch": 0.8663919952913478,
+      "grad_norm": 0.4054642617702484,
+      "learning_rate": 4.879243437576383e-06,
+      "loss": 0.4318,
+      "step": 184
+    },
+    {
+      "epoch": 0.8711006474396704,
+      "grad_norm": 0.38721156120300293,
+      "learning_rate": 4.877239145660075e-06,
+      "loss": 0.4555,
+      "step": 185
+    },
+    {
+      "epoch": 0.8758092995879929,
+      "grad_norm": 0.41428545117378235,
+      "learning_rate": 4.8752187755038384e-06,
+      "loss": 0.4505,
+      "step": 186
+    },
+    {
+      "epoch": 0.8805179517363155,
+      "grad_norm": 0.42160969972610474,
+      "learning_rate": 4.873182340772263e-06,
+      "loss": 0.4424,
+      "step": 187
+    },
+    {
+      "epoch": 0.885226603884638,
+      "grad_norm": 0.38617607951164246,
+      "learning_rate": 4.871129855238589e-06,
+      "loss": 0.4475,
+      "step": 188
+    },
+    {
+      "epoch": 0.8899352560329605,
+      "grad_norm": 0.398945152759552,
+      "learning_rate": 4.869061332784614e-06,
+      "loss": 0.4386,
+      "step": 189
+    },
+    {
+      "epoch": 0.8946439081812831,
+      "grad_norm": 0.3810630738735199,
+      "learning_rate": 4.866976787400602e-06,
+      "loss": 0.4471,
+      "step": 190
+    },
+    {
+      "epoch": 0.8993525603296056,
+      "grad_norm": 0.391017347574234,
+      "learning_rate": 4.864876233185185e-06,
+      "loss": 0.4541,
+      "step": 191
+    },
+    {
+      "epoch": 0.9040612124779281,
+      "grad_norm": 0.4350132644176483,
+      "learning_rate": 4.862759684345269e-06,
+      "loss": 0.4685,
+      "step": 192
+    },
+    {
+      "epoch": 0.9087698646262508,
+      "grad_norm": 0.3881378769874573,
+      "learning_rate": 4.860627155195941e-06,
+      "loss": 0.4369,
+      "step": 193
+    },
+    {
+      "epoch": 0.9134785167745733,
+      "grad_norm": 0.42739424109458923,
+      "learning_rate": 4.858478660160364e-06,
+      "loss": 0.4669,
+      "step": 194
+    },
+    {
+      "epoch": 0.9181871689228959,
+      "grad_norm": 0.40032345056533813,
+      "learning_rate": 4.8563142137696906e-06,
+      "loss": 0.4769,
+      "step": 195
+    },
+    {
+      "epoch": 0.9228958210712184,
+      "grad_norm": 0.4069063067436218,
+      "learning_rate": 4.854133830662955e-06,
+      "loss": 0.448,
+      "step": 196
+    },
+    {
+      "epoch": 0.9276044732195409,
+      "grad_norm": 0.38561198115348816,
+      "learning_rate": 4.85193752558698e-06,
+      "loss": 0.4421,
+      "step": 197
+    },
+    {
+      "epoch": 0.9323131253678635,
+      "grad_norm": 0.44492074847221375,
+      "learning_rate": 4.849725313396274e-06,
+      "loss": 0.4647,
+      "step": 198
+    },
+    {
+      "epoch": 0.937021777516186,
+      "grad_norm": 0.3882367014884949,
+      "learning_rate": 4.847497209052933e-06,
+      "loss": 0.4327,
+      "step": 199
+    },
+    {
+      "epoch": 0.9417304296645085,
+      "grad_norm": 0.38339805603027344,
+      "learning_rate": 4.8452532276265366e-06,
+      "loss": 0.4464,
+      "step": 200
+    },
+    {
+      "epoch": 0.9464390818128311,
+      "grad_norm": 0.3912663161754608,
+      "learning_rate": 4.842993384294047e-06,
+      "loss": 0.4656,
+      "step": 201
+    },
+    {
+      "epoch": 0.9511477339611536,
+      "grad_norm": 0.38164153695106506,
+      "learning_rate": 4.84071769433971e-06,
+      "loss": 0.4683,
+      "step": 202
+    },
+    {
+      "epoch": 0.9558563861094762,
+      "grad_norm": 0.40451687574386597,
+      "learning_rate": 4.838426173154948e-06,
+      "loss": 0.455,
+      "step": 203
+    },
+    {
+      "epoch": 0.9605650382577987,
+      "grad_norm": 0.37911689281463623,
+      "learning_rate": 4.836118836238253e-06,
+      "loss": 0.4493,
+      "step": 204
+    },
+    {
+      "epoch": 0.9652736904061212,
+      "grad_norm": 0.3586491644382477,
+      "learning_rate": 4.833795699195089e-06,
+      "loss": 0.4293,
+      "step": 205
+    },
+    {
+      "epoch": 0.9699823425544438,
+      "grad_norm": 0.3737817406654358,
+      "learning_rate": 4.83145677773778e-06,
+      "loss": 0.4433,
+      "step": 206
+    },
+    {
+      "epoch": 0.9746909947027663,
+      "grad_norm": 0.40282636880874634,
+      "learning_rate": 4.829102087685408e-06,
+      "loss": 0.4528,
+      "step": 207
+    },
+    {
+      "epoch": 0.9793996468510888,
+      "grad_norm": 0.41744890809059143,
+      "learning_rate": 4.826731644963705e-06,
+      "loss": 0.4591,
+      "step": 208
+    },
+    {
+      "epoch": 0.9841082989994114,
+      "grad_norm": 0.39800336956977844,
+      "learning_rate": 4.824345465604944e-06,
+      "loss": 0.4433,
+      "step": 209
+    },
+    {
+      "epoch": 0.9888169511477339,
+      "grad_norm": 0.41930362582206726,
+      "learning_rate": 4.821943565747831e-06,
+      "loss": 0.4362,
+      "step": 210
+    },
+    {
+      "epoch": 0.9935256032960565,
+      "grad_norm": 0.4097538888454437,
+      "learning_rate": 4.8195259616373955e-06,
+      "loss": 0.4528,
+      "step": 211
+    },
+    {
+      "epoch": 0.998234255444379,
+      "grad_norm": 0.3836863040924072,
+      "learning_rate": 4.817092669624883e-06,
+      "loss": 0.4621,
+      "step": 212
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3836863040924072,
+      "learning_rate": 4.814643706167642e-06,
+      "loss": 0.4892,
+      "step": 213
+    },
+    {
+      "epoch": 1.0047086521483226,
+      "grad_norm": 0.7609542012214661,
+      "learning_rate": 4.8121790878290135e-06,
+      "loss": 0.4259,
+      "step": 214
+    },
+    {
+      "epoch": 1.009417304296645,
+      "grad_norm": 0.3908381462097168,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 0.4192,
+      "step": 215
+    },
+    {
+      "epoch": 1.0141259564449676,
+      "grad_norm": 0.43996715545654297,
+      "learning_rate": 4.807202953290243e-06,
+      "loss": 0.4103,
+      "step": 216
+    },
+    {
+      "epoch": 1.0188346085932902,
+      "grad_norm": 0.41817402839660645,
+      "learning_rate": 4.804691470745734e-06,
+      "loss": 0.4432,
+      "step": 217
+    },
+    {
+      "epoch": 1.0235432607416126,
+      "grad_norm": 0.3807670772075653,
+      "learning_rate": 4.8021644006308735e-06,
+      "loss": 0.3983,
+      "step": 218
+    },
+    {
+      "epoch": 1.0282519128899352,
+      "grad_norm": 0.390556275844574,
+      "learning_rate": 4.799621760037268e-06,
+      "loss": 0.4002,
+      "step": 219
+    },
+    {
+      "epoch": 1.0329605650382578,
+      "grad_norm": 0.4105827212333679,
+      "learning_rate": 4.7970635661618345e-06,
+      "loss": 0.4409,
+      "step": 220
+    },
+    {
+      "epoch": 1.0376692171865805,
+      "grad_norm": 0.42910847067832947,
+      "learning_rate": 4.7944898363066865e-06,
+      "loss": 0.4316,
+      "step": 221
+    },
+    {
+      "epoch": 1.0423778693349028,
+      "grad_norm": 0.42775776982307434,
+      "learning_rate": 4.79190058787901e-06,
+      "loss": 0.4248,
+      "step": 222
+    },
+    {
+      "epoch": 1.0470865214832255,
+      "grad_norm": 0.5438376069068909,
+      "learning_rate": 4.789295838390947e-06,
+      "loss": 0.4301,
+      "step": 223
+    },
+    {
+      "epoch": 1.051795173631548,
+      "grad_norm": 0.403005450963974,
+      "learning_rate": 4.786675605459488e-06,
+      "loss": 0.4118,
+      "step": 224
+    },
+    {
+      "epoch": 1.0565038257798705,
+      "grad_norm": 0.3933987617492676,
+      "learning_rate": 4.7840399068063355e-06,
+      "loss": 0.4146,
+      "step": 225
+    },
+    {
+      "epoch": 1.061212477928193,
+      "grad_norm": 0.3724134862422943,
+      "learning_rate": 4.781388760257799e-06,
+      "loss": 0.4186,
+      "step": 226
+    },
+    {
+      "epoch": 1.0659211300765157,
+      "grad_norm": 0.3519379794597626,
+      "learning_rate": 4.778722183744665e-06,
+      "loss": 0.3863,
+      "step": 227
+    },
+    {
+      "epoch": 1.070629782224838,
+      "grad_norm": 0.4137888252735138,
+      "learning_rate": 4.77604019530208e-06,
+      "loss": 0.4398,
+      "step": 228
+    },
+    {
+      "epoch": 1.0753384343731607,
+      "grad_norm": 0.3873429596424103,
+      "learning_rate": 4.773342813069428e-06,
+      "loss": 0.4332,
+      "step": 229
+    },
+    {
+      "epoch": 1.0800470865214833,
+      "grad_norm": 0.38474398851394653,
+      "learning_rate": 4.770630055290209e-06,
+      "loss": 0.4318,
+      "step": 230
+    },
+    {
+      "epoch": 1.0847557386698057,
+      "grad_norm": 0.4013967514038086,
+      "learning_rate": 4.767901940311911e-06,
+      "loss": 0.4429,
+      "step": 231
+    },
+    {
+      "epoch": 1.0894643908181283,
+      "grad_norm": 0.3859746754169464,
+      "learning_rate": 4.76515848658589e-06,
+      "loss": 0.4095,
+      "step": 232
+    },
+    {
+      "epoch": 1.094173042966451,
+      "grad_norm": 0.36502712965011597,
+      "learning_rate": 4.762399712667247e-06,
+      "loss": 0.4093,
+      "step": 233
+    },
+    {
+      "epoch": 1.0988816951147733,
+      "grad_norm": 0.3925178349018097,
+      "learning_rate": 4.7596256372146964e-06,
+      "loss": 0.4176,
+      "step": 234
+    },
+    {
+      "epoch": 1.103590347263096,
+      "grad_norm": 0.3929302990436554,
+      "learning_rate": 4.7568362789904445e-06,
+      "loss": 0.4312,
+      "step": 235
+    },
+    {
+      "epoch": 1.1082989994114185,
+      "grad_norm": 0.3633573055267334,
+      "learning_rate": 4.75403165686006e-06,
+      "loss": 0.4449,
+      "step": 236
+    },
+    {
+      "epoch": 1.113007651559741,
+      "grad_norm": 0.4058593809604645,
+      "learning_rate": 4.751211789792349e-06,
+      "loss": 0.4242,
+      "step": 237
+    },
+    {
+      "epoch": 1.1177163037080635,
+      "grad_norm": 0.37989768385887146,
+      "learning_rate": 4.748376696859226e-06,
+      "loss": 0.4035,
+      "step": 238
+    },
+    {
+      "epoch": 1.1224249558563861,
+      "grad_norm": 0.46411606669425964,
+      "learning_rate": 4.745526397235583e-06,
+      "loss": 0.388,
+      "step": 239
+    },
+    {
+      "epoch": 1.1271336080047087,
+      "grad_norm": 0.3705718219280243,
+      "learning_rate": 4.7426609101991605e-06,
+      "loss": 0.3974,
+      "step": 240
+    },
+    {
+      "epoch": 1.1318422601530311,
+      "grad_norm": 0.3925943374633789,
+      "learning_rate": 4.7397802551304185e-06,
+      "loss": 0.4229,
+      "step": 241
+    },
+    {
+      "epoch": 1.1365509123013537,
+      "grad_norm": 0.37178972363471985,
+      "learning_rate": 4.736884451512405e-06,
+      "loss": 0.4055,
+      "step": 242
+    },
+    {
+      "epoch": 1.1412595644496764,
+      "grad_norm": 0.39106276631355286,
+      "learning_rate": 4.733973518930624e-06,
+      "loss": 0.4418,
+      "step": 243
+    },
+    {
+      "epoch": 1.1459682165979987,
+      "grad_norm": 0.3852050006389618,
+      "learning_rate": 4.7310474770729e-06,
+      "loss": 0.4066,
+      "step": 244
+    },
+    {
+      "epoch": 1.1506768687463214,
+      "grad_norm": 0.37759003043174744,
+      "learning_rate": 4.728106345729253e-06,
+      "loss": 0.3907,
+      "step": 245
+    },
+    {
+      "epoch": 1.155385520894644,
+      "grad_norm": 0.43599414825439453,
+      "learning_rate": 4.7251501447917535e-06,
+      "loss": 0.4632,
+      "step": 246
+    },
+    {
+      "epoch": 1.1600941730429664,
+      "grad_norm": 0.38698461651802063,
+      "learning_rate": 4.722178894254399e-06,
+      "loss": 0.4031,
+      "step": 247
+    },
+    {
+      "epoch": 1.164802825191289,
+      "grad_norm": 0.4270411729812622,
+      "learning_rate": 4.71919261421297e-06,
+      "loss": 0.4251,
+      "step": 248
+    },
+    {
+      "epoch": 1.1695114773396116,
+      "grad_norm": 0.4065026640892029,
+      "learning_rate": 4.716191324864899e-06,
+      "loss": 0.4295,
+      "step": 249
+    },
+    {
+      "epoch": 1.1742201294879342,
+      "grad_norm": 0.38610973954200745,
+      "learning_rate": 4.7131750465091315e-06,
+      "loss": 0.3991,
+      "step": 250
+    },
+    {
+      "epoch": 1.1789287816362566,
+      "grad_norm": 0.41599562764167786,
+      "learning_rate": 4.710143799545994e-06,
+      "loss": 0.4278,
+      "step": 251
+    },
+    {
+      "epoch": 1.1836374337845792,
+      "grad_norm": 0.429103821516037,
+      "learning_rate": 4.707097604477045e-06,
+      "loss": 0.406,
+      "step": 252
+    },
+    {
+      "epoch": 1.1883460859329018,
+      "grad_norm": 0.4257577657699585,
+      "learning_rate": 4.704036481904949e-06,
+      "loss": 0.4471,
+      "step": 253
+    },
+    {
+      "epoch": 1.1930547380812242,
+      "grad_norm": 0.38792288303375244,
+      "learning_rate": 4.700960452533329e-06,
+      "loss": 0.4114,
+      "step": 254
+    },
+    {
+      "epoch": 1.1977633902295468,
+      "grad_norm": 0.38506487011909485,
+      "learning_rate": 4.697869537166626e-06,
+      "loss": 0.4106,
+      "step": 255
+    },
+    {
+      "epoch": 1.2024720423778694,
+      "grad_norm": 0.42475154995918274,
+      "learning_rate": 4.6947637567099675e-06,
+      "loss": 0.4208,
+      "step": 256
+    },
+    {
+      "epoch": 1.2071806945261918,
+      "grad_norm": 0.3511320650577545,
+      "learning_rate": 4.691643132169014e-06,
+      "loss": 0.4231,
+      "step": 257
+    },
+    {
+      "epoch": 1.2118893466745144,
+      "grad_norm": 0.38943809270858765,
+      "learning_rate": 4.6885076846498255e-06,
+      "loss": 0.4013,
+      "step": 258
+    },
+    {
+      "epoch": 1.216597998822837,
+      "grad_norm": 0.3698880672454834,
+      "learning_rate": 4.685357435358715e-06,
+      "loss": 0.4055,
+      "step": 259
+    },
+    {
+      "epoch": 1.2213066509711594,
+      "grad_norm": 0.3700575828552246,
+      "learning_rate": 4.682192405602106e-06,
+      "loss": 0.4142,
+      "step": 260
+    },
+    {
+      "epoch": 1.226015303119482,
+      "grad_norm": 0.40088504552841187,
+      "learning_rate": 4.6790126167863875e-06,
+      "loss": 0.4258,
+      "step": 261
+    },
+    {
+      "epoch": 1.2307239552678046,
+      "grad_norm": 0.4097707271575928,
+      "learning_rate": 4.675818090417772e-06,
+      "loss": 0.4199,
+      "step": 262
+    },
+    {
+      "epoch": 1.235432607416127,
+      "grad_norm": 0.4138272702693939,
+      "learning_rate": 4.672608848102146e-06,
+      "loss": 0.4601,
+      "step": 263
+    },
+    {
+      "epoch": 1.2401412595644496,
+      "grad_norm": 0.40342196822166443,
+      "learning_rate": 4.6693849115449265e-06,
+      "loss": 0.3962,
+      "step": 264
+    },
+    {
+      "epoch": 1.2448499117127723,
+      "grad_norm": 0.39221933484077454,
+      "learning_rate": 4.666146302550916e-06,
+      "loss": 0.4273,
+      "step": 265
+    },
+    {
+      "epoch": 1.2495585638610947,
+      "grad_norm": 0.42961645126342773,
+      "learning_rate": 4.66289304302415e-06,
+      "loss": 0.3995,
+      "step": 266
+    },
+    {
+      "epoch": 1.2542672160094173,
+      "grad_norm": 0.3756173849105835,
+      "learning_rate": 4.659625154967751e-06,
+      "loss": 0.4143,
+      "step": 267
+    },
+    {
+      "epoch": 1.2589758681577399,
+      "grad_norm": 0.3766971230506897,
+      "learning_rate": 4.656342660483783e-06,
+      "loss": 0.4316,
+      "step": 268
+    },
+    {
+      "epoch": 1.2636845203060623,
+      "grad_norm": 0.3824644982814789,
+      "learning_rate": 4.6530455817730965e-06,
+      "loss": 0.4078,
+      "step": 269
+    },
+    {
+      "epoch": 1.2683931724543849,
+      "grad_norm": 0.4655967056751251,
+      "learning_rate": 4.649733941135183e-06,
+      "loss": 0.4123,
+      "step": 270
+    },
+    {
+      "epoch": 1.2731018246027075,
+      "grad_norm": 0.42232146859169006,
+      "learning_rate": 4.646407760968023e-06,
+      "loss": 0.4197,
+      "step": 271
+    },
+    {
+      "epoch": 1.27781047675103,
+      "grad_norm": 0.39129775762557983,
+      "learning_rate": 4.6430670637679295e-06,
+      "loss": 0.4297,
+      "step": 272
+    },
+    {
+      "epoch": 1.2825191288993525,
+      "grad_norm": 0.41317105293273926,
+      "learning_rate": 4.639711872129405e-06,
+      "loss": 0.4479,
+      "step": 273
+    },
+    {
+      "epoch": 1.287227781047675,
+      "grad_norm": 0.38349881768226624,
+      "learning_rate": 4.636342208744981e-06,
+      "loss": 0.4113,
+      "step": 274
+    },
+    {
+      "epoch": 1.2919364331959977,
+      "grad_norm": 0.43842265009880066,
+      "learning_rate": 4.632958096405071e-06,
+      "loss": 0.4299,
+      "step": 275
+    },
+    {
+      "epoch": 1.2966450853443203,
+      "grad_norm": 0.37726807594299316,
+      "learning_rate": 4.629559557997805e-06,
+      "loss": 0.4099,
+      "step": 276
+    },
+    {
+      "epoch": 1.3013537374926427,
+      "grad_norm": 0.3869900703430176,
+      "learning_rate": 4.62614661650889e-06,
+      "loss": 0.4244,
+      "step": 277
+    },
+    {
+      "epoch": 1.3060623896409653,
+      "grad_norm": 0.37169286608695984,
+      "learning_rate": 4.6227192950214435e-06,
+      "loss": 0.4005,
+      "step": 278
+    },
+    {
+      "epoch": 1.310771041789288,
+      "grad_norm": 0.3810769021511078,
+      "learning_rate": 4.619277616715843e-06,
+      "loss": 0.4334,
+      "step": 279
+    },
+    {
+      "epoch": 1.3154796939376103,
+      "grad_norm": 0.3881538510322571,
+      "learning_rate": 4.615821604869564e-06,
+      "loss": 0.4163,
+      "step": 280
+    },
+    {
+      "epoch": 1.320188346085933,
+      "grad_norm": 0.37338587641716003,
+      "learning_rate": 4.612351282857027e-06,
+      "loss": 0.4233,
+      "step": 281
+    },
+    {
+      "epoch": 1.3248969982342556,
+      "grad_norm": 0.3771686851978302,
+      "learning_rate": 4.6088666741494395e-06,
+      "loss": 0.4024,
+      "step": 282
+    },
+    {
+      "epoch": 1.329605650382578,
+      "grad_norm": 0.3833746910095215,
+      "learning_rate": 4.605367802314632e-06,
+      "loss": 0.3932,
+      "step": 283
+    },
+    {
+      "epoch": 1.3343143025309006,
+      "grad_norm": 0.37729066610336304,
+      "learning_rate": 4.601854691016907e-06,
+      "loss": 0.4022,
+      "step": 284
+    },
+    {
+      "epoch": 1.3390229546792232,
+      "grad_norm": 0.3654945492744446,
+      "learning_rate": 4.598327364016871e-06,
+      "loss": 0.4089,
+      "step": 285
+    },
+    {
+      "epoch": 1.3437316068275456,
+      "grad_norm": 0.39260923862457275,
+      "learning_rate": 4.5947858451712775e-06,
+      "loss": 0.4286,
+      "step": 286
+    },
+    {
+      "epoch": 1.3484402589758682,
+      "grad_norm": 0.4227854609489441,
+      "learning_rate": 4.591230158432867e-06,
+      "loss": 0.4287,
+      "step": 287
+    },
+    {
+      "epoch": 1.3531489111241908,
+      "grad_norm": 0.4016524851322174,
+      "learning_rate": 4.587660327850203e-06,
+      "loss": 0.4132,
+      "step": 288
+    },
+    {
+      "epoch": 1.3578575632725132,
+      "grad_norm": 0.3761261999607086,
+      "learning_rate": 4.584076377567506e-06,
+      "loss": 0.426,
+      "step": 289
+    },
+    {
+      "epoch": 1.3625662154208358,
+      "grad_norm": 0.3990737795829773,
+      "learning_rate": 4.580478331824499e-06,
+      "loss": 0.4224,
+      "step": 290
+    },
+    {
+      "epoch": 1.3672748675691584,
+      "grad_norm": 0.3864597976207733,
+      "learning_rate": 4.576866214956235e-06,
+      "loss": 0.4021,
+      "step": 291
+    },
+    {
+      "epoch": 1.3719835197174808,
+      "grad_norm": 0.39642560482025146,
+      "learning_rate": 4.5732400513929354e-06,
+      "loss": 0.4373,
+      "step": 292
+    },
+    {
+      "epoch": 1.3766921718658034,
+      "grad_norm": 0.3693861663341522,
+      "learning_rate": 4.569599865659829e-06,
+      "loss": 0.4239,
+      "step": 293
+    },
+    {
+      "epoch": 1.381400824014126,
+      "grad_norm": 0.374732106924057,
+      "learning_rate": 4.565945682376978e-06,
+      "loss": 0.4138,
+      "step": 294
+    },
+    {
+      "epoch": 1.3861094761624484,
+      "grad_norm": 0.3904346227645874,
+      "learning_rate": 4.562277526259117e-06,
+      "loss": 0.4371,
+      "step": 295
+    },
+    {
+      "epoch": 1.390818128310771,
+      "grad_norm": 0.42636239528656006,
+      "learning_rate": 4.558595422115486e-06,
+      "loss": 0.4326,
+      "step": 296
+    },
+    {
+      "epoch": 1.3955267804590936,
+      "grad_norm": 0.4012644290924072,
+      "learning_rate": 4.55489939484966e-06,
+      "loss": 0.417,
+      "step": 297
+    },
+    {
+      "epoch": 1.400235432607416,
+      "grad_norm": 0.3714677393436432,
+      "learning_rate": 4.551189469459382e-06,
+      "loss": 0.4112,
+      "step": 298
+    },
+    {
+      "epoch": 1.4049440847557386,
+      "grad_norm": 0.36800074577331543,
+      "learning_rate": 4.547465671036394e-06,
+      "loss": 0.4136,
+      "step": 299
+    },
+    {
+      "epoch": 1.4096527369040612,
+      "grad_norm": 0.42034563422203064,
+      "learning_rate": 4.543728024766265e-06,
+      "loss": 0.4389,
+      "step": 300
+    },
+    {
+      "epoch": 1.4143613890523836,
+      "grad_norm": 0.37657004594802856,
+      "learning_rate": 4.539976555928227e-06,
+      "loss": 0.4241,
+      "step": 301
+    },
+    {
+      "epoch": 1.4190700412007062,
+      "grad_norm": 0.3639003336429596,
+      "learning_rate": 4.536211289894995e-06,
+      "loss": 0.4055,
+      "step": 302
+    },
+    {
+      "epoch": 1.4237786933490288,
+      "grad_norm": 0.3953859210014343,
+      "learning_rate": 4.532432252132603e-06,
+      "loss": 0.4317,
+      "step": 303
+    },
+    {
+      "epoch": 1.4284873454973515,
+      "grad_norm": 0.3482641279697418,
+      "learning_rate": 4.5286394682002265e-06,
+      "loss": 0.4023,
+      "step": 304
+    },
+    {
+      "epoch": 1.4331959976456738,
+      "grad_norm": 0.41340556740760803,
+      "learning_rate": 4.524832963750015e-06,
+      "loss": 0.4301,
+      "step": 305
+    },
+    {
+      "epoch": 1.4379046497939965,
+      "grad_norm": 0.39513641595840454,
+      "learning_rate": 4.521012764526913e-06,
+      "loss": 0.4083,
+      "step": 306
+    },
+    {
+      "epoch": 1.442613301942319,
+      "grad_norm": 0.3495503067970276,
+      "learning_rate": 4.517178896368489e-06,
+      "loss": 0.3983,
+      "step": 307
+    },
+    {
+      "epoch": 1.4473219540906417,
+      "grad_norm": 0.7292077541351318,
+      "learning_rate": 4.513331385204762e-06,
+      "loss": 0.4199,
+      "step": 308
+    },
+    {
+      "epoch": 1.452030606238964,
+      "grad_norm": 0.403432160615921,
+      "learning_rate": 4.50947025705802e-06,
+      "loss": 0.4303,
+      "step": 309
+    },
+    {
+      "epoch": 1.4567392583872867,
+      "grad_norm": 0.3867107331752777,
+      "learning_rate": 4.505595538042652e-06,
+      "loss": 0.4178,
+      "step": 310
+    },
+    {
+      "epoch": 1.4614479105356093,
+      "grad_norm": 0.39861124753952026,
+      "learning_rate": 4.501707254364967e-06,
+      "loss": 0.4134,
+      "step": 311
+    },
+    {
+      "epoch": 1.4661565626839317,
+      "grad_norm": 0.3962036669254303,
+      "learning_rate": 4.497805432323015e-06,
+      "loss": 0.4302,
+      "step": 312
+    },
+    {
+      "epoch": 1.4708652148322543,
+      "grad_norm": 0.3967229127883911,
+      "learning_rate": 4.493890098306414e-06,
+      "loss": 0.4011,
+      "step": 313
+    },
+    {
+      "epoch": 1.475573866980577,
+      "grad_norm": 0.37047067284584045,
+      "learning_rate": 4.4899612787961675e-06,
+      "loss": 0.4066,
+      "step": 314
+    },
+    {
+      "epoch": 1.4802825191288993,
+      "grad_norm": 0.3553713858127594,
+      "learning_rate": 4.4860190003644895e-06,
+      "loss": 0.4214,
+      "step": 315
+    },
+    {
+      "epoch": 1.484991171277222,
+      "grad_norm": 0.3747401833534241,
+      "learning_rate": 4.4820632896746185e-06,
+      "loss": 0.4022,
+      "step": 316
+    },
+    {
+      "epoch": 1.4896998234255445,
+      "grad_norm": 0.3619938790798187,
+      "learning_rate": 4.478094173480642e-06,
+      "loss": 0.3946,
+      "step": 317
+    },
+    {
+      "epoch": 1.494408475573867,
+      "grad_norm": 0.3725791871547699,
+      "learning_rate": 4.474111678627318e-06,
+      "loss": 0.4031,
+      "step": 318
+    },
+    {
+      "epoch": 1.4991171277221895,
+      "grad_norm": 0.3898618817329407,
+      "learning_rate": 4.470115832049885e-06,
+      "loss": 0.4181,
+      "step": 319
+    },
+    {
+      "epoch": 1.5038257798705121,
+      "grad_norm": 0.3704550862312317,
+      "learning_rate": 4.466106660773884e-06,
+      "loss": 0.4043,
+      "step": 320
+    },
+    {
+      "epoch": 1.5085344320188345,
+      "grad_norm": 0.4264324903488159,
+      "learning_rate": 4.462084191914984e-06,
+      "loss": 0.401,
+      "step": 321
+    },
+    {
+      "epoch": 1.5132430841671571,
+      "grad_norm": 0.39826300740242004,
+      "learning_rate": 4.458048452678781e-06,
+      "loss": 0.3926,
+      "step": 322
+    },
+    {
+      "epoch": 1.5179517363154797,
+      "grad_norm": 0.3568350076675415,
+      "learning_rate": 4.453999470360633e-06,
+      "loss": 0.3904,
+      "step": 323
+    },
+    {
+      "epoch": 1.5226603884638021,
+      "grad_norm": 0.37382814288139343,
+      "learning_rate": 4.449937272345461e-06,
+      "loss": 0.3873,
+      "step": 324
+    },
+    {
+      "epoch": 1.5273690406121248,
+      "grad_norm": 0.41268882155418396,
+      "learning_rate": 4.445861886107567e-06,
+      "loss": 0.4063,
+      "step": 325
+    },
+    {
+      "epoch": 1.5320776927604474,
+      "grad_norm": 0.3825395107269287,
+      "learning_rate": 4.441773339210459e-06,
+      "loss": 0.4148,
+      "step": 326
+    },
+    {
+      "epoch": 1.5367863449087698,
+      "grad_norm": 0.38161659240722656,
+      "learning_rate": 4.437671659306647e-06,
+      "loss": 0.4228,
+      "step": 327
+    },
+    {
+      "epoch": 1.5414949970570924,
+      "grad_norm": 0.391406387090683,
+      "learning_rate": 4.43355687413747e-06,
+      "loss": 0.4283,
+      "step": 328
+    },
+    {
+      "epoch": 1.546203649205415,
+      "grad_norm": 0.9112624526023865,
+      "learning_rate": 4.429429011532902e-06,
+      "loss": 0.3876,
+      "step": 329
+    },
+    {
+      "epoch": 1.5509123013537374,
+      "grad_norm": 0.3513050377368927,
+      "learning_rate": 4.425288099411365e-06,
+      "loss": 0.4025,
+      "step": 330
+    },
+    {
+      "epoch": 1.5556209535020602,
+      "grad_norm": 0.40149223804473877,
+      "learning_rate": 4.421134165779538e-06,
+      "loss": 0.4183,
+      "step": 331
+    },
+    {
+      "epoch": 1.5603296056503826,
+      "grad_norm": 0.3666805624961853,
+      "learning_rate": 4.416967238732174e-06,
+      "loss": 0.4268,
+      "step": 332
+    },
+    {
+      "epoch": 1.565038257798705,
+      "grad_norm": 0.4094839096069336,
+      "learning_rate": 4.412787346451903e-06,
+      "loss": 0.4255,
+      "step": 333
+    },
+    {
+      "epoch": 1.5697469099470278,
+      "grad_norm": 0.38265460729599,
+      "learning_rate": 4.408594517209045e-06,
+      "loss": 0.4095,
+      "step": 334
+    },
+    {
+      "epoch": 1.5744555620953502,
+      "grad_norm": 0.36584508419036865,
+      "learning_rate": 4.4043887793614186e-06,
+      "loss": 0.4085,
+      "step": 335
+    },
+    {
+      "epoch": 1.5791642142436726,
+      "grad_norm": 0.3524089455604553,
+      "learning_rate": 4.400170161354146e-06,
+      "loss": 0.3903,
+      "step": 336
+    },
+    {
+      "epoch": 1.5838728663919954,
+      "grad_norm": 0.3962515592575073,
+      "learning_rate": 4.395938691719466e-06,
+      "loss": 0.4203,
+      "step": 337
+    },
+    {
+      "epoch": 1.5885815185403178,
+      "grad_norm": 0.38059180974960327,
+      "learning_rate": 4.391694399076537e-06,
+      "loss": 0.4084,
+      "step": 338
+    },
+    {
+      "epoch": 1.5932901706886404,
+      "grad_norm": 0.37096184492111206,
+      "learning_rate": 4.387437312131246e-06,
+      "loss": 0.4077,
+      "step": 339
+    },
+    {
+      "epoch": 1.597998822836963,
+      "grad_norm": 0.34766098856925964,
+      "learning_rate": 4.383167459676009e-06,
+      "loss": 0.4085,
+      "step": 340
+    },
+    {
+      "epoch": 1.6027074749852854,
+      "grad_norm": 0.3656401038169861,
+      "learning_rate": 4.378884870589585e-06,
+      "loss": 0.4098,
+      "step": 341
+    },
+    {
+      "epoch": 1.607416127133608,
+      "grad_norm": 0.43042734265327454,
+      "learning_rate": 4.374589573836875e-06,
+      "loss": 0.4143,
+      "step": 342
+    },
+    {
+      "epoch": 1.6121247792819307,
+      "grad_norm": 0.42422834038734436,
+      "learning_rate": 4.370281598468725e-06,
+      "loss": 0.3953,
+      "step": 343
+    },
+    {
+      "epoch": 1.616833431430253,
+      "grad_norm": 0.3533729910850525,
+      "learning_rate": 4.365960973621735e-06,
+      "loss": 0.4068,
+      "step": 344
+    },
+    {
+      "epoch": 1.6215420835785757,
+      "grad_norm": 0.36275553703308105,
+      "learning_rate": 4.361627728518054e-06,
+      "loss": 0.4014,
+      "step": 345
+    },
+    {
+      "epoch": 1.6262507357268983,
+      "grad_norm": 0.4120238721370697,
+      "learning_rate": 4.357281892465192e-06,
+      "loss": 0.4401,
+      "step": 346
+    },
+    {
+      "epoch": 1.6309593878752207,
+      "grad_norm": 0.3676035404205322,
+      "learning_rate": 4.352923494855813e-06,
+      "loss": 0.4034,
+      "step": 347
+    },
+    {
+      "epoch": 1.6356680400235433,
+      "grad_norm": 0.3960263729095459,
+      "learning_rate": 4.348552565167542e-06,
+      "loss": 0.3941,
+      "step": 348
+    },
+    {
+      "epoch": 1.6403766921718659,
+      "grad_norm": 0.3616948425769806,
+      "learning_rate": 4.344169132962763e-06,
+      "loss": 0.399,
+      "step": 349
+    },
+    {
+      "epoch": 1.6450853443201883,
+      "grad_norm": 0.39258840680122375,
+      "learning_rate": 4.33977322788842e-06,
+      "loss": 0.4038,
+      "step": 350
+    },
+    {
+      "epoch": 1.6497939964685109,
+      "grad_norm": 0.4016514718532562,
+      "learning_rate": 4.335364879675814e-06,
+      "loss": 0.4164,
+      "step": 351
+    },
+    {
+      "epoch": 1.6545026486168335,
+      "grad_norm": 0.38080155849456787,
+      "learning_rate": 4.330944118140406e-06,
+      "loss": 0.3953,
+      "step": 352
+    },
+    {
+      "epoch": 1.6592113007651559,
+      "grad_norm": 0.3435121774673462,
+      "learning_rate": 4.326510973181615e-06,
+      "loss": 0.391,
+      "step": 353
+    },
+    {
+      "epoch": 1.6639199529134785,
+      "grad_norm": 0.3891783356666565,
+      "learning_rate": 4.32206547478261e-06,
+      "loss": 0.4129,
+      "step": 354
+    },
+    {
+      "epoch": 1.668628605061801,
+      "grad_norm": 0.39722615480422974,
+      "learning_rate": 4.317607653010114e-06,
+      "loss": 0.4223,
+      "step": 355
+    },
+    {
+      "epoch": 1.6733372572101235,
+      "grad_norm": 0.3996773958206177,
+      "learning_rate": 4.313137538014199e-06,
+      "loss": 0.4141,
+      "step": 356
+    },
+    {
+      "epoch": 1.678045909358446,
+      "grad_norm": 0.39691588282585144,
+      "learning_rate": 4.308655160028078e-06,
+      "loss": 0.3786,
+      "step": 357
+    },
+    {
+      "epoch": 1.6827545615067687,
+      "grad_norm": 0.3999384343624115,
+      "learning_rate": 4.304160549367906e-06,
+      "loss": 0.4311,
+      "step": 358
+    },
+    {
+      "epoch": 1.687463213655091,
+      "grad_norm": 0.4473234713077545,
+      "learning_rate": 4.299653736432572e-06,
+      "loss": 0.4319,
+      "step": 359
+    },
+    {
+      "epoch": 1.6921718658034137,
+      "grad_norm": 0.38774654269218445,
+      "learning_rate": 4.295134751703493e-06,
+      "loss": 0.4166,
+      "step": 360
+    },
+    {
+      "epoch": 1.6968805179517363,
+      "grad_norm": 0.41464555263519287,
+      "learning_rate": 4.290603625744411e-06,
+      "loss": 0.4141,
+      "step": 361
+    },
+    {
+      "epoch": 1.7015891701000587,
+      "grad_norm": 0.3974348306655884,
+      "learning_rate": 4.286060389201181e-06,
+      "loss": 0.4265,
+      "step": 362
+    },
+    {
+      "epoch": 1.7062978222483816,
+      "grad_norm": 0.3926202356815338,
+      "learning_rate": 4.28150507280157e-06,
+      "loss": 0.4078,
+      "step": 363
+    },
+    {
+      "epoch": 1.711006474396704,
+      "grad_norm": 0.37342771887779236,
+      "learning_rate": 4.276937707355044e-06,
+      "loss": 0.3894,
+      "step": 364
+    },
+    {
+      "epoch": 1.7157151265450263,
+      "grad_norm": 0.37231311202049255,
+      "learning_rate": 4.272358323752563e-06,
+      "loss": 0.414,
+      "step": 365
+    },
+    {
+      "epoch": 1.7204237786933492,
+      "grad_norm": 0.3970523774623871,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.4008,
+      "step": 366
+    },
+    {
+      "epoch": 1.7251324308416716,
+      "grad_norm": 0.3824027180671692,
+      "learning_rate": 4.26316362604978e-06,
+      "loss": 0.4174,
+      "step": 367
+    },
+    {
+      "epoch": 1.729841082989994,
+      "grad_norm": 0.38245680928230286,
+      "learning_rate": 4.258548374136976e-06,
+      "loss": 0.431,
+      "step": 368
+    },
+    {
+      "epoch": 1.7345497351383168,
+      "grad_norm": 0.38463136553764343,
+      "learning_rate": 4.253921228442793e-06,
+      "loss": 0.4076,
+      "step": 369
+    },
+    {
+      "epoch": 1.7392583872866392,
+      "grad_norm": 0.3690101206302643,
+      "learning_rate": 4.249282220262507e-06,
+      "loss": 0.3978,
+      "step": 370
+    },
+    {
+      "epoch": 1.7439670394349618,
+      "grad_norm": 0.3739013969898224,
+      "learning_rate": 4.244631380971629e-06,
+      "loss": 0.4205,
+      "step": 371
+    },
+    {
+      "epoch": 1.7486756915832844,
+      "grad_norm": 0.3831653594970703,
+      "learning_rate": 4.239968742025684e-06,
+      "loss": 0.3965,
+      "step": 372
+    },
+    {
+      "epoch": 1.7533843437316068,
+      "grad_norm": 0.3808199465274811,
+      "learning_rate": 4.235294334960005e-06,
+      "loss": 0.408,
+      "step": 373
+    },
+    {
+      "epoch": 1.7580929958799294,
+      "grad_norm": 0.40046992897987366,
+      "learning_rate": 4.230608191389518e-06,
+      "loss": 0.407,
+      "step": 374
+    },
+    {
+      "epoch": 1.762801648028252,
+      "grad_norm": 0.37953296303749084,
+      "learning_rate": 4.225910343008527e-06,
+      "loss": 0.4187,
+      "step": 375
+    },
+    {
+      "epoch": 1.7675103001765744,
+      "grad_norm": 0.4310925006866455,
+      "learning_rate": 4.2212008215905e-06,
+      "loss": 0.4241,
+      "step": 376
+    },
+    {
+      "epoch": 1.772218952324897,
+      "grad_norm": 0.383286714553833,
+      "learning_rate": 4.216479658987856e-06,
+      "loss": 0.3998,
+      "step": 377
+    },
+    {
+      "epoch": 1.7769276044732196,
+      "grad_norm": 0.41131308674812317,
+      "learning_rate": 4.211746887131747e-06,
+      "loss": 0.3963,
+      "step": 378
+    },
+    {
+      "epoch": 1.781636256621542,
+      "grad_norm": 0.3883269727230072,
+      "learning_rate": 4.207002538031843e-06,
+      "loss": 0.4004,
+      "step": 379
+    },
+    {
+      "epoch": 1.7863449087698646,
+      "grad_norm": 0.39173465967178345,
+      "learning_rate": 4.202246643776116e-06,
+      "loss": 0.3966,
+      "step": 380
+    },
+    {
+      "epoch": 1.7910535609181872,
+      "grad_norm": 0.402128130197525,
+      "learning_rate": 4.1974792365306235e-06,
+      "loss": 0.3988,
+      "step": 381
+    },
+    {
+      "epoch": 1.7957622130665096,
+      "grad_norm": 0.3661210238933563,
+      "learning_rate": 4.192700348539288e-06,
+      "loss": 0.3783,
+      "step": 382
+    },
+    {
+      "epoch": 1.8004708652148322,
+      "grad_norm": 0.3758254051208496,
+      "learning_rate": 4.187910012123683e-06,
+      "loss": 0.4074,
+      "step": 383
+    },
+    {
+      "epoch": 1.8051795173631549,
+      "grad_norm": 0.37446528673171997,
+      "learning_rate": 4.183108259682811e-06,
+      "loss": 0.4084,
+      "step": 384
+    },
+    {
+      "epoch": 1.8098881695114772,
+      "grad_norm": 0.4256710410118103,
+      "learning_rate": 4.178295123692886e-06,
+      "loss": 0.4184,
+      "step": 385
+    },
+    {
+      "epoch": 1.8145968216597999,
+      "grad_norm": 0.3782687783241272,
+      "learning_rate": 4.1734706367071155e-06,
+      "loss": 0.4129,
+      "step": 386
+    },
+    {
+      "epoch": 1.8193054738081225,
+      "grad_norm": 0.43555963039398193,
+      "learning_rate": 4.168634831355474e-06,
+      "loss": 0.4243,
+      "step": 387
+    },
+    {
+      "epoch": 1.8240141259564449,
+      "grad_norm": 0.3523005247116089,
+      "learning_rate": 4.163787740344492e-06,
+      "loss": 0.3873,
+      "step": 388
+    },
+    {
+      "epoch": 1.8287227781047675,
+      "grad_norm": 0.36786767840385437,
+      "learning_rate": 4.1589293964570255e-06,
+      "loss": 0.3882,
+      "step": 389
+    },
+    {
+      "epoch": 1.83343143025309,
+      "grad_norm": 0.3621681332588196,
+      "learning_rate": 4.154059832552041e-06,
+      "loss": 0.3717,
+      "step": 390
+    },
+    {
+      "epoch": 1.8381400824014125,
+      "grad_norm": 0.349269837141037,
+      "learning_rate": 4.149179081564388e-06,
+      "loss": 0.3681,
+      "step": 391
+    },
+    {
+      "epoch": 1.8428487345497353,
+      "grad_norm": 0.38650140166282654,
+      "learning_rate": 4.144287176504582e-06,
+      "loss": 0.4063,
+      "step": 392
+    },
+    {
+      "epoch": 1.8475573866980577,
+      "grad_norm": 0.4057513475418091,
+      "learning_rate": 4.139384150458575e-06,
+      "loss": 0.4189,
+      "step": 393
+    },
+    {
+      "epoch": 1.85226603884638,
+      "grad_norm": 0.39712995290756226,
+      "learning_rate": 4.134470036587536e-06,
+      "loss": 0.4077,
+      "step": 394
+    },
+    {
+      "epoch": 1.856974690994703,
+      "grad_norm": 0.38149163126945496,
+      "learning_rate": 4.129544868127626e-06,
+      "loss": 0.4199,
+      "step": 395
+    },
+    {
+      "epoch": 1.8616833431430253,
+      "grad_norm": 0.41102463006973267,
+      "learning_rate": 4.124608678389772e-06,
+      "loss": 0.3964,
+      "step": 396
+    },
+    {
+      "epoch": 1.8663919952913477,
+      "grad_norm": 0.3947491943836212,
+      "learning_rate": 4.119661500759442e-06,
+      "loss": 0.418,
+      "step": 397
+    },
+    {
+      "epoch": 1.8711006474396705,
+      "grad_norm": 0.3679056465625763,
+      "learning_rate": 4.114703368696421e-06,
+      "loss": 0.3903,
+      "step": 398
+    },
+    {
+      "epoch": 1.875809299587993,
+      "grad_norm": 0.37078574299812317,
+      "learning_rate": 4.109734315734584e-06,
+      "loss": 0.4044,
+      "step": 399
+    },
+    {
+      "epoch": 1.8805179517363155,
+      "grad_norm": 0.38189971446990967,
+      "learning_rate": 4.104754375481665e-06,
+      "loss": 0.4182,
+      "step": 400
+    },
+    {
+      "epoch": 1.8852266038846381,
+      "grad_norm": 0.37885355949401855,
+      "learning_rate": 4.0997635816190354e-06,
+      "loss": 0.4146,
+      "step": 401
+    },
+    {
+      "epoch": 1.8899352560329605,
+      "grad_norm": 0.38311830163002014,
+      "learning_rate": 4.094761967901474e-06,
+      "loss": 0.4101,
+      "step": 402
+    },
+    {
+      "epoch": 1.8946439081812831,
+      "grad_norm": 0.3592222332954407,
+      "learning_rate": 4.0897495681569385e-06,
+      "loss": 0.4124,
+      "step": 403
+    },
+    {
+      "epoch": 1.8993525603296058,
+      "grad_norm": 0.449899822473526,
+      "learning_rate": 4.084726416286338e-06,
+      "loss": 0.4011,
+      "step": 404
+    },
+    {
+      "epoch": 1.9040612124779281,
+      "grad_norm": 0.3764320909976959,
+      "learning_rate": 4.079692546263299e-06,
+      "loss": 0.4068,
+      "step": 405
+    },
+    {
+      "epoch": 1.9087698646262508,
+      "grad_norm": 0.39345216751098633,
+      "learning_rate": 4.074647992133946e-06,
+      "loss": 0.3942,
+      "step": 406
+    },
+    {
+      "epoch": 1.9134785167745734,
+      "grad_norm": 0.4699568450450897,
+      "learning_rate": 4.069592788016657e-06,
+      "loss": 0.4245,
+      "step": 407
+    },
+    {
+      "epoch": 1.9181871689228958,
+      "grad_norm": 0.3634001910686493,
+      "learning_rate": 4.064526968101844e-06,
+      "loss": 0.4236,
+      "step": 408
+    },
+    {
+      "epoch": 1.9228958210712184,
+      "grad_norm": 0.3803250193595886,
+      "learning_rate": 4.059450566651719e-06,
+      "loss": 0.4097,
+      "step": 409
+    },
+    {
+      "epoch": 1.927604473219541,
+      "grad_norm": 0.39286285638809204,
+      "learning_rate": 4.054363618000058e-06,
+      "loss": 0.4107,
+      "step": 410
+    },
+    {
+      "epoch": 1.9323131253678634,
+      "grad_norm": 0.40081989765167236,
+      "learning_rate": 4.049266156551975e-06,
+      "loss": 0.4439,
+      "step": 411
+    },
+    {
+      "epoch": 1.937021777516186,
+      "grad_norm": 0.3855789601802826,
+      "learning_rate": 4.044158216783684e-06,
+      "loss": 0.4104,
+      "step": 412
+    },
+    {
+      "epoch": 1.9417304296645086,
+      "grad_norm": 0.33973613381385803,
+      "learning_rate": 4.03903983324227e-06,
+      "loss": 0.411,
+      "step": 413
+    },
+    {
+      "epoch": 1.946439081812831,
+      "grad_norm": 0.3574191927909851,
+      "learning_rate": 4.0339110405454535e-06,
+      "loss": 0.4028,
+      "step": 414
+    },
+    {
+      "epoch": 1.9511477339611536,
+      "grad_norm": 0.3561255633831024,
+      "learning_rate": 4.028771873381354e-06,
+      "loss": 0.4045,
+      "step": 415
+    },
+    {
+      "epoch": 1.9558563861094762,
+      "grad_norm": 0.3795716464519501,
+      "learning_rate": 4.02362236650826e-06,
+      "loss": 0.4019,
+      "step": 416
+    },
+    {
+      "epoch": 1.9605650382577986,
+      "grad_norm": 0.40633508563041687,
+      "learning_rate": 4.018462554754393e-06,
+      "loss": 0.3952,
+      "step": 417
+    },
+    {
+      "epoch": 1.9652736904061212,
+      "grad_norm": 0.35927334427833557,
+      "learning_rate": 4.013292473017666e-06,
+      "loss": 0.4009,
+      "step": 418
+    },
+    {
+      "epoch": 1.9699823425544438,
+      "grad_norm": 0.39236533641815186,
+      "learning_rate": 4.008112156265455e-06,
+      "loss": 0.4515,
+      "step": 419
+    },
+    {
+      "epoch": 1.9746909947027662,
+      "grad_norm": 0.41305258870124817,
+      "learning_rate": 4.002921639534362e-06,
+      "loss": 0.4234,
+      "step": 420
+    },
+    {
+      "epoch": 1.9793996468510888,
+      "grad_norm": 0.36336320638656616,
+      "learning_rate": 3.997720957929971e-06,
+      "loss": 0.3876,
+      "step": 421
+    },
+    {
+      "epoch": 1.9841082989994114,
+      "grad_norm": 0.5077244639396667,
+      "learning_rate": 3.992510146626618e-06,
+      "loss": 0.4333,
+      "step": 422
+    },
+    {
+      "epoch": 1.9888169511477338,
+      "grad_norm": 0.3557424247264862,
+      "learning_rate": 3.987289240867148e-06,
+      "loss": 0.3786,
+      "step": 423
+    },
+    {
+      "epoch": 1.9935256032960567,
+      "grad_norm": 0.37336745858192444,
+      "learning_rate": 3.9820582759626825e-06,
+      "loss": 0.3905,
+      "step": 424
+    },
+    {
+      "epoch": 1.998234255444379,
+      "grad_norm": 0.3858487010002136,
+      "learning_rate": 3.976817287292374e-06,
+      "loss": 0.4188,
+      "step": 425
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3858487010002136,
+      "learning_rate": 3.971566310303171e-06,
+      "loss": 0.395,
+      "step": 426
+    },
+    {
+      "epoch": 2.0047086521483224,
+      "grad_norm": 0.7181822061538696,
+      "learning_rate": 3.966305380509576e-06,
+      "loss": 0.373,
+      "step": 427
+    },
+    {
+      "epoch": 2.0094173042966452,
+      "grad_norm": 0.367645800113678,
+      "learning_rate": 3.96103453349341e-06,
+      "loss": 0.3708,
+      "step": 428
+    },
+    {
+      "epoch": 2.0141259564449676,
+      "grad_norm": 0.4022150933742523,
+      "learning_rate": 3.955753804903562e-06,
+      "loss": 0.3693,
+      "step": 429
+    },
+    {
+      "epoch": 2.01883460859329,
+      "grad_norm": 0.3676547408103943,
+      "learning_rate": 3.950463230455762e-06,
+      "loss": 0.3889,
+      "step": 430
+    },
+    {
+      "epoch": 2.023543260741613,
+      "grad_norm": 0.41727522015571594,
+      "learning_rate": 3.945162845932323e-06,
+      "loss": 0.3666,
+      "step": 431
+    },
+    {
+      "epoch": 2.0282519128899352,
+      "grad_norm": 0.396017849445343,
+      "learning_rate": 3.939852687181916e-06,
+      "loss": 0.3768,
+      "step": 432
+    },
+    {
+      "epoch": 2.0329605650382576,
+      "grad_norm": 0.3791135549545288,
+      "learning_rate": 3.9345327901193104e-06,
+      "loss": 0.3507,
+      "step": 433
+    },
+    {
+      "epoch": 2.0376692171865805,
+      "grad_norm": 0.3801550269126892,
+      "learning_rate": 3.929203190725147e-06,
+      "loss": 0.3858,
+      "step": 434
+    },
+    {
+      "epoch": 2.042377869334903,
+      "grad_norm": 0.3655739724636078,
+      "learning_rate": 3.923863925045685e-06,
+      "loss": 0.3745,
+      "step": 435
+    },
+    {
+      "epoch": 2.0470865214832252,
+      "grad_norm": 0.37982434034347534,
+      "learning_rate": 3.918515029192559e-06,
+      "loss": 0.358,
+      "step": 436
+    },
+    {
+      "epoch": 2.051795173631548,
+      "grad_norm": 0.36431175470352173,
+      "learning_rate": 3.9131565393425385e-06,
+      "loss": 0.3559,
+      "step": 437
+    },
+    {
+      "epoch": 2.0565038257798705,
+      "grad_norm": 0.3713436722755432,
+      "learning_rate": 3.907788491737281e-06,
+      "loss": 0.3753,
+      "step": 438
+    },
+    {
+      "epoch": 2.061212477928193,
+      "grad_norm": 0.3815770447254181,
+      "learning_rate": 3.902410922683084e-06,
+      "loss": 0.3456,
+      "step": 439
+    },
+    {
+      "epoch": 2.0659211300765157,
+      "grad_norm": 0.34248778223991394,
+      "learning_rate": 3.897023868550649e-06,
+      "loss": 0.3591,
+      "step": 440
+    },
+    {
+      "epoch": 2.070629782224838,
+      "grad_norm": 0.37305399775505066,
+      "learning_rate": 3.89162736577482e-06,
+      "loss": 0.3974,
+      "step": 441
+    },
+    {
+      "epoch": 2.075338434373161,
+      "grad_norm": 0.41609179973602295,
+      "learning_rate": 3.886221450854355e-06,
+      "loss": 0.3372,
+      "step": 442
+    },
+    {
+      "epoch": 2.0800470865214833,
+      "grad_norm": 0.3434038460254669,
+      "learning_rate": 3.880806160351663e-06,
+      "loss": 0.3597,
+      "step": 443
+    },
+    {
+      "epoch": 2.0847557386698057,
+      "grad_norm": 0.38777461647987366,
+      "learning_rate": 3.875381530892569e-06,
+      "loss": 0.3713,
+      "step": 444
+    },
+    {
+      "epoch": 2.0894643908181285,
+      "grad_norm": 0.48137035965919495,
+      "learning_rate": 3.869947599166059e-06,
+      "loss": 0.3803,
+      "step": 445
+    },
+    {
+      "epoch": 2.094173042966451,
+      "grad_norm": 0.4451170563697815,
+      "learning_rate": 3.864504401924031e-06,
+      "loss": 0.3766,
+      "step": 446
+    },
+    {
+      "epoch": 2.0988816951147733,
+      "grad_norm": 0.38987934589385986,
+      "learning_rate": 3.859051975981056e-06,
+      "loss": 0.3683,
+      "step": 447
+    },
+    {
+      "epoch": 2.103590347263096,
+      "grad_norm": 0.3817046582698822,
+      "learning_rate": 3.853590358214119e-06,
+      "loss": 0.3545,
+      "step": 448
+    },
+    {
+      "epoch": 2.1082989994114185,
+      "grad_norm": 0.3981139361858368,
+      "learning_rate": 3.848119585562371e-06,
+      "loss": 0.3532,
+      "step": 449
+    },
+    {
+      "epoch": 2.113007651559741,
+      "grad_norm": 0.3644142150878906,
+      "learning_rate": 3.842639695026885e-06,
+      "loss": 0.3739,
+      "step": 450
+    },
+    {
+      "epoch": 2.1177163037080637,
+      "grad_norm": 0.37675121426582336,
+      "learning_rate": 3.8371507236704e-06,
+      "loss": 0.3777,
+      "step": 451
+    },
+    {
+      "epoch": 2.122424955856386,
+      "grad_norm": 0.3827580213546753,
+      "learning_rate": 3.831652708617073e-06,
+      "loss": 0.379,
+      "step": 452
+    },
+    {
+      "epoch": 2.1271336080047085,
+      "grad_norm": 0.39171484112739563,
+      "learning_rate": 3.826145687052225e-06,
+      "loss": 0.3711,
+      "step": 453
+    },
+    {
+      "epoch": 2.1318422601530314,
+      "grad_norm": 0.3718264400959015,
+      "learning_rate": 3.820629696222096e-06,
+      "loss": 0.3604,
+      "step": 454
+    },
+    {
+      "epoch": 2.1365509123013537,
+      "grad_norm": 0.5579127669334412,
+      "learning_rate": 3.815104773433585e-06,
+      "loss": 0.3893,
+      "step": 455
+    },
+    {
+      "epoch": 2.141259564449676,
+      "grad_norm": 0.4234530031681061,
+      "learning_rate": 3.809570956054004e-06,
+      "loss": 0.3794,
+      "step": 456
+    },
+    {
+      "epoch": 2.145968216597999,
+      "grad_norm": 0.3809348940849304,
+      "learning_rate": 3.8040282815108196e-06,
+      "loss": 0.3808,
+      "step": 457
+    },
+    {
+      "epoch": 2.1506768687463214,
+      "grad_norm": 0.4055100381374359,
+      "learning_rate": 3.7984767872914076e-06,
+      "loss": 0.372,
+      "step": 458
+    },
+    {
+      "epoch": 2.1553855208946437,
+      "grad_norm": 0.36842137575149536,
+      "learning_rate": 3.79291651094279e-06,
+      "loss": 0.37,
+      "step": 459
+    },
+    {
+      "epoch": 2.1600941730429666,
+      "grad_norm": 0.35666927695274353,
+      "learning_rate": 3.78734749007139e-06,
+      "loss": 0.358,
+      "step": 460
+    },
+    {
+      "epoch": 2.164802825191289,
+      "grad_norm": 0.33630210161209106,
+      "learning_rate": 3.7817697623427707e-06,
+      "loss": 0.3551,
+      "step": 461
+    },
+    {
+      "epoch": 2.1695114773396114,
+      "grad_norm": 0.35024064779281616,
+      "learning_rate": 3.776183365481385e-06,
+      "loss": 0.3696,
+      "step": 462
+    },
+    {
+      "epoch": 2.174220129487934,
+      "grad_norm": 0.40014365315437317,
+      "learning_rate": 3.7705883372703177e-06,
+      "loss": 0.379,
+      "step": 463
+    },
+    {
+      "epoch": 2.1789287816362566,
+      "grad_norm": 0.3819674253463745,
+      "learning_rate": 3.7649847155510314e-06,
+      "loss": 0.3687,
+      "step": 464
+    },
+    {
+      "epoch": 2.183637433784579,
+      "grad_norm": 0.3578038215637207,
+      "learning_rate": 3.7593725382231107e-06,
+      "loss": 0.3694,
+      "step": 465
+    },
+    {
+      "epoch": 2.188346085932902,
+      "grad_norm": 0.35197731852531433,
+      "learning_rate": 3.7537518432440034e-06,
+      "loss": 0.3647,
+      "step": 466
+    },
+    {
+      "epoch": 2.193054738081224,
+      "grad_norm": 0.3819369673728943,
+      "learning_rate": 3.7481226686287676e-06,
+      "loss": 0.3761,
+      "step": 467
+    },
+    {
+      "epoch": 2.1977633902295466,
+      "grad_norm": 0.3859822750091553,
+      "learning_rate": 3.742485052449812e-06,
+      "loss": 0.3709,
+      "step": 468
+    },
+    {
+      "epoch": 2.2024720423778694,
+      "grad_norm": 0.3550996482372284,
+      "learning_rate": 3.736839032836638e-06,
+      "loss": 0.3708,
+      "step": 469
+    },
+    {
+      "epoch": 2.207180694526192,
+      "grad_norm": 0.4063422977924347,
+      "learning_rate": 3.731184647975585e-06,
+      "loss": 0.3898,
+      "step": 470
+    },
+    {
+      "epoch": 2.211889346674514,
+      "grad_norm": 0.36760836839675903,
+      "learning_rate": 3.725521936109567e-06,
+      "loss": 0.3679,
+      "step": 471
+    },
+    {
+      "epoch": 2.216597998822837,
+      "grad_norm": 0.37566909193992615,
+      "learning_rate": 3.7198509355378208e-06,
+      "loss": 0.3659,
+      "step": 472
+    },
+    {
+      "epoch": 2.2213066509711594,
+      "grad_norm": 0.3421824276447296,
+      "learning_rate": 3.7141716846156406e-06,
+      "loss": 0.3556,
+      "step": 473
+    },
+    {
+      "epoch": 2.226015303119482,
+      "grad_norm": 0.36609965562820435,
+      "learning_rate": 3.70848422175412e-06,
+      "loss": 0.3769,
+      "step": 474
+    },
+    {
+      "epoch": 2.2307239552678046,
+      "grad_norm": 0.3647328019142151,
+      "learning_rate": 3.702788585419896e-06,
+      "loss": 0.391,
+      "step": 475
+    },
+    {
+      "epoch": 2.235432607416127,
+      "grad_norm": 0.3609878122806549,
+      "learning_rate": 3.6970848141348857e-06,
+      "loss": 0.3516,
+      "step": 476
+    },
+    {
+      "epoch": 2.24014125956445,
+      "grad_norm": 0.36001378297805786,
+      "learning_rate": 3.6913729464760235e-06,
+      "loss": 0.3583,
+      "step": 477
+    },
+    {
+      "epoch": 2.2448499117127723,
+      "grad_norm": 0.3540048897266388,
+      "learning_rate": 3.6856530210750063e-06,
+      "loss": 0.3841,
+      "step": 478
+    },
+    {
+      "epoch": 2.2495585638610947,
+      "grad_norm": 0.3976086974143982,
+      "learning_rate": 3.679925076618027e-06,
+      "loss": 0.3686,
+      "step": 479
+    },
+    {
+      "epoch": 2.2542672160094175,
+      "grad_norm": 0.3643709421157837,
+      "learning_rate": 3.674189151845515e-06,
+      "loss": 0.3686,
+      "step": 480
+    },
+    {
+      "epoch": 2.25897586815774,
+      "grad_norm": 0.38158276677131653,
+      "learning_rate": 3.668445285551874e-06,
+      "loss": 0.3621,
+      "step": 481
+    },
+    {
+      "epoch": 2.2636845203060623,
+      "grad_norm": 0.3460032343864441,
+      "learning_rate": 3.662693516585219e-06,
+      "loss": 0.3837,
+      "step": 482
+    },
+    {
+      "epoch": 2.268393172454385,
+      "grad_norm": 0.3750056326389313,
+      "learning_rate": 3.6569338838471137e-06,
+      "loss": 0.3531,
+      "step": 483
+    },
+    {
+      "epoch": 2.2731018246027075,
+      "grad_norm": 0.3516237139701843,
+      "learning_rate": 3.6511664262923096e-06,
+      "loss": 0.3937,
+      "step": 484
+    },
+    {
+      "epoch": 2.27781047675103,
+      "grad_norm": 0.3862423300743103,
+      "learning_rate": 3.645391182928479e-06,
+      "loss": 0.3679,
+      "step": 485
+    },
+    {
+      "epoch": 2.2825191288993527,
+      "grad_norm": 0.34053516387939453,
+      "learning_rate": 3.6396081928159514e-06,
+      "loss": 0.3742,
+      "step": 486
+    },
+    {
+      "epoch": 2.287227781047675,
+      "grad_norm": 0.3533380925655365,
+      "learning_rate": 3.6338174950674555e-06,
+      "loss": 0.3806,
+      "step": 487
+    },
+    {
+      "epoch": 2.2919364331959975,
+      "grad_norm": 0.40273189544677734,
+      "learning_rate": 3.6280191288478437e-06,
+      "loss": 0.3668,
+      "step": 488
+    },
+    {
+      "epoch": 2.2966450853443203,
+      "grad_norm": 0.35413995385169983,
+      "learning_rate": 3.622213133373839e-06,
+      "loss": 0.3591,
+      "step": 489
+    },
+    {
+      "epoch": 2.3013537374926427,
+      "grad_norm": 0.33633658289909363,
+      "learning_rate": 3.6163995479137605e-06,
+      "loss": 0.3739,
+      "step": 490
+    },
+    {
+      "epoch": 2.306062389640965,
+      "grad_norm": 0.3628805875778198,
+      "learning_rate": 3.6105784117872633e-06,
+      "loss": 0.3657,
+      "step": 491
+    },
+    {
+      "epoch": 2.310771041789288,
+      "grad_norm": 0.34639039635658264,
+      "learning_rate": 3.604749764365069e-06,
+      "loss": 0.3684,
+      "step": 492
+    },
+    {
+      "epoch": 2.3154796939376103,
+      "grad_norm": 0.37136930227279663,
+      "learning_rate": 3.5989136450687035e-06,
+      "loss": 0.3991,
+      "step": 493
+    },
+    {
+      "epoch": 2.3201883460859327,
+      "grad_norm": 0.4088273048400879,
+      "learning_rate": 3.593070093370227e-06,
+      "loss": 0.382,
+      "step": 494
+    },
+    {
+      "epoch": 2.3248969982342556,
+      "grad_norm": 0.39885440468788147,
+      "learning_rate": 3.5872191487919662e-06,
+      "loss": 0.3816,
+      "step": 495
+    },
+    {
+      "epoch": 2.329605650382578,
+      "grad_norm": 0.36275333166122437,
+      "learning_rate": 3.581360850906253e-06,
+      "loss": 0.3491,
+      "step": 496
+    },
+    {
+      "epoch": 2.3343143025309008,
+      "grad_norm": 0.3472638726234436,
+      "learning_rate": 3.575495239335148e-06,
+      "loss": 0.3589,
+      "step": 497
+    },
+    {
+      "epoch": 2.339022954679223,
+      "grad_norm": 0.3820279836654663,
+      "learning_rate": 3.5696223537501817e-06,
+      "loss": 0.3826,
+      "step": 498
+    },
+    {
+      "epoch": 2.3437316068275456,
+      "grad_norm": 0.3917175829410553,
+      "learning_rate": 3.5637422338720785e-06,
+      "loss": 0.3624,
+      "step": 499
+    },
+    {
+      "epoch": 2.3484402589758684,
+      "grad_norm": 0.3817431628704071,
+      "learning_rate": 3.5578549194704916e-06,
+      "loss": 0.3846,
+      "step": 500
+    },
+    {
+      "epoch": 2.353148911124191,
+      "grad_norm": 0.358780175447464,
+      "learning_rate": 3.5519604503637362e-06,
+      "loss": 0.3693,
+      "step": 501
+    },
+    {
+      "epoch": 2.357857563272513,
+      "grad_norm": 0.3502993583679199,
+      "learning_rate": 3.5460588664185138e-06,
+      "loss": 0.3618,
+      "step": 502
+    },
+    {
+      "epoch": 2.362566215420836,
+      "grad_norm": 0.35879674553871155,
+      "learning_rate": 3.5401502075496506e-06,
+      "loss": 0.3847,
+      "step": 503
+    },
+    {
+      "epoch": 2.3672748675691584,
+      "grad_norm": 0.38985323905944824,
+      "learning_rate": 3.534234513719821e-06,
+      "loss": 0.3622,
+      "step": 504
+    },
+    {
+      "epoch": 2.371983519717481,
+      "grad_norm": 0.34615063667297363,
+      "learning_rate": 3.528311824939279e-06,
+      "loss": 0.3682,
+      "step": 505
+    },
+    {
+      "epoch": 2.3766921718658036,
+      "grad_norm": 0.35256028175354004,
+      "learning_rate": 3.5223821812655906e-06,
+      "loss": 0.3681,
+      "step": 506
+    },
+    {
+      "epoch": 2.381400824014126,
+      "grad_norm": 0.3890756666660309,
+      "learning_rate": 3.5164456228033583e-06,
+      "loss": 0.356,
+      "step": 507
+    },
+    {
+      "epoch": 2.3861094761624484,
+      "grad_norm": 0.3554323613643646,
+      "learning_rate": 3.510502189703955e-06,
+      "loss": 0.3812,
+      "step": 508
+    },
+    {
+      "epoch": 2.3908181283107712,
+      "grad_norm": 0.3922102153301239,
+      "learning_rate": 3.504551922165246e-06,
+      "loss": 0.3658,
+      "step": 509
+    },
+    {
+      "epoch": 2.3955267804590936,
+      "grad_norm": 0.3660562336444855,
+      "learning_rate": 3.4985948604313243e-06,
+      "loss": 0.3775,
+      "step": 510
+    },
+    {
+      "epoch": 2.400235432607416,
+      "grad_norm": 0.36027809977531433,
+      "learning_rate": 3.492631044792232e-06,
+      "loss": 0.3703,
+      "step": 511
+    },
+    {
+      "epoch": 2.404944084755739,
+      "grad_norm": 0.3643735647201538,
+      "learning_rate": 3.4866605155836913e-06,
+      "loss": 0.3538,
+      "step": 512
+    },
+    {
+      "epoch": 2.4096527369040612,
+      "grad_norm": 0.36816880106925964,
+      "learning_rate": 3.480683313186833e-06,
+      "loss": 0.3819,
+      "step": 513
+    },
+    {
+      "epoch": 2.4143613890523836,
+      "grad_norm": 0.3588556945323944,
+      "learning_rate": 3.474699478027918e-06,
+      "loss": 0.3531,
+      "step": 514
+    },
+    {
+      "epoch": 2.4190700412007065,
+      "grad_norm": 0.3689442574977875,
+      "learning_rate": 3.4687090505780707e-06,
+      "loss": 0.3831,
+      "step": 515
+    },
+    {
+      "epoch": 2.423778693349029,
+      "grad_norm": 0.37948766350746155,
+      "learning_rate": 3.4627120713529986e-06,
+      "loss": 0.3383,
+      "step": 516
+    },
+    {
+      "epoch": 2.4284873454973512,
+      "grad_norm": 0.3453068435192108,
+      "learning_rate": 3.4567085809127247e-06,
+      "loss": 0.3733,
+      "step": 517
+    },
+    {
+      "epoch": 2.433195997645674,
+      "grad_norm": 0.3405013382434845,
+      "learning_rate": 3.450698619861308e-06,
+      "loss": 0.3748,
+      "step": 518
+    },
+    {
+      "epoch": 2.4379046497939965,
+      "grad_norm": 0.364133358001709,
+      "learning_rate": 3.444682228846571e-06,
+      "loss": 0.3774,
+      "step": 519
+    },
+    {
+      "epoch": 2.442613301942319,
+      "grad_norm": 0.3705088198184967,
+      "learning_rate": 3.4386594485598257e-06,
+      "loss": 0.3658,
+      "step": 520
+    },
+    {
+      "epoch": 2.4473219540906417,
+      "grad_norm": 0.36289486289024353,
+      "learning_rate": 3.4326303197355955e-06,
+      "loss": 0.3869,
+      "step": 521
+    },
+    {
+      "epoch": 2.452030606238964,
+      "grad_norm": 0.40453988313674927,
+      "learning_rate": 3.426594883151344e-06,
+      "loss": 0.3551,
+      "step": 522
+    },
+    {
+      "epoch": 2.4567392583872865,
+      "grad_norm": 0.34340184926986694,
+      "learning_rate": 3.4205531796271953e-06,
+      "loss": 0.3661,
+      "step": 523
+    },
+    {
+      "epoch": 2.4614479105356093,
+      "grad_norm": 0.37485653162002563,
+      "learning_rate": 3.4145052500256595e-06,
+      "loss": 0.3637,
+      "step": 524
+    },
+    {
+      "epoch": 2.4661565626839317,
+      "grad_norm": 0.3690657913684845,
+      "learning_rate": 3.4084511352513557e-06,
+      "loss": 0.3726,
+      "step": 525
+    },
+    {
+      "epoch": 2.470865214832254,
+      "grad_norm": 0.41782814264297485,
+      "learning_rate": 3.402390876250737e-06,
+      "loss": 0.3766,
+      "step": 526
+    },
+    {
+      "epoch": 2.475573866980577,
+      "grad_norm": 0.3712344169616699,
+      "learning_rate": 3.3963245140118116e-06,
+      "loss": 0.3825,
+      "step": 527
+    },
+    {
+      "epoch": 2.4802825191288993,
+      "grad_norm": 0.40473493933677673,
+      "learning_rate": 3.3902520895638674e-06,
+      "loss": 0.3752,
+      "step": 528
+    },
+    {
+      "epoch": 2.4849911712772217,
+      "grad_norm": 0.38602185249328613,
+      "learning_rate": 3.3841736439771926e-06,
+      "loss": 0.3885,
+      "step": 529
+    },
+    {
+      "epoch": 2.4896998234255445,
+      "grad_norm": 0.40137943625450134,
+      "learning_rate": 3.3780892183627977e-06,
+      "loss": 0.3641,
+      "step": 530
+    },
+    {
+      "epoch": 2.494408475573867,
+      "grad_norm": 0.3662640154361725,
+      "learning_rate": 3.371998853872142e-06,
+      "loss": 0.3968,
+      "step": 531
+    },
+    {
+      "epoch": 2.4991171277221893,
+      "grad_norm": 0.4132643938064575,
+      "learning_rate": 3.365902591696848e-06,
+      "loss": 0.3614,
+      "step": 532
+    },
+    {
+      "epoch": 2.503825779870512,
+      "grad_norm": 0.35764506459236145,
+      "learning_rate": 3.3598004730684287e-06,
+      "loss": 0.3765,
+      "step": 533
+    },
+    {
+      "epoch": 2.5085344320188345,
+      "grad_norm": 0.4405362904071808,
+      "learning_rate": 3.3536925392580066e-06,
+      "loss": 0.3579,
+      "step": 534
+    },
+    {
+      "epoch": 2.513243084167157,
+      "grad_norm": 0.38784265518188477,
+      "learning_rate": 3.3475788315760344e-06,
+      "loss": 0.3801,
+      "step": 535
+    },
+    {
+      "epoch": 2.5179517363154797,
+      "grad_norm": 0.3985704183578491,
+      "learning_rate": 3.3414593913720163e-06,
+      "loss": 0.3645,
+      "step": 536
+    },
+    {
+      "epoch": 2.522660388463802,
+      "grad_norm": 0.3759413957595825,
+      "learning_rate": 3.3353342600342263e-06,
+      "loss": 0.3976,
+      "step": 537
+    },
+    {
+      "epoch": 2.5273690406121245,
+      "grad_norm": 0.3630805015563965,
+      "learning_rate": 3.329203478989431e-06,
+      "loss": 0.3709,
+      "step": 538
+    },
+    {
+      "epoch": 2.5320776927604474,
+      "grad_norm": 0.35673877596855164,
+      "learning_rate": 3.32306708970261e-06,
+      "loss": 0.3843,
+      "step": 539
+    },
+    {
+      "epoch": 2.5367863449087698,
+      "grad_norm": 0.43144941329956055,
+      "learning_rate": 3.31692513367667e-06,
+      "loss": 0.3495,
+      "step": 540
+    },
+    {
+      "epoch": 2.541494997057092,
+      "grad_norm": 0.4056400656700134,
+      "learning_rate": 3.310777652452171e-06,
+      "loss": 0.3768,
+      "step": 541
+    },
+    {
+      "epoch": 2.546203649205415,
+      "grad_norm": 0.3569755256175995,
+      "learning_rate": 3.304624687607041e-06,
+      "loss": 0.372,
+      "step": 542
+    },
+    {
+      "epoch": 2.5509123013537374,
+      "grad_norm": 0.3697240352630615,
+      "learning_rate": 3.298466280756295e-06,
+      "loss": 0.3829,
+      "step": 543
+    },
+    {
+      "epoch": 2.55562095350206,
+      "grad_norm": 0.3861812353134155,
+      "learning_rate": 3.2923024735517575e-06,
+      "loss": 0.3752,
+      "step": 544
+    },
+    {
+      "epoch": 2.5603296056503826,
+      "grad_norm": 0.40910521149635315,
+      "learning_rate": 3.286133307681773e-06,
+      "loss": 0.3814,
+      "step": 545
+    },
+    {
+      "epoch": 2.565038257798705,
+      "grad_norm": 0.3837888538837433,
+      "learning_rate": 3.2799588248709346e-06,
+      "loss": 0.3923,
+      "step": 546
+    },
+    {
+      "epoch": 2.569746909947028,
+      "grad_norm": 0.39909642934799194,
+      "learning_rate": 3.2737790668797907e-06,
+      "loss": 0.3834,
+      "step": 547
+    },
+    {
+      "epoch": 2.57445556209535,
+      "grad_norm": 0.3567914664745331,
+      "learning_rate": 3.2675940755045717e-06,
+      "loss": 0.354,
+      "step": 548
+    },
+    {
+      "epoch": 2.5791642142436726,
+      "grad_norm": 0.3389582633972168,
+      "learning_rate": 3.261403892576901e-06,
+      "loss": 0.3763,
+      "step": 549
+    },
+    {
+      "epoch": 2.5838728663919954,
+      "grad_norm": 0.41281071305274963,
+      "learning_rate": 3.255208559963517e-06,
+      "loss": 0.3846,
+      "step": 550
+    },
+    {
+      "epoch": 2.588581518540318,
+      "grad_norm": 0.40536680817604065,
+      "learning_rate": 3.2490081195659858e-06,
+      "loss": 0.3766,
+      "step": 551
+    },
+    {
+      "epoch": 2.5932901706886407,
+      "grad_norm": 0.3592185080051422,
+      "learning_rate": 3.2428026133204187e-06,
+      "loss": 0.3508,
+      "step": 552
+    },
+    {
+      "epoch": 2.597998822836963,
+      "grad_norm": 0.3820286691188812,
+      "learning_rate": 3.2365920831971925e-06,
+      "loss": 0.3753,
+      "step": 553
+    },
+    {
+      "epoch": 2.6027074749852854,
+      "grad_norm": 0.36180534958839417,
+      "learning_rate": 3.230376571200659e-06,
+      "loss": 0.3611,
+      "step": 554
+    },
+    {
+      "epoch": 2.6074161271336083,
+      "grad_norm": 0.38467833399772644,
+      "learning_rate": 3.2241561193688673e-06,
+      "loss": 0.3751,
+      "step": 555
+    },
+    {
+      "epoch": 2.6121247792819307,
+      "grad_norm": 0.3356960117816925,
+      "learning_rate": 3.2179307697732753e-06,
+      "loss": 0.3549,
+      "step": 556
+    },
+    {
+      "epoch": 2.616833431430253,
+      "grad_norm": 0.367064893245697,
+      "learning_rate": 3.2117005645184668e-06,
+      "loss": 0.4021,
+      "step": 557
+    },
+    {
+      "epoch": 2.621542083578576,
+      "grad_norm": 0.39036962389945984,
+      "learning_rate": 3.205465545741865e-06,
+      "loss": 0.3493,
+      "step": 558
+    },
+    {
+      "epoch": 2.6262507357268983,
+      "grad_norm": 0.34206390380859375,
+      "learning_rate": 3.199225755613452e-06,
+      "loss": 0.3913,
+      "step": 559
+    },
+    {
+      "epoch": 2.6309593878752207,
+      "grad_norm": 0.3853800296783447,
+      "learning_rate": 3.1929812363354766e-06,
+      "loss": 0.3516,
+      "step": 560
+    },
+    {
+      "epoch": 2.6356680400235435,
+      "grad_norm": 0.3480147421360016,
+      "learning_rate": 3.186732030142177e-06,
+      "loss": 0.3711,
+      "step": 561
+    },
+    {
+      "epoch": 2.640376692171866,
+      "grad_norm": 0.3732787072658539,
+      "learning_rate": 3.180478179299487e-06,
+      "loss": 0.3781,
+      "step": 562
+    },
+    {
+      "epoch": 2.6450853443201883,
+      "grad_norm": 0.3606315851211548,
+      "learning_rate": 3.174219726104757e-06,
+      "loss": 0.3601,
+      "step": 563
+    },
+    {
+      "epoch": 2.649793996468511,
+      "grad_norm": 0.35798346996307373,
+      "learning_rate": 3.167956712886463e-06,
+      "loss": 0.3765,
+      "step": 564
+    },
+    {
+      "epoch": 2.6545026486168335,
+      "grad_norm": 0.3613319396972656,
+      "learning_rate": 3.161689182003924e-06,
+      "loss": 0.3635,
+      "step": 565
+    },
+    {
+      "epoch": 2.659211300765156,
+      "grad_norm": 0.34538811445236206,
+      "learning_rate": 3.1554171758470114e-06,
+      "loss": 0.3575,
+      "step": 566
+    },
+    {
+      "epoch": 2.6639199529134787,
+      "grad_norm": 0.3859974443912506,
+      "learning_rate": 3.1491407368358682e-06,
+      "loss": 0.3805,
+      "step": 567
+    },
+    {
+      "epoch": 2.668628605061801,
+      "grad_norm": 0.364109605550766,
+      "learning_rate": 3.142859907420615e-06,
+      "loss": 0.3682,
+      "step": 568
+    },
+    {
+      "epoch": 2.6733372572101235,
+      "grad_norm": 0.36927446722984314,
+      "learning_rate": 3.1365747300810693e-06,
+      "loss": 0.3737,
+      "step": 569
+    },
+    {
+      "epoch": 2.6780459093584463,
+      "grad_norm": 0.38519367575645447,
+      "learning_rate": 3.130285247326454e-06,
+      "loss": 0.3788,
+      "step": 570
+    },
+    {
+      "epoch": 2.6827545615067687,
+      "grad_norm": 0.36593741178512573,
+      "learning_rate": 3.12399150169511e-06,
+      "loss": 0.3744,
+      "step": 571
+    },
+    {
+      "epoch": 2.687463213655091,
+      "grad_norm": 0.3665541112422943,
+      "learning_rate": 3.1176935357542136e-06,
+      "loss": 0.3641,
+      "step": 572
+    },
+    {
+      "epoch": 2.692171865803414,
+      "grad_norm": 0.35215452313423157,
+      "learning_rate": 3.1113913920994803e-06,
+      "loss": 0.3428,
+      "step": 573
+    },
+    {
+      "epoch": 2.6968805179517363,
+      "grad_norm": 0.3475217819213867,
+      "learning_rate": 3.1050851133548854e-06,
+      "loss": 0.3541,
+      "step": 574
+    },
+    {
+      "epoch": 2.7015891701000587,
+      "grad_norm": 0.35495033860206604,
+      "learning_rate": 3.0987747421723678e-06,
+      "loss": 0.3647,
+      "step": 575
+    },
+    {
+      "epoch": 2.7062978222483816,
+      "grad_norm": 0.3700904846191406,
+      "learning_rate": 3.0924603212315475e-06,
+      "loss": 0.3758,
+      "step": 576
+    },
+    {
+      "epoch": 2.711006474396704,
+      "grad_norm": 0.3657042682170868,
+      "learning_rate": 3.0861418932394344e-06,
+      "loss": 0.3644,
+      "step": 577
+    },
+    {
+      "epoch": 2.7157151265450263,
+      "grad_norm": 0.350494384765625,
+      "learning_rate": 3.0798195009301385e-06,
+      "loss": 0.3724,
+      "step": 578
+    },
+    {
+      "epoch": 2.720423778693349,
+      "grad_norm": 0.3735278248786926,
+      "learning_rate": 3.0734931870645844e-06,
+      "loss": 0.3653,
+      "step": 579
+    },
+    {
+      "epoch": 2.7251324308416716,
+      "grad_norm": 0.35524168610572815,
+      "learning_rate": 3.0671629944302168e-06,
+      "loss": 0.3526,
+      "step": 580
+    },
+    {
+      "epoch": 2.729841082989994,
+      "grad_norm": 0.35376468300819397,
+      "learning_rate": 3.0608289658407165e-06,
+      "loss": 0.3601,
+      "step": 581
+    },
+    {
+      "epoch": 2.734549735138317,
+      "grad_norm": 0.3474513292312622,
+      "learning_rate": 3.054491144135707e-06,
+      "loss": 0.3796,
+      "step": 582
+    },
+    {
+      "epoch": 2.739258387286639,
+      "grad_norm": 0.3583899438381195,
+      "learning_rate": 3.0481495721804683e-06,
+      "loss": 0.3577,
+      "step": 583
+    },
+    {
+      "epoch": 2.7439670394349616,
+      "grad_norm": 0.34981340169906616,
+      "learning_rate": 3.0418042928656416e-06,
+      "loss": 0.3727,
+      "step": 584
+    },
+    {
+      "epoch": 2.7486756915832844,
+      "grad_norm": 0.34836986660957336,
+      "learning_rate": 3.035455349106945e-06,
+      "loss": 0.3737,
+      "step": 585
+    },
+    {
+      "epoch": 2.753384343731607,
+      "grad_norm": 0.3657343089580536,
+      "learning_rate": 3.0291027838448793e-06,
+      "loss": 0.3694,
+      "step": 586
+    },
+    {
+      "epoch": 2.758092995879929,
+      "grad_norm": 0.36264339089393616,
+      "learning_rate": 3.022746640044441e-06,
+      "loss": 0.3314,
+      "step": 587
+    },
+    {
+      "epoch": 2.762801648028252,
+      "grad_norm": 0.32704487442970276,
+      "learning_rate": 3.0163869606948276e-06,
+      "loss": 0.3826,
+      "step": 588
+    },
+    {
+      "epoch": 2.7675103001765744,
+      "grad_norm": 0.37806132435798645,
+      "learning_rate": 3.0100237888091493e-06,
+      "loss": 0.3816,
+      "step": 589
+    },
+    {
+      "epoch": 2.772218952324897,
+      "grad_norm": 0.3723447620868683,
+      "learning_rate": 3.0036571674241393e-06,
+      "loss": 0.3776,
+      "step": 590
+    },
+    {
+      "epoch": 2.7769276044732196,
+      "grad_norm": 0.3555864691734314,
+      "learning_rate": 2.997287139599859e-06,
+      "loss": 0.3956,
+      "step": 591
+    },
+    {
+      "epoch": 2.781636256621542,
+      "grad_norm": 0.3734078109264374,
+      "learning_rate": 2.9909137484194113e-06,
+      "loss": 0.3913,
+      "step": 592
+    },
+    {
+      "epoch": 2.7863449087698644,
+      "grad_norm": 0.3938598930835724,
+      "learning_rate": 2.984537036988644e-06,
+      "loss": 0.377,
+      "step": 593
+    },
+    {
+      "epoch": 2.7910535609181872,
+      "grad_norm": 0.4086543321609497,
+      "learning_rate": 2.9781570484358634e-06,
+      "loss": 0.3846,
+      "step": 594
+    },
+    {
+      "epoch": 2.7957622130665096,
+      "grad_norm": 0.3646213710308075,
+      "learning_rate": 2.9717738259115403e-06,
+      "loss": 0.3559,
+      "step": 595
+    },
+    {
+      "epoch": 2.800470865214832,
+      "grad_norm": 0.35105398297309875,
+      "learning_rate": 2.965387412588017e-06,
+      "loss": 0.3567,
+      "step": 596
+    },
+    {
+      "epoch": 2.805179517363155,
+      "grad_norm": 0.36393171548843384,
+      "learning_rate": 2.9589978516592164e-06,
+      "loss": 0.3653,
+      "step": 597
+    },
+    {
+      "epoch": 2.8098881695114772,
+      "grad_norm": 0.34818121790885925,
+      "learning_rate": 2.952605186340352e-06,
+      "loss": 0.3786,
+      "step": 598
+    },
+    {
+      "epoch": 2.8145968216597996,
+      "grad_norm": 0.4391355812549591,
+      "learning_rate": 2.9462094598676298e-06,
+      "loss": 0.3702,
+      "step": 599
+    },
+    {
+      "epoch": 2.8193054738081225,
+      "grad_norm": 0.37328672409057617,
+      "learning_rate": 2.9398107154979637e-06,
+      "loss": 0.3575,
+      "step": 600
+    },
+    {
+      "epoch": 2.824014125956445,
+      "grad_norm": 0.3919398784637451,
+      "learning_rate": 2.9334089965086765e-06,
+      "loss": 0.3779,
+      "step": 601
+    },
+    {
+      "epoch": 2.8287227781047672,
+      "grad_norm": 0.3888506293296814,
+      "learning_rate": 2.9270043461972104e-06,
+      "loss": 0.3684,
+      "step": 602
+    },
+    {
+      "epoch": 2.83343143025309,
+      "grad_norm": 0.3589399456977844,
+      "learning_rate": 2.920596807880832e-06,
+      "loss": 0.3716,
+      "step": 603
+    },
+    {
+      "epoch": 2.8381400824014125,
+      "grad_norm": 0.37757810950279236,
+      "learning_rate": 2.9141864248963433e-06,
+      "loss": 0.3924,
+      "step": 604
+    },
+    {
+      "epoch": 2.8428487345497353,
+      "grad_norm": 0.37059202790260315,
+      "learning_rate": 2.9077732405997834e-06,
+      "loss": 0.3805,
+      "step": 605
+    },
+    {
+      "epoch": 2.8475573866980577,
+      "grad_norm": 0.38783740997314453,
+      "learning_rate": 2.9013572983661382e-06,
+      "loss": 0.3586,
+      "step": 606
+    },
+    {
+      "epoch": 2.85226603884638,
+      "grad_norm": 0.3548484146595001,
+      "learning_rate": 2.8949386415890486e-06,
+      "loss": 0.3559,
+      "step": 607
+    },
+    {
+      "epoch": 2.856974690994703,
+      "grad_norm": 0.36312374472618103,
+      "learning_rate": 2.8885173136805126e-06,
+      "loss": 0.3662,
+      "step": 608
+    },
+    {
+      "epoch": 2.8616833431430253,
+      "grad_norm": 0.3536983132362366,
+      "learning_rate": 2.882093358070597e-06,
+      "loss": 0.3674,
+      "step": 609
+    },
+    {
+      "epoch": 2.8663919952913477,
+      "grad_norm": 0.3506225347518921,
+      "learning_rate": 2.8756668182071364e-06,
+      "loss": 0.3916,
+      "step": 610
+    },
+    {
+      "epoch": 2.8711006474396705,
+      "grad_norm": 0.37823688983917236,
+      "learning_rate": 2.869237737555448e-06,
+      "loss": 0.3695,
+      "step": 611
+    },
+    {
+      "epoch": 2.875809299587993,
+      "grad_norm": 0.3676494061946869,
+      "learning_rate": 2.8628061595980323e-06,
+      "loss": 0.376,
+      "step": 612
+    },
+    {
+      "epoch": 2.8805179517363158,
+      "grad_norm": 0.3870524764060974,
+      "learning_rate": 2.856372127834279e-06,
+      "loss": 0.3725,
+      "step": 613
+    },
+    {
+      "epoch": 2.885226603884638,
+      "grad_norm": 0.4425090253353119,
+      "learning_rate": 2.8499356857801747e-06,
+      "loss": 0.3688,
+      "step": 614
+    },
+    {
+      "epoch": 2.8899352560329605,
+      "grad_norm": 0.3530326187610626,
+      "learning_rate": 2.843496876968007e-06,
+      "loss": 0.3619,
+      "step": 615
+    },
+    {
+      "epoch": 2.8946439081812834,
+      "grad_norm": 0.3702718913555145,
+      "learning_rate": 2.837055744946072e-06,
+      "loss": 0.3791,
+      "step": 616
+    },
+    {
+      "epoch": 2.8993525603296058,
+      "grad_norm": 0.37180230021476746,
+      "learning_rate": 2.8306123332783793e-06,
+      "loss": 0.3692,
+      "step": 617
+    },
+    {
+      "epoch": 2.904061212477928,
+      "grad_norm": 0.3533918559551239,
+      "learning_rate": 2.8241666855443527e-06,
+      "loss": 0.3567,
+      "step": 618
+    },
+    {
+      "epoch": 2.908769864626251,
+      "grad_norm": 0.3372352719306946,
+      "learning_rate": 2.8177188453385445e-06,
+      "loss": 0.3761,
+      "step": 619
+    },
+    {
+      "epoch": 2.9134785167745734,
+      "grad_norm": 0.3824419379234314,
+      "learning_rate": 2.811268856270332e-06,
+      "loss": 0.3925,
+      "step": 620
+    },
+    {
+      "epoch": 2.9181871689228958,
+      "grad_norm": 0.36836275458335876,
+      "learning_rate": 2.8048167619636287e-06,
+      "loss": 0.3481,
+      "step": 621
+    },
+    {
+      "epoch": 2.9228958210712186,
+      "grad_norm": 0.343496173620224,
+      "learning_rate": 2.7983626060565833e-06,
+      "loss": 0.3725,
+      "step": 622
+    },
+    {
+      "epoch": 2.927604473219541,
+      "grad_norm": 0.3810299336910248,
+      "learning_rate": 2.7919064322012922e-06,
+      "loss": 0.347,
+      "step": 623
+    },
+    {
+      "epoch": 2.9323131253678634,
+      "grad_norm": 0.3711015284061432,
+      "learning_rate": 2.785448284063497e-06,
+      "loss": 0.3632,
+      "step": 624
+    },
+    {
+      "epoch": 2.937021777516186,
+      "grad_norm": 0.3700212836265564,
+      "learning_rate": 2.778988205322292e-06,
+      "loss": 0.3812,
+      "step": 625
+    },
+    {
+      "epoch": 2.9417304296645086,
+      "grad_norm": 0.44081220030784607,
+      "learning_rate": 2.772526239669831e-06,
+      "loss": 0.3225,
+      "step": 626
+    },
+    {
+      "epoch": 2.946439081812831,
+      "grad_norm": 0.314218670129776,
+      "learning_rate": 2.7660624308110293e-06,
+      "loss": 0.3503,
+      "step": 627
+    },
+    {
+      "epoch": 2.951147733961154,
+      "grad_norm": 0.35966551303863525,
+      "learning_rate": 2.7595968224632672e-06,
+      "loss": 0.4044,
+      "step": 628
+    },
+    {
+      "epoch": 2.955856386109476,
+      "grad_norm": 0.4296911656856537,
+      "learning_rate": 2.7531294583560968e-06,
+      "loss": 0.3862,
+      "step": 629
+    },
+    {
+      "epoch": 2.9605650382577986,
+      "grad_norm": 0.3642086088657379,
+      "learning_rate": 2.746660382230944e-06,
+      "loss": 0.3549,
+      "step": 630
+    },
+    {
+      "epoch": 2.9652736904061214,
+      "grad_norm": 0.3510330021381378,
+      "learning_rate": 2.740189637840816e-06,
+      "loss": 0.3491,
+      "step": 631
+    },
+    {
+      "epoch": 2.969982342554444,
+      "grad_norm": 0.34799647331237793,
+      "learning_rate": 2.73371726895e-06,
+      "loss": 0.3785,
+      "step": 632
+    },
+    {
+      "epoch": 2.974690994702766,
+      "grad_norm": 0.3923639953136444,
+      "learning_rate": 2.727243319333774e-06,
+      "loss": 0.3686,
+      "step": 633
+    },
+    {
+      "epoch": 2.979399646851089,
+      "grad_norm": 0.3745047152042389,
+      "learning_rate": 2.720767832778104e-06,
+      "loss": 0.3814,
+      "step": 634
+    },
+    {
+      "epoch": 2.9841082989994114,
+      "grad_norm": 0.3802978992462158,
+      "learning_rate": 2.7142908530793523e-06,
+      "loss": 0.3692,
+      "step": 635
+    },
+    {
+      "epoch": 2.988816951147734,
+      "grad_norm": 0.38678938150405884,
+      "learning_rate": 2.7078124240439795e-06,
+      "loss": 0.3703,
+      "step": 636
+    },
+    {
+      "epoch": 2.9935256032960567,
+      "grad_norm": 0.3405422866344452,
+      "learning_rate": 2.701332589488249e-06,
+      "loss": 0.3778,
+      "step": 637
+    },
+    {
+      "epoch": 2.998234255444379,
+      "grad_norm": 0.36346471309661865,
+      "learning_rate": 2.694851393237931e-06,
+      "loss": 0.3601,
+      "step": 638
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.5746592879295349,
+      "learning_rate": 2.688368879128004e-06,
+      "loss": 0.3771,
+      "step": 639
+    },
+    {
+      "epoch": 3.0047086521483224,
+      "grad_norm": 0.49112632870674133,
+      "learning_rate": 2.68188509100236e-06,
+      "loss": 0.332,
+      "step": 640
+    },
+    {
+      "epoch": 3.0094173042966452,
+      "grad_norm": 0.3928791284561157,
+      "learning_rate": 2.675400072713508e-06,
+      "loss": 0.3545,
+      "step": 641
+    },
+    {
+      "epoch": 3.0141259564449676,
+      "grad_norm": 0.36599087715148926,
+      "learning_rate": 2.6689138681222793e-06,
+      "loss": 0.3242,
+      "step": 642
+    },
+    {
+      "epoch": 3.01883460859329,
+      "grad_norm": 0.3617115914821625,
+      "learning_rate": 2.6624265210975232e-06,
+      "loss": 0.3175,
+      "step": 643
+    },
+    {
+      "epoch": 3.023543260741613,
+      "grad_norm": 0.39354369044303894,
+      "learning_rate": 2.655938075515821e-06,
+      "loss": 0.3444,
+      "step": 644
+    },
+    {
+      "epoch": 3.0282519128899352,
+      "grad_norm": 0.40006500482559204,
+      "learning_rate": 2.649448575261182e-06,
+      "loss": 0.3396,
+      "step": 645
+    },
+    {
+      "epoch": 3.0329605650382576,
+      "grad_norm": 0.37696173787117004,
+      "learning_rate": 2.6429580642247478e-06,
+      "loss": 0.3492,
+      "step": 646
+    },
+    {
+      "epoch": 3.0376692171865805,
+      "grad_norm": 0.4124061167240143,
+      "learning_rate": 2.6364665863044984e-06,
+      "loss": 0.3512,
+      "step": 647
+    },
+    {
+      "epoch": 3.042377869334903,
+      "grad_norm": 0.3561529219150543,
+      "learning_rate": 2.629974185404951e-06,
+      "loss": 0.3312,
+      "step": 648
+    },
+    {
+      "epoch": 3.0470865214832252,
+      "grad_norm": 0.4442432224750519,
+      "learning_rate": 2.623480905436867e-06,
+      "loss": 0.3456,
+      "step": 649
+    },
+    {
+      "epoch": 3.051795173631548,
+      "grad_norm": 0.359831839799881,
+      "learning_rate": 2.6169867903169528e-06,
+      "loss": 0.3253,
+      "step": 650
+    },
+    {
+      "epoch": 3.0565038257798705,
+      "grad_norm": 0.424878865480423,
+      "learning_rate": 2.610491883967562e-06,
+      "loss": 0.3518,
+      "step": 651
+    },
+    {
+      "epoch": 3.061212477928193,
+      "grad_norm": 0.37126424908638,
+      "learning_rate": 2.6039962303164024e-06,
+      "loss": 0.3327,
+      "step": 652
+    },
+    {
+      "epoch": 3.0659211300765157,
+      "grad_norm": 0.394135445356369,
+      "learning_rate": 2.5974998732962324e-06,
+      "loss": 0.328,
+      "step": 653
+    },
+    {
+      "epoch": 3.070629782224838,
+      "grad_norm": 0.3592108488082886,
+      "learning_rate": 2.5910028568445724e-06,
+      "loss": 0.3294,
+      "step": 654
+    },
+    {
+      "epoch": 3.075338434373161,
+      "grad_norm": 0.3625464141368866,
+      "learning_rate": 2.584505224903397e-06,
+      "loss": 0.3415,
+      "step": 655
+    },
+    {
+      "epoch": 3.0800470865214833,
+      "grad_norm": 0.3701664209365845,
+      "learning_rate": 2.578007021418848e-06,
+      "loss": 0.3416,
+      "step": 656
+    },
+    {
+      "epoch": 3.0847557386698057,
+      "grad_norm": 0.35726526379585266,
+      "learning_rate": 2.571508290340932e-06,
+      "loss": 0.3283,
+      "step": 657
+    },
+    {
+      "epoch": 3.0894643908181285,
+      "grad_norm": 0.35142824053764343,
+      "learning_rate": 2.5650090756232227e-06,
+      "loss": 0.3253,
+      "step": 658
+    },
+    {
+      "epoch": 3.094173042966451,
+      "grad_norm": 0.3839946985244751,
+      "learning_rate": 2.5585094212225646e-06,
+      "loss": 0.3233,
+      "step": 659
+    },
+    {
+      "epoch": 3.0988816951147733,
+      "grad_norm": 0.3479538857936859,
+      "learning_rate": 2.5520093710987785e-06,
+      "loss": 0.3306,
+      "step": 660
+    },
+    {
+      "epoch": 3.103590347263096,
+      "grad_norm": 0.36447158455848694,
+      "learning_rate": 2.5455089692143593e-06,
+      "loss": 0.3431,
+      "step": 661
+    },
+    {
+      "epoch": 3.1082989994114185,
+      "grad_norm": 0.3611982762813568,
+      "learning_rate": 2.539008259534182e-06,
+      "loss": 0.3345,
+      "step": 662
+    },
+    {
+      "epoch": 3.113007651559741,
+      "grad_norm": 0.35439133644104004,
+      "learning_rate": 2.5325072860252016e-06,
+      "loss": 0.3425,
+      "step": 663
+    },
+    {
+      "epoch": 3.1177163037080637,
+      "grad_norm": 0.3768078684806824,
+      "learning_rate": 2.526006092656161e-06,
+      "loss": 0.3444,
+      "step": 664
+    },
+    {
+      "epoch": 3.122424955856386,
+      "grad_norm": 0.38114500045776367,
+      "learning_rate": 2.5195047233972856e-06,
+      "loss": 0.349,
+      "step": 665
+    },
+    {
+      "epoch": 3.1271336080047085,
+      "grad_norm": 0.4064490497112274,
+      "learning_rate": 2.5130032222199956e-06,
+      "loss": 0.3698,
+      "step": 666
+    },
+    {
+      "epoch": 3.1318422601530314,
+      "grad_norm": 0.35883423686027527,
+      "learning_rate": 2.5065016330965985e-06,
+      "loss": 0.3194,
+      "step": 667
+    },
+    {
+      "epoch": 3.1365509123013537,
+      "grad_norm": 0.35054928064346313,
+      "learning_rate": 2.5e-06,
+      "loss": 0.328,
+      "step": 668
+    },
+    {
+      "epoch": 3.141259564449676,
+      "grad_norm": 0.34279653429985046,
+      "learning_rate": 2.493498366903402e-06,
+      "loss": 0.3407,
+      "step": 669
+    },
+    {
+      "epoch": 3.145968216597999,
+      "grad_norm": 0.37415286898612976,
+      "learning_rate": 2.4869967777800057e-06,
+      "loss": 0.3459,
+      "step": 670
+    },
+    {
+      "epoch": 3.1506768687463214,
+      "grad_norm": 0.3572427034378052,
+      "learning_rate": 2.480495276602714e-06,
+      "loss": 0.3294,
+      "step": 671
+    },
+    {
+      "epoch": 3.1553855208946437,
+      "grad_norm": 0.3583246171474457,
+      "learning_rate": 2.47399390734384e-06,
+      "loss": 0.3342,
+      "step": 672
+    },
+    {
+      "epoch": 3.1600941730429666,
+      "grad_norm": 0.35233941674232483,
+      "learning_rate": 2.467492713974799e-06,
+      "loss": 0.3351,
+      "step": 673
+    },
+    {
+      "epoch": 3.164802825191289,
+      "grad_norm": 0.36822763085365295,
+      "learning_rate": 2.4609917404658198e-06,
+      "loss": 0.3422,
+      "step": 674
+    },
+    {
+      "epoch": 3.1695114773396114,
+      "grad_norm": 0.35386916995048523,
+      "learning_rate": 2.4544910307856415e-06,
+      "loss": 0.3372,
+      "step": 675
+    },
+    {
+      "epoch": 3.174220129487934,
+      "grad_norm": 0.3638751208782196,
+      "learning_rate": 2.447990628901222e-06,
+      "loss": 0.3344,
+      "step": 676
+    },
+    {
+      "epoch": 3.1789287816362566,
+      "grad_norm": 0.3515467047691345,
+      "learning_rate": 2.441490578777436e-06,
+      "loss": 0.3302,
+      "step": 677
+    },
+    {
+      "epoch": 3.183637433784579,
+      "grad_norm": 0.3626198172569275,
+      "learning_rate": 2.434990924376778e-06,
+      "loss": 0.3488,
+      "step": 678
+    },
+    {
+      "epoch": 3.188346085932902,
+      "grad_norm": 0.3596525490283966,
+      "learning_rate": 2.428491709659069e-06,
+      "loss": 0.3231,
+      "step": 679
+    },
+    {
+      "epoch": 3.193054738081224,
+      "grad_norm": 0.37080711126327515,
+      "learning_rate": 2.421992978581152e-06,
+      "loss": 0.3353,
+      "step": 680
+    },
+    {
+      "epoch": 3.1977633902295466,
+      "grad_norm": 0.3591707944869995,
+      "learning_rate": 2.4154947750966035e-06,
+      "loss": 0.34,
+      "step": 681
+    },
+    {
+      "epoch": 3.2024720423778694,
+      "grad_norm": 0.3675726354122162,
+      "learning_rate": 2.408997143155429e-06,
+      "loss": 0.3358,
+      "step": 682
+    },
+    {
+      "epoch": 3.207180694526192,
+      "grad_norm": 0.3385520875453949,
+      "learning_rate": 2.402500126703768e-06,
+      "loss": 0.3151,
+      "step": 683
+    },
+    {
+      "epoch": 3.211889346674514,
+      "grad_norm": 0.38244402408599854,
+      "learning_rate": 2.396003769683599e-06,
+      "loss": 0.3489,
+      "step": 684
+    },
+    {
+      "epoch": 3.216597998822837,
+      "grad_norm": 0.35723546147346497,
+      "learning_rate": 2.389508116032438e-06,
+      "loss": 0.3407,
+      "step": 685
+    },
+    {
+      "epoch": 3.2213066509711594,
+      "grad_norm": 0.3416253328323364,
+      "learning_rate": 2.3830132096830476e-06,
+      "loss": 0.3124,
+      "step": 686
+    },
+    {
+      "epoch": 3.226015303119482,
+      "grad_norm": 0.3942345678806305,
+      "learning_rate": 2.376519094563134e-06,
+      "loss": 0.3166,
+      "step": 687
+    },
+    {
+      "epoch": 3.2307239552678046,
+      "grad_norm": 0.35811132192611694,
+      "learning_rate": 2.3700258145950495e-06,
+      "loss": 0.3317,
+      "step": 688
+    },
+    {
+      "epoch": 3.235432607416127,
+      "grad_norm": 0.35040491819381714,
+      "learning_rate": 2.363533413695503e-06,
+      "loss": 0.318,
+      "step": 689
+    },
+    {
+      "epoch": 3.24014125956445,
+      "grad_norm": 0.35471948981285095,
+      "learning_rate": 2.3570419357752522e-06,
+      "loss": 0.3278,
+      "step": 690
+    },
+    {
+      "epoch": 3.2448499117127723,
+      "grad_norm": 0.35799640417099,
+      "learning_rate": 2.3505514247388185e-06,
+      "loss": 0.3304,
+      "step": 691
+    },
+    {
+      "epoch": 3.2495585638610947,
+      "grad_norm": 0.3335202634334564,
+      "learning_rate": 2.3440619244841798e-06,
+      "loss": 0.3171,
+      "step": 692
+    },
+    {
+      "epoch": 3.2542672160094175,
+      "grad_norm": 0.42852750420570374,
+      "learning_rate": 2.3375734789024776e-06,
+      "loss": 0.332,
+      "step": 693
+    },
+    {
+      "epoch": 3.25897586815774,
+      "grad_norm": 0.37598463892936707,
+      "learning_rate": 2.3310861318777216e-06,
+      "loss": 0.3563,
+      "step": 694
+    },
+    {
+      "epoch": 3.2636845203060623,
+      "grad_norm": 0.3953750729560852,
+      "learning_rate": 2.3245999272864924e-06,
+      "loss": 0.3267,
+      "step": 695
+    },
+    {
+      "epoch": 3.268393172454385,
+      "grad_norm": 0.3668225407600403,
+      "learning_rate": 2.3181149089976407e-06,
+      "loss": 0.3204,
+      "step": 696
+    },
+    {
+      "epoch": 3.2731018246027075,
+      "grad_norm": 0.3694913387298584,
+      "learning_rate": 2.311631120871997e-06,
+      "loss": 0.3472,
+      "step": 697
+    },
+    {
+      "epoch": 3.27781047675103,
+      "grad_norm": 0.35975244641304016,
+      "learning_rate": 2.30514860676207e-06,
+      "loss": 0.339,
+      "step": 698
+    },
+    {
+      "epoch": 3.2825191288993527,
+      "grad_norm": 0.3759594261646271,
+      "learning_rate": 2.298667410511751e-06,
+      "loss": 0.3412,
+      "step": 699
+    },
+    {
+      "epoch": 3.287227781047675,
+      "grad_norm": 0.37436795234680176,
+      "learning_rate": 2.2921875759560213e-06,
+      "loss": 0.3279,
+      "step": 700
+    },
+    {
+      "epoch": 3.2919364331959975,
+      "grad_norm": 0.3595412075519562,
+      "learning_rate": 2.2857091469206485e-06,
+      "loss": 0.3427,
+      "step": 701
+    },
+    {
+      "epoch": 3.2966450853443203,
+      "grad_norm": 0.3599635660648346,
+      "learning_rate": 2.279232167221897e-06,
+      "loss": 0.3353,
+      "step": 702
+    },
+    {
+      "epoch": 3.3013537374926427,
+      "grad_norm": 0.35342299938201904,
+      "learning_rate": 2.272756680666227e-06,
+      "loss": 0.3212,
+      "step": 703
+    },
+    {
+      "epoch": 3.306062389640965,
+      "grad_norm": 0.37875843048095703,
+      "learning_rate": 2.2662827310499995e-06,
+      "loss": 0.3215,
+      "step": 704
+    },
+    {
+      "epoch": 3.310771041789288,
+      "grad_norm": 0.3415927290916443,
+      "learning_rate": 2.2598103621591846e-06,
+      "loss": 0.3317,
+      "step": 705
+    },
+    {
+      "epoch": 3.3154796939376103,
+      "grad_norm": 0.38792863488197327,
+      "learning_rate": 2.2533396177690565e-06,
+      "loss": 0.3326,
+      "step": 706
+    },
+    {
+      "epoch": 3.3201883460859327,
+      "grad_norm": 0.3674851655960083,
+      "learning_rate": 2.2468705416439045e-06,
+      "loss": 0.3306,
+      "step": 707
+    },
+    {
+      "epoch": 3.3248969982342556,
+      "grad_norm": 0.3733600974082947,
+      "learning_rate": 2.2404031775367336e-06,
+      "loss": 0.3438,
+      "step": 708
+    },
+    {
+      "epoch": 3.329605650382578,
+      "grad_norm": 0.37565818428993225,
+      "learning_rate": 2.233937569188971e-06,
+      "loss": 0.3412,
+      "step": 709
+    },
+    {
+      "epoch": 3.3343143025309008,
+      "grad_norm": 0.3375566303730011,
+      "learning_rate": 2.2274737603301696e-06,
+      "loss": 0.32,
+      "step": 710
+    },
+    {
+      "epoch": 3.339022954679223,
+      "grad_norm": 0.3649482727050781,
+      "learning_rate": 2.2210117946777087e-06,
+      "loss": 0.3354,
+      "step": 711
+    },
+    {
+      "epoch": 3.3437316068275456,
+      "grad_norm": 0.3563418686389923,
+      "learning_rate": 2.2145517159365044e-06,
+      "loss": 0.3369,
+      "step": 712
+    },
+    {
+      "epoch": 3.3484402589758684,
+      "grad_norm": 0.377107173204422,
+      "learning_rate": 2.2080935677987086e-06,
+      "loss": 0.3392,
+      "step": 713
+    },
+    {
+      "epoch": 3.353148911124191,
+      "grad_norm": 0.4007817804813385,
+      "learning_rate": 2.2016373939434166e-06,
+      "loss": 0.3316,
+      "step": 714
+    },
+    {
+      "epoch": 3.357857563272513,
+      "grad_norm": 0.359765887260437,
+      "learning_rate": 2.1951832380363726e-06,
+      "loss": 0.3307,
+      "step": 715
+    },
+    {
+      "epoch": 3.362566215420836,
+      "grad_norm": 0.33300676941871643,
+      "learning_rate": 2.188731143729669e-06,
+      "loss": 0.32,
+      "step": 716
+    },
+    {
+      "epoch": 3.3672748675691584,
+      "grad_norm": 0.36447378993034363,
+      "learning_rate": 2.182281154661457e-06,
+      "loss": 0.3478,
+      "step": 717
+    },
+    {
+      "epoch": 3.371983519717481,
+      "grad_norm": 0.33058932423591614,
+      "learning_rate": 2.1758333144556473e-06,
+      "loss": 0.2869,
+      "step": 718
+    },
+    {
+      "epoch": 3.3766921718658036,
+      "grad_norm": 0.363322913646698,
+      "learning_rate": 2.1693876667216215e-06,
+      "loss": 0.3304,
+      "step": 719
+    },
+    {
+      "epoch": 3.381400824014126,
+      "grad_norm": 0.34668809175491333,
+      "learning_rate": 2.1629442550539283e-06,
+      "loss": 0.3273,
+      "step": 720
+    },
+    {
+      "epoch": 3.3861094761624484,
+      "grad_norm": 0.3586620092391968,
+      "learning_rate": 2.1565031230319937e-06,
+      "loss": 0.3431,
+      "step": 721
+    },
+    {
+      "epoch": 3.3908181283107712,
+      "grad_norm": 0.33611762523651123,
+      "learning_rate": 2.1500643142198265e-06,
+      "loss": 0.3263,
+      "step": 722
+    },
+    {
+      "epoch": 3.3955267804590936,
+      "grad_norm": 0.3559138774871826,
+      "learning_rate": 2.1436278721657216e-06,
+      "loss": 0.3299,
+      "step": 723
+    },
+    {
+      "epoch": 3.400235432607416,
+      "grad_norm": 0.38343891501426697,
+      "learning_rate": 2.137193840401968e-06,
+      "loss": 0.3226,
+      "step": 724
+    },
+    {
+      "epoch": 3.404944084755739,
+      "grad_norm": 0.3612920045852661,
+      "learning_rate": 2.130762262444553e-06,
+      "loss": 0.3481,
+      "step": 725
+    },
+    {
+      "epoch": 3.4096527369040612,
+      "grad_norm": 0.34701016545295715,
+      "learning_rate": 2.1243331817928645e-06,
+      "loss": 0.3279,
+      "step": 726
+    },
+    {
+      "epoch": 3.4143613890523836,
+      "grad_norm": 0.32958680391311646,
+      "learning_rate": 2.1179066419294043e-06,
+      "loss": 0.3205,
+      "step": 727
+    },
+    {
+      "epoch": 3.4190700412007065,
+      "grad_norm": 0.37069225311279297,
+      "learning_rate": 2.1114826863194882e-06,
+      "loss": 0.338,
+      "step": 728
+    },
+    {
+      "epoch": 3.423778693349029,
+      "grad_norm": 0.3406302034854889,
+      "learning_rate": 2.1050613584109518e-06,
+      "loss": 0.3331,
+      "step": 729
+    },
+    {
+      "epoch": 3.4284873454973512,
+      "grad_norm": 0.32335618138313293,
+      "learning_rate": 2.0986427016338626e-06,
+      "loss": 0.3124,
+      "step": 730
+    },
+    {
+      "epoch": 3.433195997645674,
+      "grad_norm": 0.35719034075737,
+      "learning_rate": 2.0922267594002182e-06,
+      "loss": 0.3433,
+      "step": 731
+    },
+    {
+      "epoch": 3.4379046497939965,
+      "grad_norm": 0.3510840833187103,
+      "learning_rate": 2.085813575103657e-06,
+      "loss": 0.3244,
+      "step": 732
+    },
+    {
+      "epoch": 3.442613301942319,
+      "grad_norm": 0.3347994387149811,
+      "learning_rate": 2.0794031921191683e-06,
+      "loss": 0.3289,
+      "step": 733
+    },
+    {
+      "epoch": 3.4473219540906417,
+      "grad_norm": 0.3473336696624756,
+      "learning_rate": 2.072995653802791e-06,
+      "loss": 0.3251,
+      "step": 734
+    },
+    {
+      "epoch": 3.452030606238964,
+      "grad_norm": 0.3559500277042389,
+      "learning_rate": 2.0665910034913243e-06,
+      "loss": 0.3256,
+      "step": 735
+    },
+    {
+      "epoch": 3.4567392583872865,
+      "grad_norm": 0.3657936155796051,
+      "learning_rate": 2.060189284502037e-06,
+      "loss": 0.3485,
+      "step": 736
+    },
+    {
+      "epoch": 3.4614479105356093,
+      "grad_norm": 0.346377968788147,
+      "learning_rate": 2.05379054013237e-06,
+      "loss": 0.3246,
+      "step": 737
+    },
+    {
+      "epoch": 3.4661565626839317,
+      "grad_norm": 0.3639799654483795,
+      "learning_rate": 2.0473948136596487e-06,
+      "loss": 0.3281,
+      "step": 738
+    },
+    {
+      "epoch": 3.470865214832254,
+      "grad_norm": 0.35108593106269836,
+      "learning_rate": 2.041002148340784e-06,
+      "loss": 0.3182,
+      "step": 739
+    },
+    {
+      "epoch": 3.475573866980577,
+      "grad_norm": 0.3662627637386322,
+      "learning_rate": 2.034612587411984e-06,
+      "loss": 0.3391,
+      "step": 740
+    },
+    {
+      "epoch": 3.4802825191288993,
+      "grad_norm": 0.37259167432785034,
+      "learning_rate": 2.02822617408846e-06,
+      "loss": 0.3422,
+      "step": 741
+    },
+    {
+      "epoch": 3.4849911712772217,
+      "grad_norm": 0.3728717267513275,
+      "learning_rate": 2.021842951564137e-06,
+      "loss": 0.3438,
+      "step": 742
+    },
+    {
+      "epoch": 3.4896998234255445,
+      "grad_norm": 0.36524805426597595,
+      "learning_rate": 2.015462963011357e-06,
+      "loss": 0.3286,
+      "step": 743
+    },
+    {
+      "epoch": 3.494408475573867,
+      "grad_norm": 0.3473794758319855,
+      "learning_rate": 2.0090862515805896e-06,
+      "loss": 0.3347,
+      "step": 744
+    },
+    {
+      "epoch": 3.4991171277221893,
+      "grad_norm": 0.35428687930107117,
+      "learning_rate": 2.0027128604001417e-06,
+      "loss": 0.3229,
+      "step": 745
+    },
+    {
+      "epoch": 3.503825779870512,
+      "grad_norm": 0.3526255190372467,
+      "learning_rate": 1.9963428325758615e-06,
+      "loss": 0.3407,
+      "step": 746
+    },
+    {
+      "epoch": 3.5085344320188345,
+      "grad_norm": 0.4010574519634247,
+      "learning_rate": 1.989976211190851e-06,
+      "loss": 0.3357,
+      "step": 747
+    },
+    {
+      "epoch": 3.513243084167157,
+      "grad_norm": 0.3535029888153076,
+      "learning_rate": 1.983613039305173e-06,
+      "loss": 0.3479,
+      "step": 748
+    },
+    {
+      "epoch": 3.5179517363154797,
+      "grad_norm": 0.36888188123703003,
+      "learning_rate": 1.97725335995556e-06,
+      "loss": 0.3457,
+      "step": 749
+    },
+    {
+      "epoch": 3.522660388463802,
+      "grad_norm": 0.35986262559890747,
+      "learning_rate": 1.9708972161551215e-06,
+      "loss": 0.3304,
+      "step": 750
+    },
+    {
+      "epoch": 3.5273690406121245,
+      "grad_norm": 0.35900959372520447,
+      "learning_rate": 1.9645446508930556e-06,
+      "loss": 0.3415,
+      "step": 751
+    },
+    {
+      "epoch": 3.5320776927604474,
+      "grad_norm": 0.39998960494995117,
+      "learning_rate": 1.9581957071343593e-06,
+      "loss": 0.3662,
+      "step": 752
+    },
+    {
+      "epoch": 3.5367863449087698,
+      "grad_norm": 0.3455185890197754,
+      "learning_rate": 1.951850427819533e-06,
+      "loss": 0.3445,
+      "step": 753
+    },
+    {
+      "epoch": 3.541494997057092,
+      "grad_norm": 0.35920363664627075,
+      "learning_rate": 1.9455088558642932e-06,
+      "loss": 0.3572,
+      "step": 754
+    },
+    {
+      "epoch": 3.546203649205415,
+      "grad_norm": 0.3520212471485138,
+      "learning_rate": 1.9391710341592847e-06,
+      "loss": 0.3413,
+      "step": 755
+    },
+    {
+      "epoch": 3.5509123013537374,
+      "grad_norm": 0.36848849058151245,
+      "learning_rate": 1.9328370055697836e-06,
+      "loss": 0.3304,
+      "step": 756
+    },
+    {
+      "epoch": 3.55562095350206,
+      "grad_norm": 0.3467768132686615,
+      "learning_rate": 1.9265068129354164e-06,
+      "loss": 0.3133,
+      "step": 757
+    },
+    {
+      "epoch": 3.5603296056503826,
+      "grad_norm": 0.38214123249053955,
+      "learning_rate": 1.920180499069862e-06,
+      "loss": 0.353,
+      "step": 758
+    },
+    {
+      "epoch": 3.565038257798705,
+      "grad_norm": 0.33890092372894287,
+      "learning_rate": 1.9138581067605664e-06,
+      "loss": 0.3144,
+      "step": 759
+    },
+    {
+      "epoch": 3.569746909947028,
+      "grad_norm": 0.36359307169914246,
+      "learning_rate": 1.9075396787684534e-06,
+      "loss": 0.3176,
+      "step": 760
+    },
+    {
+      "epoch": 3.57445556209535,
+      "grad_norm": 0.3256242871284485,
+      "learning_rate": 1.901225257827633e-06,
+      "loss": 0.3186,
+      "step": 761
+    },
+    {
+      "epoch": 3.5791642142436726,
+      "grad_norm": 0.33996862173080444,
+      "learning_rate": 1.8949148866451154e-06,
+      "loss": 0.329,
+      "step": 762
+    },
+    {
+      "epoch": 3.5838728663919954,
+      "grad_norm": 0.38446974754333496,
+      "learning_rate": 1.8886086079005201e-06,
+      "loss": 0.3502,
+      "step": 763
+    },
+    {
+      "epoch": 3.588581518540318,
+      "grad_norm": 0.33746618032455444,
+      "learning_rate": 1.8823064642457877e-06,
+      "loss": 0.3172,
+      "step": 764
+    },
+    {
+      "epoch": 3.5932901706886407,
+      "grad_norm": 0.3711576759815216,
+      "learning_rate": 1.8760084983048898e-06,
+      "loss": 0.3489,
+      "step": 765
+    },
+    {
+      "epoch": 3.597998822836963,
+      "grad_norm": 0.36693471670150757,
+      "learning_rate": 1.8697147526735468e-06,
+      "loss": 0.3418,
+      "step": 766
+    },
+    {
+      "epoch": 3.6027074749852854,
+      "grad_norm": 0.40748727321624756,
+      "learning_rate": 1.8634252699189311e-06,
+      "loss": 0.3443,
+      "step": 767
+    },
+    {
+      "epoch": 3.6074161271336083,
+      "grad_norm": 0.3639236092567444,
+      "learning_rate": 1.8571400925793855e-06,
+      "loss": 0.3396,
+      "step": 768
+    },
+    {
+      "epoch": 3.6121247792819307,
+      "grad_norm": 0.35893839597702026,
+      "learning_rate": 1.8508592631641328e-06,
+      "loss": 0.3369,
+      "step": 769
+    },
+    {
+      "epoch": 3.616833431430253,
+      "grad_norm": 0.3774881064891815,
+      "learning_rate": 1.8445828241529884e-06,
+      "loss": 0.3311,
+      "step": 770
+    },
+    {
+      "epoch": 3.621542083578576,
+      "grad_norm": 0.3896530866622925,
+      "learning_rate": 1.8383108179960769e-06,
+      "loss": 0.337,
+      "step": 771
+    },
+    {
+      "epoch": 3.6262507357268983,
+      "grad_norm": 0.365604966878891,
+      "learning_rate": 1.8320432871135378e-06,
+      "loss": 0.3357,
+      "step": 772
+    },
+    {
+      "epoch": 3.6309593878752207,
+      "grad_norm": 0.7592654228210449,
+      "learning_rate": 1.825780273895244e-06,
+      "loss": 0.3632,
+      "step": 773
+    },
+    {
+      "epoch": 3.6356680400235435,
+      "grad_norm": 0.4070909023284912,
+      "learning_rate": 1.8195218207005139e-06,
+      "loss": 0.3431,
+      "step": 774
+    },
+    {
+      "epoch": 3.640376692171866,
+      "grad_norm": 0.3568059206008911,
+      "learning_rate": 1.8132679698578238e-06,
+      "loss": 0.3218,
+      "step": 775
+    },
+    {
+      "epoch": 3.6450853443201883,
+      "grad_norm": 0.3874458074569702,
+      "learning_rate": 1.807018763664524e-06,
+      "loss": 0.3487,
+      "step": 776
+    },
+    {
+      "epoch": 3.649793996468511,
+      "grad_norm": 0.6760424971580505,
+      "learning_rate": 1.800774244386549e-06,
+      "loss": 0.3396,
+      "step": 777
+    },
+    {
+      "epoch": 3.6545026486168335,
+      "grad_norm": 0.3860941529273987,
+      "learning_rate": 1.7945344542581356e-06,
+      "loss": 0.3687,
+      "step": 778
+    },
+    {
+      "epoch": 3.659211300765156,
+      "grad_norm": 0.399251252412796,
+      "learning_rate": 1.7882994354815343e-06,
+      "loss": 0.3452,
+      "step": 779
+    },
+    {
+      "epoch": 3.6639199529134787,
+      "grad_norm": 0.3464195430278778,
+      "learning_rate": 1.7820692302267251e-06,
+      "loss": 0.3219,
+      "step": 780
+    },
+    {
+      "epoch": 3.668628605061801,
+      "grad_norm": 0.33919650316238403,
+      "learning_rate": 1.7758438806311329e-06,
+      "loss": 0.3099,
+      "step": 781
+    },
+    {
+      "epoch": 3.6733372572101235,
+      "grad_norm": 0.35061123967170715,
+      "learning_rate": 1.7696234287993416e-06,
+      "loss": 0.3229,
+      "step": 782
+    },
+    {
+      "epoch": 3.6780459093584463,
+      "grad_norm": 0.3360130488872528,
+      "learning_rate": 1.763407916802809e-06,
+      "loss": 0.3225,
+      "step": 783
+    },
+    {
+      "epoch": 3.6827545615067687,
+      "grad_norm": 0.3464799225330353,
+      "learning_rate": 1.7571973866795817e-06,
+      "loss": 0.3228,
+      "step": 784
+    },
+    {
+      "epoch": 3.687463213655091,
+      "grad_norm": 0.3488558232784271,
+      "learning_rate": 1.750991880434015e-06,
+      "loss": 0.3488,
+      "step": 785
+    },
+    {
+      "epoch": 3.692171865803414,
+      "grad_norm": 0.5446019768714905,
+      "learning_rate": 1.7447914400364835e-06,
+      "loss": 0.3491,
+      "step": 786
+    },
+    {
+      "epoch": 3.6968805179517363,
+      "grad_norm": 0.3364548981189728,
+      "learning_rate": 1.7385961074230994e-06,
+      "loss": 0.3192,
+      "step": 787
+    },
+    {
+      "epoch": 3.7015891701000587,
+      "grad_norm": 0.34718620777130127,
+      "learning_rate": 1.7324059244954294e-06,
+      "loss": 0.3368,
+      "step": 788
+    },
+    {
+      "epoch": 3.7062978222483816,
+      "grad_norm": 0.35897183418273926,
+      "learning_rate": 1.7262209331202095e-06,
+      "loss": 0.3414,
+      "step": 789
+    },
+    {
+      "epoch": 3.711006474396704,
+      "grad_norm": 0.37337997555732727,
+      "learning_rate": 1.7200411751290662e-06,
+      "loss": 0.3426,
+      "step": 790
+    },
+    {
+      "epoch": 3.7157151265450263,
+      "grad_norm": 0.34406691789627075,
+      "learning_rate": 1.7138666923182274e-06,
+      "loss": 0.319,
+      "step": 791
+    },
+    {
+      "epoch": 3.720423778693349,
+      "grad_norm": 0.36620885133743286,
+      "learning_rate": 1.7076975264482434e-06,
+      "loss": 0.3587,
+      "step": 792
+    },
+    {
+      "epoch": 3.7251324308416716,
+      "grad_norm": 0.351634681224823,
+      "learning_rate": 1.7015337192437053e-06,
+      "loss": 0.3332,
+      "step": 793
+    },
+    {
+      "epoch": 3.729841082989994,
+      "grad_norm": 0.386874794960022,
+      "learning_rate": 1.6953753123929597e-06,
+      "loss": 0.3433,
+      "step": 794
+    },
+    {
+      "epoch": 3.734549735138317,
+      "grad_norm": 0.34270480275154114,
+      "learning_rate": 1.6892223475478293e-06,
+      "loss": 0.3339,
+      "step": 795
+    },
+    {
+      "epoch": 3.739258387286639,
+      "grad_norm": 0.3725406229496002,
+      "learning_rate": 1.6830748663233306e-06,
+      "loss": 0.3551,
+      "step": 796
+    },
+    {
+      "epoch": 3.7439670394349616,
+      "grad_norm": 0.34803131222724915,
+      "learning_rate": 1.6769329102973913e-06,
+      "loss": 0.3187,
+      "step": 797
+    },
+    {
+      "epoch": 3.7486756915832844,
+      "grad_norm": 0.3435935080051422,
+      "learning_rate": 1.670796521010569e-06,
+      "loss": 0.3218,
+      "step": 798
+    },
+    {
+      "epoch": 3.753384343731607,
+      "grad_norm": 0.3839336931705475,
+      "learning_rate": 1.6646657399657745e-06,
+      "loss": 0.3581,
+      "step": 799
+    },
+    {
+      "epoch": 3.758092995879929,
+      "grad_norm": 0.37887904047966003,
+      "learning_rate": 1.6585406086279847e-06,
+      "loss": 0.3501,
+      "step": 800
+    },
+    {
+      "epoch": 3.762801648028252,
+      "grad_norm": 0.3566491901874542,
+      "learning_rate": 1.652421168423966e-06,
+      "loss": 0.3467,
+      "step": 801
+    },
+    {
+      "epoch": 3.7675103001765744,
+      "grad_norm": 0.37119993567466736,
+      "learning_rate": 1.6463074607419943e-06,
+      "loss": 0.3442,
+      "step": 802
+    },
+    {
+      "epoch": 3.772218952324897,
+      "grad_norm": 0.34793585538864136,
+      "learning_rate": 1.6401995269315712e-06,
+      "loss": 0.3411,
+      "step": 803
+    },
+    {
+      "epoch": 3.7769276044732196,
+      "grad_norm": 0.3532625734806061,
+      "learning_rate": 1.6340974083031525e-06,
+      "loss": 0.3417,
+      "step": 804
+    },
+    {
+      "epoch": 3.781636256621542,
+      "grad_norm": 0.36595654487609863,
+      "learning_rate": 1.6280011461278586e-06,
+      "loss": 0.3621,
+      "step": 805
+    },
+    {
+      "epoch": 3.7863449087698644,
+      "grad_norm": 0.365357905626297,
+      "learning_rate": 1.6219107816372027e-06,
+      "loss": 0.3328,
+      "step": 806
+    },
+    {
+      "epoch": 3.7910535609181872,
+      "grad_norm": 0.33471792936325073,
+      "learning_rate": 1.6158263560228089e-06,
+      "loss": 0.3258,
+      "step": 807
+    },
+    {
+      "epoch": 3.7957622130665096,
+      "grad_norm": 0.35250982642173767,
+      "learning_rate": 1.6097479104361328e-06,
+      "loss": 0.3396,
+      "step": 808
+    },
+    {
+      "epoch": 3.800470865214832,
+      "grad_norm": 0.3582373261451721,
+      "learning_rate": 1.6036754859881888e-06,
+      "loss": 0.3208,
+      "step": 809
+    },
+    {
+      "epoch": 3.805179517363155,
+      "grad_norm": 0.35405614972114563,
+      "learning_rate": 1.5976091237492637e-06,
+      "loss": 0.3355,
+      "step": 810
+    },
+    {
+      "epoch": 3.8098881695114772,
+      "grad_norm": 0.3517574071884155,
+      "learning_rate": 1.5915488647486453e-06,
+      "loss": 0.3364,
+      "step": 811
+    },
+    {
+      "epoch": 3.8145968216597996,
+      "grad_norm": 0.3884425163269043,
+      "learning_rate": 1.5854947499743414e-06,
+      "loss": 0.3548,
+      "step": 812
+    },
+    {
+      "epoch": 3.8193054738081225,
+      "grad_norm": 0.3678779900074005,
+      "learning_rate": 1.5794468203728053e-06,
+      "loss": 0.3499,
+      "step": 813
+    },
+    {
+      "epoch": 3.824014125956445,
+      "grad_norm": 0.3617662489414215,
+      "learning_rate": 1.5734051168486565e-06,
+      "loss": 0.3305,
+      "step": 814
+    },
+    {
+      "epoch": 3.8287227781047672,
+      "grad_norm": 0.33246317505836487,
+      "learning_rate": 1.5673696802644053e-06,
+      "loss": 0.3065,
+      "step": 815
+    },
+    {
+      "epoch": 3.83343143025309,
+      "grad_norm": 0.3756829798221588,
+      "learning_rate": 1.5613405514401757e-06,
+      "loss": 0.351,
+      "step": 816
+    },
+    {
+      "epoch": 3.8381400824014125,
+      "grad_norm": 0.3410739004611969,
+      "learning_rate": 1.5553177711534296e-06,
+      "loss": 0.32,
+      "step": 817
+    },
+    {
+      "epoch": 3.8428487345497353,
+      "grad_norm": 0.3635987937450409,
+      "learning_rate": 1.5493013801386924e-06,
+      "loss": 0.3337,
+      "step": 818
+    },
+    {
+      "epoch": 3.8475573866980577,
+      "grad_norm": 0.3543144166469574,
+      "learning_rate": 1.5432914190872757e-06,
+      "loss": 0.3409,
+      "step": 819
+    },
+    {
+      "epoch": 3.85226603884638,
+      "grad_norm": 0.3444572687149048,
+      "learning_rate": 1.5372879286470022e-06,
+      "loss": 0.3264,
+      "step": 820
+    },
+    {
+      "epoch": 3.856974690994703,
+      "grad_norm": 0.360392689704895,
+      "learning_rate": 1.5312909494219308e-06,
+      "loss": 0.338,
+      "step": 821
+    },
+    {
+      "epoch": 3.8616833431430253,
+      "grad_norm": 0.3705150783061981,
+      "learning_rate": 1.5253005219720822e-06,
+      "loss": 0.3071,
+      "step": 822
+    },
+    {
+      "epoch": 3.8663919952913477,
+      "grad_norm": 0.37416768074035645,
+      "learning_rate": 1.519316686813168e-06,
+      "loss": 0.3564,
+      "step": 823
+    },
+    {
+      "epoch": 3.8711006474396705,
+      "grad_norm": 0.3697960078716278,
+      "learning_rate": 1.5133394844163093e-06,
+      "loss": 0.3426,
+      "step": 824
+    },
+    {
+      "epoch": 3.875809299587993,
+      "grad_norm": 0.3538321554660797,
+      "learning_rate": 1.5073689552077692e-06,
+      "loss": 0.3291,
+      "step": 825
+    },
+    {
+      "epoch": 3.8805179517363158,
+      "grad_norm": 0.3262041211128235,
+      "learning_rate": 1.5014051395686768e-06,
+      "loss": 0.311,
+      "step": 826
+    },
+    {
+      "epoch": 3.885226603884638,
+      "grad_norm": 0.35173299908638,
+      "learning_rate": 1.4954480778347543e-06,
+      "loss": 0.3331,
+      "step": 827
+    },
+    {
+      "epoch": 3.8899352560329605,
+      "grad_norm": 0.36466389894485474,
+      "learning_rate": 1.489497810296046e-06,
+      "loss": 0.3419,
+      "step": 828
+    },
+    {
+      "epoch": 3.8946439081812834,
+      "grad_norm": 0.35103070735931396,
+      "learning_rate": 1.4835543771966422e-06,
+      "loss": 0.3433,
+      "step": 829
+    },
+    {
+      "epoch": 3.8993525603296058,
+      "grad_norm": 0.35364291071891785,
+      "learning_rate": 1.4776178187344107e-06,
+      "loss": 0.3171,
+      "step": 830
+    },
+    {
+      "epoch": 3.904061212477928,
+      "grad_norm": 0.3403826951980591,
+      "learning_rate": 1.4716881750607215e-06,
+      "loss": 0.3275,
+      "step": 831
+    },
+    {
+      "epoch": 3.908769864626251,
+      "grad_norm": 0.38063591718673706,
+      "learning_rate": 1.4657654862801798e-06,
+      "loss": 0.3314,
+      "step": 832
+    },
+    {
+      "epoch": 3.9134785167745734,
+      "grad_norm": 0.3796558976173401,
+      "learning_rate": 1.4598497924503498e-06,
+      "loss": 0.3609,
+      "step": 833
+    },
+    {
+      "epoch": 3.9181871689228958,
+      "grad_norm": 0.33824312686920166,
+      "learning_rate": 1.4539411335814868e-06,
+      "loss": 0.3182,
+      "step": 834
+    },
+    {
+      "epoch": 3.9228958210712186,
+      "grad_norm": 0.3416818082332611,
+      "learning_rate": 1.4480395496362648e-06,
+      "loss": 0.3303,
+      "step": 835
+    },
+    {
+      "epoch": 3.927604473219541,
+      "grad_norm": 0.3429725766181946,
+      "learning_rate": 1.4421450805295084e-06,
+      "loss": 0.3218,
+      "step": 836
+    },
+    {
+      "epoch": 3.9323131253678634,
+      "grad_norm": 0.3443184792995453,
+      "learning_rate": 1.4362577661279225e-06,
+      "loss": 0.3344,
+      "step": 837
+    },
+    {
+      "epoch": 3.937021777516186,
+      "grad_norm": 0.36939215660095215,
+      "learning_rate": 1.4303776462498187e-06,
+      "loss": 0.3574,
+      "step": 838
+    },
+    {
+      "epoch": 3.9417304296645086,
+      "grad_norm": 0.33150216937065125,
+      "learning_rate": 1.4245047606648518e-06,
+      "loss": 0.3155,
+      "step": 839
+    },
+    {
+      "epoch": 3.946439081812831,
+      "grad_norm": 0.3554579019546509,
+      "learning_rate": 1.4186391490937482e-06,
+      "loss": 0.3458,
+      "step": 840
+    },
+    {
+      "epoch": 3.951147733961154,
+      "grad_norm": 0.36509135365486145,
+      "learning_rate": 1.4127808512080331e-06,
+      "loss": 0.333,
+      "step": 841
+    },
+    {
+      "epoch": 3.955856386109476,
+      "grad_norm": 0.3694615960121155,
+      "learning_rate": 1.406929906629774e-06,
+      "loss": 0.3463,
+      "step": 842
+    },
+    {
+      "epoch": 3.9605650382577986,
+      "grad_norm": 0.3466503918170929,
+      "learning_rate": 1.401086354931297e-06,
+      "loss": 0.3342,
+      "step": 843
+    },
+    {
+      "epoch": 3.9652736904061214,
+      "grad_norm": 0.3704316020011902,
+      "learning_rate": 1.3952502356349323e-06,
+      "loss": 0.3585,
+      "step": 844
+    },
+    {
+      "epoch": 3.969982342554444,
+      "grad_norm": 0.36790791153907776,
+      "learning_rate": 1.3894215882127371e-06,
+      "loss": 0.3296,
+      "step": 845
+    },
+    {
+      "epoch": 3.974690994702766,
+      "grad_norm": 0.3574499189853668,
+      "learning_rate": 1.3836004520862404e-06,
+      "loss": 0.3374,
+      "step": 846
+    },
+    {
+      "epoch": 3.979399646851089,
+      "grad_norm": 0.3674068748950958,
+      "learning_rate": 1.3777868666261619e-06,
+      "loss": 0.3401,
+      "step": 847
+    },
+    {
+      "epoch": 3.9841082989994114,
+      "grad_norm": 0.3557773530483246,
+      "learning_rate": 1.3719808711521573e-06,
+      "loss": 0.3377,
+      "step": 848
+    },
+    {
+      "epoch": 3.988816951147734,
+      "grad_norm": 0.4119923412799835,
+      "learning_rate": 1.3661825049325462e-06,
+      "loss": 0.3274,
+      "step": 849
+    },
+    {
+      "epoch": 3.9935256032960567,
+      "grad_norm": 0.35379621386528015,
+      "learning_rate": 1.3603918071840488e-06,
+      "loss": 0.3417,
+      "step": 850
+    },
+    {
+      "epoch": 3.998234255444379,
+      "grad_norm": 0.35988909006118774,
+      "learning_rate": 1.3546088170715222e-06,
+      "loss": 0.3274,
+      "step": 851
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.35988909006118774,
+      "learning_rate": 1.3488335737076914e-06,
+      "loss": 0.3568,
+      "step": 852
+    },
+    {
+      "epoch": 4.004708652148323,
+      "grad_norm": 0.682213544845581,
+      "learning_rate": 1.343066116152887e-06,
+      "loss": 0.3152,
+      "step": 853
+    },
+    {
+      "epoch": 4.009417304296645,
+      "grad_norm": 0.3923402428627014,
+      "learning_rate": 1.3373064834147819e-06,
+      "loss": 0.3242,
+      "step": 854
+    },
+    {
+      "epoch": 4.014125956444968,
+      "grad_norm": 0.3617793917655945,
+      "learning_rate": 1.3315547144481263e-06,
+      "loss": 0.3177,
+      "step": 855
+    },
+    {
+      "epoch": 4.0188346085932904,
+      "grad_norm": 0.3607487678527832,
+      "learning_rate": 1.3258108481544849e-06,
+      "loss": 0.3238,
+      "step": 856
+    },
+    {
+      "epoch": 4.023543260741612,
+      "grad_norm": 0.37872302532196045,
+      "learning_rate": 1.3200749233819738e-06,
+      "loss": 0.3383,
+      "step": 857
+    },
+    {
+      "epoch": 4.028251912889935,
+      "grad_norm": 0.3622148036956787,
+      "learning_rate": 1.314346978924994e-06,
+      "loss": 0.3033,
+      "step": 858
+    },
+    {
+      "epoch": 4.032960565038258,
+      "grad_norm": 0.3558484613895416,
+      "learning_rate": 1.3086270535239776e-06,
+      "loss": 0.3075,
+      "step": 859
+    },
+    {
+      "epoch": 4.03766921718658,
+      "grad_norm": 0.3856630027294159,
+      "learning_rate": 1.3029151858651145e-06,
+      "loss": 0.3165,
+      "step": 860
+    },
+    {
+      "epoch": 4.042377869334903,
+      "grad_norm": 0.38117724657058716,
+      "learning_rate": 1.2972114145801046e-06,
+      "loss": 0.3155,
+      "step": 861
+    },
+    {
+      "epoch": 4.047086521483226,
+      "grad_norm": 0.38278907537460327,
+      "learning_rate": 1.2915157782458804e-06,
+      "loss": 0.3145,
+      "step": 862
+    },
+    {
+      "epoch": 4.051795173631548,
+      "grad_norm": 0.3633948862552643,
+      "learning_rate": 1.2858283153843608e-06,
+      "loss": 0.2881,
+      "step": 863
+    },
+    {
+      "epoch": 4.0565038257798705,
+      "grad_norm": 0.35489869117736816,
+      "learning_rate": 1.280149064462179e-06,
+      "loss": 0.3093,
+      "step": 864
+    },
+    {
+      "epoch": 4.061212477928193,
+      "grad_norm": 0.3667081296443939,
+      "learning_rate": 1.2744780638904336e-06,
+      "loss": 0.3368,
+      "step": 865
+    },
+    {
+      "epoch": 4.065921130076515,
+      "grad_norm": 0.3660667836666107,
+      "learning_rate": 1.2688153520244162e-06,
+      "loss": 0.3093,
+      "step": 866
+    },
+    {
+      "epoch": 4.070629782224838,
+      "grad_norm": 0.36665669083595276,
+      "learning_rate": 1.2631609671633632e-06,
+      "loss": 0.2986,
+      "step": 867
+    },
+    {
+      "epoch": 4.075338434373161,
+      "grad_norm": 0.35874009132385254,
+      "learning_rate": 1.2575149475501891e-06,
+      "loss": 0.3223,
+      "step": 868
+    },
+    {
+      "epoch": 4.080047086521483,
+      "grad_norm": 0.38370969891548157,
+      "learning_rate": 1.251877331371233e-06,
+      "loss": 0.2954,
+      "step": 869
+    },
+    {
+      "epoch": 4.084755738669806,
+      "grad_norm": 0.3539574146270752,
+      "learning_rate": 1.2462481567559966e-06,
+      "loss": 0.2904,
+      "step": 870
+    },
+    {
+      "epoch": 4.0894643908181285,
+      "grad_norm": 0.36115357279777527,
+      "learning_rate": 1.2406274617768893e-06,
+      "loss": 0.3022,
+      "step": 871
+    },
+    {
+      "epoch": 4.0941730429664505,
+      "grad_norm": 0.3756008446216583,
+      "learning_rate": 1.235015284448969e-06,
+      "loss": 0.3245,
+      "step": 872
+    },
+    {
+      "epoch": 4.098881695114773,
+      "grad_norm": 0.3537841737270355,
+      "learning_rate": 1.2294116627296827e-06,
+      "loss": 0.2973,
+      "step": 873
+    },
+    {
+      "epoch": 4.103590347263096,
+      "grad_norm": 0.3482496738433838,
+      "learning_rate": 1.2238166345186152e-06,
+      "loss": 0.3072,
+      "step": 874
+    },
+    {
+      "epoch": 4.108298999411418,
+      "grad_norm": 0.35395577549934387,
+      "learning_rate": 1.2182302376572295e-06,
+      "loss": 0.3076,
+      "step": 875
+    },
+    {
+      "epoch": 4.113007651559741,
+      "grad_norm": 0.36247384548187256,
+      "learning_rate": 1.212652509928611e-06,
+      "loss": 0.3147,
+      "step": 876
+    },
+    {
+      "epoch": 4.117716303708064,
+      "grad_norm": 0.37728843092918396,
+      "learning_rate": 1.2070834890572105e-06,
+      "loss": 0.3286,
+      "step": 877
+    },
+    {
+      "epoch": 4.122424955856386,
+      "grad_norm": 0.3559524416923523,
+      "learning_rate": 1.2015232127085932e-06,
+      "loss": 0.2873,
+      "step": 878
+    },
+    {
+      "epoch": 4.1271336080047085,
+      "grad_norm": 0.34051647782325745,
+      "learning_rate": 1.1959717184891803e-06,
+      "loss": 0.3166,
+      "step": 879
+    },
+    {
+      "epoch": 4.131842260153031,
+      "grad_norm": 0.3496254086494446,
+      "learning_rate": 1.1904290439459974e-06,
+      "loss": 0.3051,
+      "step": 880
+    },
+    {
+      "epoch": 4.136550912301354,
+      "grad_norm": 0.346172958612442,
+      "learning_rate": 1.1848952265664157e-06,
+      "loss": 0.285,
+      "step": 881
+    },
+    {
+      "epoch": 4.141259564449676,
+      "grad_norm": 0.35215529799461365,
+      "learning_rate": 1.1793703037779056e-06,
+      "loss": 0.2974,
+      "step": 882
+    },
+    {
+      "epoch": 4.145968216597999,
+      "grad_norm": 0.3513922095298767,
+      "learning_rate": 1.173854312947775e-06,
+      "loss": 0.3216,
+      "step": 883
+    },
+    {
+      "epoch": 4.150676868746322,
+      "grad_norm": 0.3996025621891022,
+      "learning_rate": 1.1683472913829286e-06,
+      "loss": 0.308,
+      "step": 884
+    },
+    {
+      "epoch": 4.155385520894644,
+      "grad_norm": 0.3480626940727234,
+      "learning_rate": 1.1628492763296006e-06,
+      "loss": 0.3152,
+      "step": 885
+    },
+    {
+      "epoch": 4.160094173042967,
+      "grad_norm": 0.3710547089576721,
+      "learning_rate": 1.1573603049731156e-06,
+      "loss": 0.3255,
+      "step": 886
+    },
+    {
+      "epoch": 4.164802825191289,
+      "grad_norm": 0.3572455942630768,
+      "learning_rate": 1.15188041443763e-06,
+      "loss": 0.3115,
+      "step": 887
+    },
+    {
+      "epoch": 4.169511477339611,
+      "grad_norm": 0.45938190817832947,
+      "learning_rate": 1.1464096417858821e-06,
+      "loss": 0.3172,
+      "step": 888
+    },
+    {
+      "epoch": 4.174220129487934,
+      "grad_norm": 0.3499940037727356,
+      "learning_rate": 1.1409480240189442e-06,
+      "loss": 0.3136,
+      "step": 889
+    },
+    {
+      "epoch": 4.178928781636257,
+      "grad_norm": 0.35575103759765625,
+      "learning_rate": 1.135495598075969e-06,
+      "loss": 0.2879,
+      "step": 890
+    },
+    {
+      "epoch": 4.183637433784579,
+      "grad_norm": 0.33789294958114624,
+      "learning_rate": 1.130052400833943e-06,
+      "loss": 0.312,
+      "step": 891
+    },
+    {
+      "epoch": 4.188346085932902,
+      "grad_norm": 0.35234612226486206,
+      "learning_rate": 1.1246184691074317e-06,
+      "loss": 0.2925,
+      "step": 892
+    },
+    {
+      "epoch": 4.193054738081225,
+      "grad_norm": 0.34792086482048035,
+      "learning_rate": 1.119193839648337e-06,
+      "loss": 0.3019,
+      "step": 893
+    },
+    {
+      "epoch": 4.197763390229547,
+      "grad_norm": 0.34613871574401855,
+      "learning_rate": 1.1137785491456455e-06,
+      "loss": 0.286,
+      "step": 894
+    },
+    {
+      "epoch": 4.202472042377869,
+      "grad_norm": 0.3518754839897156,
+      "learning_rate": 1.1083726342251802e-06,
+      "loss": 0.3076,
+      "step": 895
+    },
+    {
+      "epoch": 4.207180694526192,
+      "grad_norm": 0.3639785647392273,
+      "learning_rate": 1.102976131449352e-06,
+      "loss": 0.331,
+      "step": 896
+    },
+    {
+      "epoch": 4.211889346674514,
+      "grad_norm": 0.3593629002571106,
+      "learning_rate": 1.0975890773169156e-06,
+      "loss": 0.3312,
+      "step": 897
+    },
+    {
+      "epoch": 4.216597998822837,
+      "grad_norm": 0.37032291293144226,
+      "learning_rate": 1.0922115082627197e-06,
+      "loss": 0.3361,
+      "step": 898
+    },
+    {
+      "epoch": 4.22130665097116,
+      "grad_norm": 0.3786473274230957,
+      "learning_rate": 1.0868434606574621e-06,
+      "loss": 0.3094,
+      "step": 899
+    },
+    {
+      "epoch": 4.226015303119482,
+      "grad_norm": 0.35703983902931213,
+      "learning_rate": 1.0814849708074416e-06,
+      "loss": 0.297,
+      "step": 900
+    },
+    {
+      "epoch": 4.230723955267805,
+      "grad_norm": 0.3440445363521576,
+      "learning_rate": 1.0761360749543158e-06,
+      "loss": 0.2927,
+      "step": 901
+    },
+    {
+      "epoch": 4.2354326074161275,
+      "grad_norm": 0.3606710433959961,
+      "learning_rate": 1.070796809274853e-06,
+      "loss": 0.3129,
+      "step": 902
+    },
+    {
+      "epoch": 4.240141259564449,
+      "grad_norm": 0.35009413957595825,
+      "learning_rate": 1.0654672098806906e-06,
+      "loss": 0.2728,
+      "step": 903
+    },
+    {
+      "epoch": 4.244849911712772,
+      "grad_norm": 0.3345305919647217,
+      "learning_rate": 1.0601473128180855e-06,
+      "loss": 0.2922,
+      "step": 904
+    },
+    {
+      "epoch": 4.249558563861095,
+      "grad_norm": 0.3608264625072479,
+      "learning_rate": 1.0548371540676768e-06,
+      "loss": 0.3196,
+      "step": 905
+    },
+    {
+      "epoch": 4.254267216009417,
+      "grad_norm": 0.48483705520629883,
+      "learning_rate": 1.0495367695442394e-06,
+      "loss": 0.3183,
+      "step": 906
+    },
+    {
+      "epoch": 4.25897586815774,
+      "grad_norm": 0.3715311288833618,
+      "learning_rate": 1.0442461950964373e-06,
+      "loss": 0.3269,
+      "step": 907
+    },
+    {
+      "epoch": 4.263684520306063,
+      "grad_norm": 0.3627394139766693,
+      "learning_rate": 1.038965466506591e-06,
+      "loss": 0.3047,
+      "step": 908
+    },
+    {
+      "epoch": 4.268393172454385,
+      "grad_norm": 0.3572857975959778,
+      "learning_rate": 1.033694619490424e-06,
+      "loss": 0.2977,
+      "step": 909
+    },
+    {
+      "epoch": 4.2731018246027075,
+      "grad_norm": 0.3349437117576599,
+      "learning_rate": 1.0284336896968306e-06,
+      "loss": 0.2997,
+      "step": 910
+    },
+    {
+      "epoch": 4.27781047675103,
+      "grad_norm": 0.3537830412387848,
+      "learning_rate": 1.0231827127076272e-06,
+      "loss": 0.3101,
+      "step": 911
+    },
+    {
+      "epoch": 4.282519128899352,
+      "grad_norm": 0.36906588077545166,
+      "learning_rate": 1.0179417240373181e-06,
+      "loss": 0.3291,
+      "step": 912
+    },
+    {
+      "epoch": 4.287227781047675,
+      "grad_norm": 0.39401233196258545,
+      "learning_rate": 1.0127107591328523e-06,
+      "loss": 0.3171,
+      "step": 913
+    },
+    {
+      "epoch": 4.291936433195998,
+      "grad_norm": 0.36123374104499817,
+      "learning_rate": 1.0074898533733834e-06,
+      "loss": 0.3193,
+      "step": 914
+    },
+    {
+      "epoch": 4.29664508534432,
+      "grad_norm": 0.36208614706993103,
+      "learning_rate": 1.0022790420700298e-06,
+      "loss": 0.2998,
+      "step": 915
+    },
+    {
+      "epoch": 4.301353737492643,
+      "grad_norm": 0.34652724862098694,
+      "learning_rate": 9.970783604656384e-07,
+      "loss": 0.3012,
+      "step": 916
+    },
+    {
+      "epoch": 4.3060623896409655,
+      "grad_norm": 0.3494437038898468,
+      "learning_rate": 9.918878437345445e-07,
+      "loss": 0.3079,
+      "step": 917
+    },
+    {
+      "epoch": 4.3107710417892875,
+      "grad_norm": 0.3824324905872345,
+      "learning_rate": 9.867075269823354e-07,
+      "loss": 0.2922,
+      "step": 918
+    },
+    {
+      "epoch": 4.31547969393761,
+      "grad_norm": 0.3468039035797119,
+      "learning_rate": 9.81537445245608e-07,
+      "loss": 0.3119,
+      "step": 919
+    },
+    {
+      "epoch": 4.320188346085933,
+      "grad_norm": 0.3550470471382141,
+      "learning_rate": 9.763776334917398e-07,
+      "loss": 0.2967,
+      "step": 920
+    },
+    {
+      "epoch": 4.324896998234255,
+      "grad_norm": 0.36313554644584656,
+      "learning_rate": 9.712281266186463e-07,
+      "loss": 0.2942,
+      "step": 921
+    },
+    {
+      "epoch": 4.329605650382578,
+      "grad_norm": 0.3467397093772888,
+      "learning_rate": 9.66088959454547e-07,
+      "loss": 0.311,
+      "step": 922
+    },
+    {
+      "epoch": 4.334314302530901,
+      "grad_norm": 0.335706502199173,
+      "learning_rate": 9.609601667577304e-07,
+      "loss": 0.2938,
+      "step": 923
+    },
+    {
+      "epoch": 4.339022954679223,
+      "grad_norm": 0.3451956510543823,
+      "learning_rate": 9.558417832163164e-07,
+      "loss": 0.3155,
+      "step": 924
+    },
+    {
+      "epoch": 4.3437316068275456,
+      "grad_norm": 0.3699392080307007,
+      "learning_rate": 9.507338434480265e-07,
+      "loss": 0.3255,
+      "step": 925
+    },
+    {
+      "epoch": 4.348440258975868,
+      "grad_norm": 0.364408940076828,
+      "learning_rate": 9.45636381999942e-07,
+      "loss": 0.289,
+      "step": 926
+    },
+    {
+      "epoch": 4.35314891112419,
+      "grad_norm": 0.33332404494285583,
+      "learning_rate": 9.405494333482817e-07,
+      "loss": 0.2991,
+      "step": 927
+    },
+    {
+      "epoch": 4.357857563272513,
+      "grad_norm": 0.36770814657211304,
+      "learning_rate": 9.354730318981561e-07,
+      "loss": 0.3088,
+      "step": 928
+    },
+    {
+      "epoch": 4.362566215420836,
+      "grad_norm": 0.35132384300231934,
+      "learning_rate": 9.304072119833441e-07,
+      "loss": 0.2852,
+      "step": 929
+    },
+    {
+      "epoch": 4.367274867569158,
+      "grad_norm": 0.327019602060318,
+      "learning_rate": 9.253520078660541e-07,
+      "loss": 0.2888,
+      "step": 930
+    },
+    {
+      "epoch": 4.371983519717481,
+      "grad_norm": 0.3622664511203766,
+      "learning_rate": 9.203074537367007e-07,
+      "loss": 0.3145,
+      "step": 931
+    },
+    {
+      "epoch": 4.376692171865804,
+      "grad_norm": 0.37222006916999817,
+      "learning_rate": 9.152735837136631e-07,
+      "loss": 0.3165,
+      "step": 932
+    },
+    {
+      "epoch": 4.381400824014126,
+      "grad_norm": 0.35063692927360535,
+      "learning_rate": 9.102504318430628e-07,
+      "loss": 0.3102,
+      "step": 933
+    },
+    {
+      "epoch": 4.386109476162448,
+      "grad_norm": 0.3508763909339905,
+      "learning_rate": 9.052380320985274e-07,
+      "loss": 0.3056,
+      "step": 934
+    },
+    {
+      "epoch": 4.390818128310771,
+      "grad_norm": 0.3588324785232544,
+      "learning_rate": 9.002364183809658e-07,
+      "loss": 0.3055,
+      "step": 935
+    },
+    {
+      "epoch": 4.395526780459093,
+      "grad_norm": 0.34880831837654114,
+      "learning_rate": 8.952456245183361e-07,
+      "loss": 0.3052,
+      "step": 936
+    },
+    {
+      "epoch": 4.400235432607416,
+      "grad_norm": 0.35600703954696655,
+      "learning_rate": 8.902656842654164e-07,
+      "loss": 0.2837,
+      "step": 937
+    },
+    {
+      "epoch": 4.404944084755739,
+      "grad_norm": 0.35241153836250305,
+      "learning_rate": 8.85296631303579e-07,
+      "loss": 0.3087,
+      "step": 938
+    },
+    {
+      "epoch": 4.409652736904061,
+      "grad_norm": 0.36634692549705505,
+      "learning_rate": 8.803384992405584e-07,
+      "loss": 0.3088,
+      "step": 939
+    },
+    {
+      "epoch": 4.414361389052384,
+      "grad_norm": 0.3660341501235962,
+      "learning_rate": 8.753913216102286e-07,
+      "loss": 0.3025,
+      "step": 940
+    },
+    {
+      "epoch": 4.4190700412007065,
+      "grad_norm": 0.3535331189632416,
+      "learning_rate": 8.704551318723744e-07,
+      "loss": 0.2905,
+      "step": 941
+    },
+    {
+      "epoch": 4.423778693349028,
+      "grad_norm": 0.3493208587169647,
+      "learning_rate": 8.655299634124648e-07,
+      "loss": 0.3032,
+      "step": 942
+    },
+    {
+      "epoch": 4.428487345497351,
+      "grad_norm": 0.35454756021499634,
+      "learning_rate": 8.606158495414258e-07,
+      "loss": 0.304,
+      "step": 943
+    },
+    {
+      "epoch": 4.433195997645674,
+      "grad_norm": 0.3448980450630188,
+      "learning_rate": 8.55712823495419e-07,
+      "loss": 0.3018,
+      "step": 944
+    },
+    {
+      "epoch": 4.437904649793996,
+      "grad_norm": 0.3334943950176239,
+      "learning_rate": 8.508209184356114e-07,
+      "loss": 0.2951,
+      "step": 945
+    },
+    {
+      "epoch": 4.442613301942319,
+      "grad_norm": 0.36533334851264954,
+      "learning_rate": 8.459401674479594e-07,
+      "loss": 0.3263,
+      "step": 946
+    },
+    {
+      "epoch": 4.447321954090642,
+      "grad_norm": 0.36448344588279724,
+      "learning_rate": 8.410706035429747e-07,
+      "loss": 0.287,
+      "step": 947
+    },
+    {
+      "epoch": 4.452030606238964,
+      "grad_norm": 0.33500760793685913,
+      "learning_rate": 8.362122596555089e-07,
+      "loss": 0.3082,
+      "step": 948
+    },
+    {
+      "epoch": 4.4567392583872865,
+      "grad_norm": 0.3625440299510956,
+      "learning_rate": 8.313651686445256e-07,
+      "loss": 0.2886,
+      "step": 949
+    },
+    {
+      "epoch": 4.461447910535609,
+      "grad_norm": 0.3715088367462158,
+      "learning_rate": 8.265293632928856e-07,
+      "loss": 0.3394,
+      "step": 950
+    },
+    {
+      "epoch": 4.466156562683932,
+      "grad_norm": 0.3713008761405945,
+      "learning_rate": 8.217048763071139e-07,
+      "loss": 0.3004,
+      "step": 951
+    },
+    {
+      "epoch": 4.470865214832254,
+      "grad_norm": 0.3510104715824127,
+      "learning_rate": 8.168917403171891e-07,
+      "loss": 0.3035,
+      "step": 952
+    },
+    {
+      "epoch": 4.475573866980577,
+      "grad_norm": 0.3302716612815857,
+      "learning_rate": 8.120899878763181e-07,
+      "loss": 0.2726,
+      "step": 953
+    },
+    {
+      "epoch": 4.4802825191289,
+      "grad_norm": 0.33889657258987427,
+      "learning_rate": 8.072996514607126e-07,
+      "loss": 0.2998,
+      "step": 954
+    },
+    {
+      "epoch": 4.484991171277222,
+      "grad_norm": 0.33504074811935425,
+      "learning_rate": 8.025207634693771e-07,
+      "loss": 0.2969,
+      "step": 955
+    },
+    {
+      "epoch": 4.4896998234255445,
+      "grad_norm": 0.3616447448730469,
+      "learning_rate": 7.97753356223884e-07,
+      "loss": 0.3177,
+      "step": 956
+    },
+    {
+      "epoch": 4.494408475573867,
+      "grad_norm": 0.36567652225494385,
+      "learning_rate": 7.92997461968158e-07,
+      "loss": 0.3116,
+      "step": 957
+    },
+    {
+      "epoch": 4.499117127722189,
+      "grad_norm": 0.38666656613349915,
+      "learning_rate": 7.882531128682539e-07,
+      "loss": 0.3331,
+      "step": 958
+    },
+    {
+      "epoch": 4.503825779870512,
+      "grad_norm": 0.3525826334953308,
+      "learning_rate": 7.835203410121444e-07,
+      "loss": 0.2967,
+      "step": 959
+    },
+    {
+      "epoch": 4.508534432018835,
+      "grad_norm": 0.3484201431274414,
+      "learning_rate": 7.787991784095e-07,
+      "loss": 0.3125,
+      "step": 960
+    },
+    {
+      "epoch": 4.513243084167157,
+      "grad_norm": 0.3635729253292084,
+      "learning_rate": 7.740896569914741e-07,
+      "loss": 0.3158,
+      "step": 961
+    },
+    {
+      "epoch": 4.51795173631548,
+      "grad_norm": 0.3555293679237366,
+      "learning_rate": 7.693918086104826e-07,
+      "loss": 0.2998,
+      "step": 962
+    },
+    {
+      "epoch": 4.522660388463803,
+      "grad_norm": 0.367501437664032,
+      "learning_rate": 7.647056650399951e-07,
+      "loss": 0.2916,
+      "step": 963
+    },
+    {
+      "epoch": 4.5273690406121245,
+      "grad_norm": 0.3521662652492523,
+      "learning_rate": 7.600312579743161e-07,
+      "loss": 0.3251,
+      "step": 964
+    },
+    {
+      "epoch": 4.532077692760447,
+      "grad_norm": 0.5477653741836548,
+      "learning_rate": 7.553686190283716e-07,
+      "loss": 0.2994,
+      "step": 965
+    },
+    {
+      "epoch": 4.53678634490877,
+      "grad_norm": 0.3754388391971588,
+      "learning_rate": 7.507177797374929e-07,
+      "loss": 0.3392,
+      "step": 966
+    },
+    {
+      "epoch": 4.541494997057092,
+      "grad_norm": 0.3783737123012543,
+      "learning_rate": 7.460787715572085e-07,
+      "loss": 0.3113,
+      "step": 967
+    },
+    {
+      "epoch": 4.546203649205415,
+      "grad_norm": 0.35153132677078247,
+      "learning_rate": 7.414516258630245e-07,
+      "loss": 0.296,
+      "step": 968
+    },
+    {
+      "epoch": 4.550912301353738,
+      "grad_norm": 0.34727081656455994,
+      "learning_rate": 7.368363739502213e-07,
+      "loss": 0.3004,
+      "step": 969
+    },
+    {
+      "epoch": 4.55562095350206,
+      "grad_norm": 0.3494728207588196,
+      "learning_rate": 7.322330470336314e-07,
+      "loss": 0.2846,
+      "step": 970
+    },
+    {
+      "epoch": 4.560329605650383,
+      "grad_norm": 0.346566766500473,
+      "learning_rate": 7.276416762474373e-07,
+      "loss": 0.3206,
+      "step": 971
+    },
+    {
+      "epoch": 4.565038257798705,
+      "grad_norm": 0.34576454758644104,
+      "learning_rate": 7.230622926449565e-07,
+      "loss": 0.3072,
+      "step": 972
+    },
+    {
+      "epoch": 4.569746909947027,
+      "grad_norm": 0.35648053884506226,
+      "learning_rate": 7.184949271984299e-07,
+      "loss": 0.3186,
+      "step": 973
+    },
+    {
+      "epoch": 4.57445556209535,
+      "grad_norm": 0.36599478125572205,
+      "learning_rate": 7.139396107988195e-07,
+      "loss": 0.3108,
+      "step": 974
+    },
+    {
+      "epoch": 4.579164214243673,
+      "grad_norm": 0.3868202567100525,
+      "learning_rate": 7.093963742555899e-07,
+      "loss": 0.3204,
+      "step": 975
+    },
+    {
+      "epoch": 4.583872866391995,
+      "grad_norm": 0.3510901629924774,
+      "learning_rate": 7.04865248296508e-07,
+      "loss": 0.3177,
+      "step": 976
+    },
+    {
+      "epoch": 4.588581518540318,
+      "grad_norm": 0.37985342741012573,
+      "learning_rate": 7.003462635674291e-07,
+      "loss": 0.3227,
+      "step": 977
+    },
+    {
+      "epoch": 4.593290170688641,
+      "grad_norm": 0.3746037781238556,
+      "learning_rate": 6.958394506320948e-07,
+      "loss": 0.324,
+      "step": 978
+    },
+    {
+      "epoch": 4.597998822836963,
+      "grad_norm": 0.4108274579048157,
+      "learning_rate": 6.913448399719225e-07,
+      "loss": 0.2989,
+      "step": 979
+    },
+    {
+      "epoch": 4.602707474985285,
+      "grad_norm": 0.34104016423225403,
+      "learning_rate": 6.868624619858022e-07,
+      "loss": 0.2942,
+      "step": 980
+    },
+    {
+      "epoch": 4.607416127133608,
+      "grad_norm": 0.3577067255973816,
+      "learning_rate": 6.823923469898863e-07,
+      "loss": 0.3104,
+      "step": 981
+    },
+    {
+      "epoch": 4.61212477928193,
+      "grad_norm": 0.3417036235332489,
+      "learning_rate": 6.779345252173908e-07,
+      "loss": 0.3032,
+      "step": 982
+    },
+    {
+      "epoch": 4.616833431430253,
+      "grad_norm": 0.3480812609195709,
+      "learning_rate": 6.734890268183855e-07,
+      "loss": 0.3072,
+      "step": 983
+    },
+    {
+      "epoch": 4.621542083578576,
+      "grad_norm": 0.3750641644001007,
+      "learning_rate": 6.690558818595944e-07,
+      "loss": 0.317,
+      "step": 984
+    },
+    {
+      "epoch": 4.626250735726898,
+      "grad_norm": 0.36166512966156006,
+      "learning_rate": 6.646351203241869e-07,
+      "loss": 0.3044,
+      "step": 985
+    },
+    {
+      "epoch": 4.630959387875221,
+      "grad_norm": 0.3501746356487274,
+      "learning_rate": 6.602267721115807e-07,
+      "loss": 0.3005,
+      "step": 986
+    },
+    {
+      "epoch": 4.6356680400235435,
+      "grad_norm": 0.39372000098228455,
+      "learning_rate": 6.558308670372368e-07,
+      "loss": 0.3077,
+      "step": 987
+    },
+    {
+      "epoch": 4.640376692171865,
+      "grad_norm": 0.3307865262031555,
+      "learning_rate": 6.514474348324582e-07,
+      "loss": 0.2942,
+      "step": 988
+    },
+    {
+      "epoch": 4.645085344320188,
+      "grad_norm": 0.36881086230278015,
+      "learning_rate": 6.470765051441874e-07,
+      "loss": 0.3006,
+      "step": 989
+    },
+    {
+      "epoch": 4.649793996468511,
+      "grad_norm": 0.34105730056762695,
+      "learning_rate": 6.427181075348085e-07,
+      "loss": 0.2903,
+      "step": 990
+    },
+    {
+      "epoch": 4.654502648616833,
+      "grad_norm": 0.32339656352996826,
+      "learning_rate": 6.383722714819468e-07,
+      "loss": 0.2845,
+      "step": 991
+    },
+    {
+      "epoch": 4.659211300765156,
+      "grad_norm": 0.3328464925289154,
+      "learning_rate": 6.340390263782656e-07,
+      "loss": 0.293,
+      "step": 992
+    },
+    {
+      "epoch": 4.663919952913479,
+      "grad_norm": 0.3342823088169098,
+      "learning_rate": 6.297184015312754e-07,
+      "loss": 0.3026,
+      "step": 993
+    },
+    {
+      "epoch": 4.6686286050618016,
+      "grad_norm": 0.3467091917991638,
+      "learning_rate": 6.254104261631255e-07,
+      "loss": 0.3055,
+      "step": 994
+    },
+    {
+      "epoch": 4.6733372572101235,
+      "grad_norm": 0.3539992868900299,
+      "learning_rate": 6.21115129410416e-07,
+      "loss": 0.312,
+      "step": 995
+    },
+    {
+      "epoch": 4.678045909358446,
+      "grad_norm": 0.3500078320503235,
+      "learning_rate": 6.168325403239913e-07,
+      "loss": 0.2952,
+      "step": 996
+    },
+    {
+      "epoch": 4.682754561506769,
+      "grad_norm": 0.3606589138507843,
+      "learning_rate": 6.125626878687555e-07,
+      "loss": 0.3298,
+      "step": 997
+    },
+    {
+      "epoch": 4.687463213655091,
+      "grad_norm": 0.3727583587169647,
+      "learning_rate": 6.083056009234631e-07,
+      "loss": 0.3092,
+      "step": 998
+    },
+    {
+      "epoch": 4.692171865803414,
+      "grad_norm": 0.32863423228263855,
+      "learning_rate": 6.040613082805347e-07,
+      "loss": 0.2909,
+      "step": 999
+    },
+    {
+      "epoch": 4.696880517951737,
+      "grad_norm": 0.3464837670326233,
+      "learning_rate": 5.998298386458546e-07,
+      "loss": 0.3022,
+      "step": 1000
+    },
+    {
+      "epoch": 4.701589170100059,
+      "grad_norm": 0.3487091064453125,
+      "learning_rate": 5.95611220638582e-07,
+      "loss": 0.3026,
+      "step": 1001
+    },
+    {
+      "epoch": 4.706297822248382,
+      "grad_norm": 0.33962616324424744,
+      "learning_rate": 5.914054827909549e-07,
+      "loss": 0.2786,
+      "step": 1002
+    },
+    {
+      "epoch": 4.711006474396704,
+      "grad_norm": 0.3359411060810089,
+      "learning_rate": 5.872126535480977e-07,
+      "loss": 0.3236,
+      "step": 1003
+    },
+    {
+      "epoch": 4.715715126545026,
+      "grad_norm": 0.3856370449066162,
+      "learning_rate": 5.830327612678266e-07,
+      "loss": 0.3197,
+      "step": 1004
+    },
+    {
+      "epoch": 4.720423778693349,
+      "grad_norm": 0.36933499574661255,
+      "learning_rate": 5.788658342204628e-07,
+      "loss": 0.3065,
+      "step": 1005
+    },
+    {
+      "epoch": 4.725132430841672,
+      "grad_norm": 0.3596036732196808,
+      "learning_rate": 5.747119005886362e-07,
+      "loss": 0.3224,
+      "step": 1006
+    },
+    {
+      "epoch": 4.729841082989994,
+      "grad_norm": 0.3620109558105469,
+      "learning_rate": 5.705709884670982e-07,
+      "loss": 0.315,
+      "step": 1007
+    },
+    {
+      "epoch": 4.734549735138317,
+      "grad_norm": 0.3684552311897278,
+      "learning_rate": 5.664431258625305e-07,
+      "loss": 0.3108,
+      "step": 1008
+    },
+    {
+      "epoch": 4.73925838728664,
+      "grad_norm": 0.3661479949951172,
+      "learning_rate": 5.623283406933535e-07,
+      "loss": 0.3171,
+      "step": 1009
+    },
+    {
+      "epoch": 4.743967039434962,
+      "grad_norm": 0.3574175238609314,
+      "learning_rate": 5.582266607895423e-07,
+      "loss": 0.3035,
+      "step": 1010
+    },
+    {
+      "epoch": 4.748675691583284,
+      "grad_norm": 0.3499114513397217,
+      "learning_rate": 5.541381138924326e-07,
+      "loss": 0.306,
+      "step": 1011
+    },
+    {
+      "epoch": 4.753384343731607,
+      "grad_norm": 0.32387298345565796,
+      "learning_rate": 5.500627276545406e-07,
+      "loss": 0.2865,
+      "step": 1012
+    },
+    {
+      "epoch": 4.758092995879929,
+      "grad_norm": 0.3871992230415344,
+      "learning_rate": 5.460005296393672e-07,
+      "loss": 0.3512,
+      "step": 1013
+    },
+    {
+      "epoch": 4.762801648028252,
+      "grad_norm": 0.3580347001552582,
+      "learning_rate": 5.419515473212192e-07,
+      "loss": 0.3052,
+      "step": 1014
+    },
+    {
+      "epoch": 4.767510300176575,
+      "grad_norm": 0.34808874130249023,
+      "learning_rate": 5.379158080850164e-07,
+      "loss": 0.309,
+      "step": 1015
+    },
+    {
+      "epoch": 4.772218952324897,
+      "grad_norm": 0.35262778401374817,
+      "learning_rate": 5.338933392261158e-07,
+      "loss": 0.2998,
+      "step": 1016
+    },
+    {
+      "epoch": 4.77692760447322,
+      "grad_norm": 0.36279359459877014,
+      "learning_rate": 5.298841679501163e-07,
+      "loss": 0.3137,
+      "step": 1017
+    },
+    {
+      "epoch": 4.7816362566215425,
+      "grad_norm": 0.36099672317504883,
+      "learning_rate": 5.258883213726829e-07,
+      "loss": 0.3212,
+      "step": 1018
+    },
+    {
+      "epoch": 4.786344908769864,
+      "grad_norm": 0.37556588649749756,
+      "learning_rate": 5.219058265193577e-07,
+      "loss": 0.3251,
+      "step": 1019
+    },
+    {
+      "epoch": 4.791053560918187,
+      "grad_norm": 0.3478699028491974,
+      "learning_rate": 5.179367103253821e-07,
+      "loss": 0.3224,
+      "step": 1020
+    },
+    {
+      "epoch": 4.79576221306651,
+      "grad_norm": 0.35103946924209595,
+      "learning_rate": 5.139809996355111e-07,
+      "loss": 0.2948,
+      "step": 1021
+    },
+    {
+      "epoch": 4.800470865214832,
+      "grad_norm": 0.33818110823631287,
+      "learning_rate": 5.100387212038324e-07,
+      "loss": 0.309,
+      "step": 1022
+    },
+    {
+      "epoch": 4.805179517363155,
+      "grad_norm": 0.371110737323761,
+      "learning_rate": 5.061099016935872e-07,
+      "loss": 0.3232,
+      "step": 1023
+    },
+    {
+      "epoch": 4.809888169511478,
+      "grad_norm": 0.3651929497718811,
+      "learning_rate": 5.02194567676986e-07,
+      "loss": 0.2903,
+      "step": 1024
+    },
+    {
+      "epoch": 4.8145968216598,
+      "grad_norm": 0.4997953772544861,
+      "learning_rate": 4.982927456350339e-07,
+      "loss": 0.309,
+      "step": 1025
+    },
+    {
+      "epoch": 4.8193054738081225,
+      "grad_norm": 0.33667755126953125,
+      "learning_rate": 4.944044619573482e-07,
+      "loss": 0.2909,
+      "step": 1026
+    },
+    {
+      "epoch": 4.824014125956445,
+      "grad_norm": 0.34498387575149536,
+      "learning_rate": 4.905297429419809e-07,
+      "loss": 0.2959,
+      "step": 1027
+    },
+    {
+      "epoch": 4.828722778104767,
+      "grad_norm": 0.3375857174396515,
+      "learning_rate": 4.866686147952388e-07,
+      "loss": 0.2796,
+      "step": 1028
+    },
+    {
+      "epoch": 4.83343143025309,
+      "grad_norm": 0.3379746079444885,
+      "learning_rate": 4.828211036315109e-07,
+      "loss": 0.2916,
+      "step": 1029
+    },
+    {
+      "epoch": 4.838140082401413,
+      "grad_norm": 0.35453835129737854,
+      "learning_rate": 4.789872354730873e-07,
+      "loss": 0.3017,
+      "step": 1030
+    },
+    {
+      "epoch": 4.842848734549735,
+      "grad_norm": 0.3590046465396881,
+      "learning_rate": 4.7516703624998577e-07,
+      "loss": 0.3063,
+      "step": 1031
+    },
+    {
+      "epoch": 4.847557386698058,
+      "grad_norm": 0.3402198851108551,
+      "learning_rate": 4.713605317997741e-07,
+      "loss": 0.3033,
+      "step": 1032
+    },
+    {
+      "epoch": 4.8522660388463805,
+      "grad_norm": 0.37940713763237,
+      "learning_rate": 4.675677478673987e-07,
+      "loss": 0.2982,
+      "step": 1033
+    },
+    {
+      "epoch": 4.8569746909947025,
+      "grad_norm": 0.34546539187431335,
+      "learning_rate": 4.637887101050054e-07,
+      "loss": 0.3008,
+      "step": 1034
+    },
+    {
+      "epoch": 4.861683343143025,
+      "grad_norm": 0.36888861656188965,
+      "learning_rate": 4.6002344407177377e-07,
+      "loss": 0.325,
+      "step": 1035
+    },
+    {
+      "epoch": 4.866391995291348,
+      "grad_norm": 0.35262590646743774,
+      "learning_rate": 4.5627197523373494e-07,
+      "loss": 0.2916,
+      "step": 1036
+    },
+    {
+      "epoch": 4.87110064743967,
+      "grad_norm": 0.35279375314712524,
+      "learning_rate": 4.5253432896360667e-07,
+      "loss": 0.3095,
+      "step": 1037
+    },
+    {
+      "epoch": 4.875809299587993,
+      "grad_norm": 0.36938080191612244,
+      "learning_rate": 4.4881053054061875e-07,
+      "loss": 0.3157,
+      "step": 1038
+    },
+    {
+      "epoch": 4.880517951736316,
+      "grad_norm": 0.5250746607780457,
+      "learning_rate": 4.451006051503406e-07,
+      "loss": 0.3154,
+      "step": 1039
+    },
+    {
+      "epoch": 4.885226603884638,
+      "grad_norm": 0.37330305576324463,
+      "learning_rate": 4.414045778845144e-07,
+      "loss": 0.3022,
+      "step": 1040
+    },
+    {
+      "epoch": 4.8899352560329605,
+      "grad_norm": 0.337289422750473,
+      "learning_rate": 4.3772247374088363e-07,
+      "loss": 0.2992,
+      "step": 1041
+    },
+    {
+      "epoch": 4.894643908181283,
+      "grad_norm": 0.3455512225627899,
+      "learning_rate": 4.340543176230233e-07,
+      "loss": 0.2925,
+      "step": 1042
+    },
+    {
+      "epoch": 4.899352560329605,
+      "grad_norm": 0.37175658345222473,
+      "learning_rate": 4.3040013434017203e-07,
+      "loss": 0.3187,
+      "step": 1043
+    },
+    {
+      "epoch": 4.904061212477928,
+      "grad_norm": 0.3691946268081665,
+      "learning_rate": 4.2675994860706477e-07,
+      "loss": 0.2933,
+      "step": 1044
+    },
+    {
+      "epoch": 4.908769864626251,
+      "grad_norm": 0.401774138212204,
+      "learning_rate": 4.2313378504376584e-07,
+      "loss": 0.3189,
+      "step": 1045
+    },
+    {
+      "epoch": 4.913478516774573,
+      "grad_norm": 0.3765166699886322,
+      "learning_rate": 4.1952166817550184e-07,
+      "loss": 0.318,
+      "step": 1046
+    },
+    {
+      "epoch": 4.918187168922896,
+      "grad_norm": 0.3326210677623749,
+      "learning_rate": 4.1592362243249443e-07,
+      "loss": 0.291,
+      "step": 1047
+    },
+    {
+      "epoch": 4.922895821071219,
+      "grad_norm": 0.35546964406967163,
+      "learning_rate": 4.123396721497977e-07,
+      "loss": 0.3273,
+      "step": 1048
+    },
+    {
+      "epoch": 4.9276044732195405,
+      "grad_norm": 0.37602460384368896,
+      "learning_rate": 4.0876984156713264e-07,
+      "loss": 0.325,
+      "step": 1049
+    },
+    {
+      "epoch": 4.932313125367863,
+      "grad_norm": 0.3632354736328125,
+      "learning_rate": 4.0521415482872305e-07,
+      "loss": 0.2885,
+      "step": 1050
+    },
+    {
+      "epoch": 4.937021777516186,
+      "grad_norm": 0.3401520848274231,
+      "learning_rate": 4.0167263598312997e-07,
+      "loss": 0.3023,
+      "step": 1051
+    },
+    {
+      "epoch": 4.941730429664508,
+      "grad_norm": 0.3465152084827423,
+      "learning_rate": 3.981453089830936e-07,
+      "loss": 0.317,
+      "step": 1052
+    },
+    {
+      "epoch": 4.946439081812831,
+      "grad_norm": 0.35135287046432495,
+      "learning_rate": 3.9463219768536814e-07,
+      "loss": 0.3104,
+      "step": 1053
+    },
+    {
+      "epoch": 4.951147733961154,
+      "grad_norm": 0.3682619631290436,
+      "learning_rate": 3.9113332585056177e-07,
+      "loss": 0.3014,
+      "step": 1054
+    },
+    {
+      "epoch": 4.955856386109476,
+      "grad_norm": 0.34490036964416504,
+      "learning_rate": 3.8764871714297324e-07,
+      "loss": 0.3045,
+      "step": 1055
+    },
+    {
+      "epoch": 4.960565038257799,
+      "grad_norm": 0.3499850332736969,
+      "learning_rate": 3.841783951304365e-07,
+      "loss": 0.3049,
+      "step": 1056
+    },
+    {
+      "epoch": 4.965273690406121,
+      "grad_norm": 0.34722211956977844,
+      "learning_rate": 3.8072238328415753e-07,
+      "loss": 0.3027,
+      "step": 1057
+    },
+    {
+      "epoch": 4.969982342554443,
+      "grad_norm": 0.3570912480354309,
+      "learning_rate": 3.7728070497855595e-07,
+      "loss": 0.2905,
+      "step": 1058
+    },
+    {
+      "epoch": 4.974690994702766,
+      "grad_norm": 0.34860408306121826,
+      "learning_rate": 3.738533834911104e-07,
+      "loss": 0.3028,
+      "step": 1059
+    },
+    {
+      "epoch": 4.979399646851089,
+      "grad_norm": 0.362990140914917,
+      "learning_rate": 3.7044044200219566e-07,
+      "loss": 0.3244,
+      "step": 1060
+    },
+    {
+      "epoch": 4.984108298999411,
+      "grad_norm": 0.3437008261680603,
+      "learning_rate": 3.6704190359493075e-07,
+      "loss": 0.298,
+      "step": 1061
+    },
+    {
+      "epoch": 4.988816951147734,
+      "grad_norm": 0.33662015199661255,
+      "learning_rate": 3.6365779125501875e-07,
+      "loss": 0.2919,
+      "step": 1062
+    },
+    {
+      "epoch": 4.993525603296057,
+      "grad_norm": 0.33186161518096924,
+      "learning_rate": 3.602881278705953e-07,
+      "loss": 0.29,
+      "step": 1063
+    },
+    {
+      "epoch": 4.998234255444379,
+      "grad_norm": 0.3635331094264984,
+      "learning_rate": 3.5693293623207084e-07,
+      "loss": 0.2941,
+      "step": 1064
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.3635331094264984,
+      "learning_rate": 3.5359223903197846e-07,
+      "loss": 0.333,
+      "step": 1065
+    },
+    {
+      "epoch": 5.004708652148323,
+      "grad_norm": 0.6811550259590149,
+      "learning_rate": 3.5026605886481743e-07,
+      "loss": 0.2784,
+      "step": 1066
+    },
+    {
+      "epoch": 5.009417304296645,
+      "grad_norm": 0.3763936460018158,
+      "learning_rate": 3.4695441822690414e-07,
+      "loss": 0.2991,
+      "step": 1067
+    },
+    {
+      "epoch": 5.014125956444968,
+      "grad_norm": 0.3745104968547821,
+      "learning_rate": 3.4365733951621796e-07,
+      "loss": 0.285,
+      "step": 1068
+    },
+    {
+      "epoch": 5.0188346085932904,
+      "grad_norm": 0.3565599322319031,
+      "learning_rate": 3.4037484503225e-07,
+      "loss": 0.3127,
+      "step": 1069
+    },
+    {
+      "epoch": 5.023543260741612,
+      "grad_norm": 0.36555394530296326,
+      "learning_rate": 3.3710695697585115e-07,
+      "loss": 0.2864,
+      "step": 1070
+    },
+    {
+      "epoch": 5.028251912889935,
+      "grad_norm": 0.3286389708518982,
+      "learning_rate": 3.338536974490841e-07,
+      "loss": 0.2963,
+      "step": 1071
+    },
+    {
+      "epoch": 5.032960565038258,
+      "grad_norm": 0.36263808608055115,
+      "learning_rate": 3.306150884550732e-07,
+      "loss": 0.2767,
+      "step": 1072
+    },
+    {
+      "epoch": 5.03766921718658,
+      "grad_norm": 0.3278793394565582,
+      "learning_rate": 3.273911518978545e-07,
+      "loss": 0.3168,
+      "step": 1073
+    },
+    {
+      "epoch": 5.042377869334903,
+      "grad_norm": 0.3660132586956024,
+      "learning_rate": 3.2418190958222877e-07,
+      "loss": 0.3031,
+      "step": 1074
+    },
+    {
+      "epoch": 5.047086521483226,
+      "grad_norm": 0.3466164469718933,
+      "learning_rate": 3.2098738321361287e-07,
+      "loss": 0.2895,
+      "step": 1075
+    },
+    {
+      "epoch": 5.051795173631548,
+      "grad_norm": 0.35571199655532837,
+      "learning_rate": 3.1780759439789507e-07,
+      "loss": 0.3096,
+      "step": 1076
+    },
+    {
+      "epoch": 5.0565038257798705,
+      "grad_norm": 0.3681715428829193,
+      "learning_rate": 3.146425646412851e-07,
+      "loss": 0.2763,
+      "step": 1077
+    },
+    {
+      "epoch": 5.061212477928193,
+      "grad_norm": 0.4345608055591583,
+      "learning_rate": 3.114923153501748e-07,
+      "loss": 0.2946,
+      "step": 1078
+    },
+    {
+      "epoch": 5.065921130076515,
+      "grad_norm": 0.34654700756073,
+      "learning_rate": 3.083568678309862e-07,
+      "loss": 0.2803,
+      "step": 1079
+    },
+    {
+      "epoch": 5.070629782224838,
+      "grad_norm": 0.3382911682128906,
+      "learning_rate": 3.0523624329003327e-07,
+      "loss": 0.3039,
+      "step": 1080
+    },
+    {
+      "epoch": 5.075338434373161,
+      "grad_norm": 0.37458857893943787,
+      "learning_rate": 3.0213046283337375e-07,
+      "loss": 0.2849,
+      "step": 1081
+    },
+    {
+      "epoch": 5.080047086521483,
+      "grad_norm": 0.3375467360019684,
+      "learning_rate": 2.990395474666724e-07,
+      "loss": 0.279,
+      "step": 1082
+    },
+    {
+      "epoch": 5.084755738669806,
+      "grad_norm": 0.3381260633468628,
+      "learning_rate": 2.9596351809505125e-07,
+      "loss": 0.2917,
+      "step": 1083
+    },
+    {
+      "epoch": 5.0894643908181285,
+      "grad_norm": 0.33453017473220825,
+      "learning_rate": 2.9290239552295543e-07,
+      "loss": 0.2831,
+      "step": 1084
+    },
+    {
+      "epoch": 5.0941730429664505,
+      "grad_norm": 0.3335488736629486,
+      "learning_rate": 2.8985620045400676e-07,
+      "loss": 0.2847,
+      "step": 1085
+    },
+    {
+      "epoch": 5.098881695114773,
+      "grad_norm": 0.3345257639884949,
+      "learning_rate": 2.868249534908682e-07,
+      "loss": 0.2958,
+      "step": 1086
+    },
+    {
+      "epoch": 5.103590347263096,
+      "grad_norm": 0.3563317060470581,
+      "learning_rate": 2.8380867513510155e-07,
+      "loss": 0.2815,
+      "step": 1087
+    },
+    {
+      "epoch": 5.108298999411418,
+      "grad_norm": 0.34579357504844666,
+      "learning_rate": 2.8080738578703054e-07,
+      "loss": 0.294,
+      "step": 1088
+    },
+    {
+      "epoch": 5.113007651559741,
+      "grad_norm": 0.35512229800224304,
+      "learning_rate": 2.778211057456018e-07,
+      "loss": 0.2861,
+      "step": 1089
+    },
+    {
+      "epoch": 5.117716303708064,
+      "grad_norm": 0.35786810517311096,
+      "learning_rate": 2.7484985520824655e-07,
+      "loss": 0.2982,
+      "step": 1090
+    },
+    {
+      "epoch": 5.122424955856386,
+      "grad_norm": 0.38589707016944885,
+      "learning_rate": 2.7189365427074757e-07,
+      "loss": 0.3011,
+      "step": 1091
+    },
+    {
+      "epoch": 5.1271336080047085,
+      "grad_norm": 0.3488099277019501,
+      "learning_rate": 2.6895252292709977e-07,
+      "loss": 0.2959,
+      "step": 1092
+    },
+    {
+      "epoch": 5.131842260153031,
+      "grad_norm": 0.34902307391166687,
+      "learning_rate": 2.660264810693772e-07,
+      "loss": 0.2982,
+      "step": 1093
+    },
+    {
+      "epoch": 5.136550912301354,
+      "grad_norm": 0.3481535017490387,
+      "learning_rate": 2.6311554848759524e-07,
+      "loss": 0.2757,
+      "step": 1094
+    },
+    {
+      "epoch": 5.141259564449676,
+      "grad_norm": 0.3484564423561096,
+      "learning_rate": 2.602197448695823e-07,
+      "loss": 0.3075,
+      "step": 1095
+    },
+    {
+      "epoch": 5.145968216597999,
+      "grad_norm": 0.3666880130767822,
+      "learning_rate": 2.573390898008399e-07,
+      "loss": 0.2895,
+      "step": 1096
+    },
+    {
+      "epoch": 5.150676868746322,
+      "grad_norm": 0.34689509868621826,
+      "learning_rate": 2.544736027644176e-07,
+      "loss": 0.2761,
+      "step": 1097
+    },
+    {
+      "epoch": 5.155385520894644,
+      "grad_norm": 0.518316388130188,
+      "learning_rate": 2.516233031407739e-07,
+      "loss": 0.2793,
+      "step": 1098
+    },
+    {
+      "epoch": 5.160094173042967,
+      "grad_norm": 0.33427321910858154,
+      "learning_rate": 2.4878821020765136e-07,
+      "loss": 0.2726,
+      "step": 1099
+    },
+    {
+      "epoch": 5.164802825191289,
+      "grad_norm": 0.37131553888320923,
+      "learning_rate": 2.459683431399404e-07,
+      "loss": 0.2972,
+      "step": 1100
+    },
+    {
+      "epoch": 5.169511477339611,
+      "grad_norm": 0.3701353967189789,
+      "learning_rate": 2.431637210095564e-07,
+      "loss": 0.2999,
+      "step": 1101
+    },
+    {
+      "epoch": 5.174220129487934,
+      "grad_norm": 0.3545023500919342,
+      "learning_rate": 2.403743627853039e-07,
+      "loss": 0.2899,
+      "step": 1102
+    },
+    {
+      "epoch": 5.178928781636257,
+      "grad_norm": 0.34868332743644714,
+      "learning_rate": 2.376002873327532e-07,
+      "loss": 0.2826,
+      "step": 1103
+    },
+    {
+      "epoch": 5.183637433784579,
+      "grad_norm": 0.33814677596092224,
+      "learning_rate": 2.348415134141102e-07,
+      "loss": 0.2897,
+      "step": 1104
+    },
+    {
+      "epoch": 5.188346085932902,
+      "grad_norm": 0.36784449219703674,
+      "learning_rate": 2.3209805968809007e-07,
+      "loss": 0.3012,
+      "step": 1105
+    },
+    {
+      "epoch": 5.193054738081225,
+      "grad_norm": 0.38873499631881714,
+      "learning_rate": 2.293699447097919e-07,
+      "loss": 0.2915,
+      "step": 1106
+    },
+    {
+      "epoch": 5.197763390229547,
+      "grad_norm": 0.3387572169303894,
+      "learning_rate": 2.2665718693057197e-07,
+      "loss": 0.2875,
+      "step": 1107
+    },
+    {
+      "epoch": 5.202472042377869,
+      "grad_norm": 0.6812853217124939,
+      "learning_rate": 2.2395980469792106e-07,
+      "loss": 0.298,
+      "step": 1108
+    },
+    {
+      "epoch": 5.207180694526192,
+      "grad_norm": 0.35651031136512756,
+      "learning_rate": 2.2127781625533612e-07,
+      "loss": 0.2699,
+      "step": 1109
+    },
+    {
+      "epoch": 5.211889346674514,
+      "grad_norm": 0.3304518163204193,
+      "learning_rate": 2.186112397422016e-07,
+      "loss": 0.3072,
+      "step": 1110
+    },
+    {
+      "epoch": 5.216597998822837,
+      "grad_norm": 0.35859474539756775,
+      "learning_rate": 2.159600931936645e-07,
+      "loss": 0.2944,
+      "step": 1111
+    },
+    {
+      "epoch": 5.22130665097116,
+      "grad_norm": 0.376880943775177,
+      "learning_rate": 2.1332439454051278e-07,
+      "loss": 0.2902,
+      "step": 1112
+    },
+    {
+      "epoch": 5.226015303119482,
+      "grad_norm": 0.3474944829940796,
+      "learning_rate": 2.1070416160905299e-07,
+      "loss": 0.2913,
+      "step": 1113
+    },
+    {
+      "epoch": 5.230723955267805,
+      "grad_norm": 0.3503904640674591,
+      "learning_rate": 2.0809941212099143e-07,
+      "loss": 0.2856,
+      "step": 1114
+    },
+    {
+      "epoch": 5.2354326074161275,
+      "grad_norm": 0.33939892053604126,
+      "learning_rate": 2.0551016369331378e-07,
+      "loss": 0.2986,
+      "step": 1115
+    },
+    {
+      "epoch": 5.240141259564449,
+      "grad_norm": 0.36083248257637024,
+      "learning_rate": 2.0293643383816564e-07,
+      "loss": 0.2937,
+      "step": 1116
+    },
+    {
+      "epoch": 5.244849911712772,
+      "grad_norm": 0.35307255387306213,
+      "learning_rate": 2.0037823996273324e-07,
+      "loss": 0.2882,
+      "step": 1117
+    },
+    {
+      "epoch": 5.249558563861095,
+      "grad_norm": 0.33921727538108826,
+      "learning_rate": 1.9783559936912777e-07,
+      "loss": 0.2891,
+      "step": 1118
+    },
+    {
+      "epoch": 5.254267216009417,
+      "grad_norm": 0.35980817675590515,
+      "learning_rate": 1.9530852925426614e-07,
+      "loss": 0.2823,
+      "step": 1119
+    },
+    {
+      "epoch": 5.25897586815774,
+      "grad_norm": 0.4645998477935791,
+      "learning_rate": 1.927970467097573e-07,
+      "loss": 0.318,
+      "step": 1120
+    },
+    {
+      "epoch": 5.263684520306063,
+      "grad_norm": 0.37159132957458496,
+      "learning_rate": 1.9030116872178317e-07,
+      "loss": 0.2979,
+      "step": 1121
+    },
+    {
+      "epoch": 5.268393172454385,
+      "grad_norm": 0.3535681962966919,
+      "learning_rate": 1.878209121709873e-07,
+      "loss": 0.2854,
+      "step": 1122
+    },
+    {
+      "epoch": 5.2731018246027075,
+      "grad_norm": 0.36089959740638733,
+      "learning_rate": 1.853562938323586e-07,
+      "loss": 0.2858,
+      "step": 1123
+    },
+    {
+      "epoch": 5.27781047675103,
+      "grad_norm": 0.34678971767425537,
+      "learning_rate": 1.8290733037511721e-07,
+      "loss": 0.2914,
+      "step": 1124
+    },
+    {
+      "epoch": 5.282519128899352,
+      "grad_norm": 0.33591699600219727,
+      "learning_rate": 1.8047403836260523e-07,
+      "loss": 0.276,
+      "step": 1125
+    },
+    {
+      "epoch": 5.287227781047675,
+      "grad_norm": 0.34570711851119995,
+      "learning_rate": 1.7805643425216984e-07,
+      "loss": 0.2921,
+      "step": 1126
+    },
+    {
+      "epoch": 5.291936433195998,
+      "grad_norm": 0.3546205759048462,
+      "learning_rate": 1.7565453439505664e-07,
+      "loss": 0.2616,
+      "step": 1127
+    },
+    {
+      "epoch": 5.29664508534432,
+      "grad_norm": 0.3535788655281067,
+      "learning_rate": 1.7326835503629542e-07,
+      "loss": 0.2924,
+      "step": 1128
+    },
+    {
+      "epoch": 5.301353737492643,
+      "grad_norm": 0.36527425050735474,
+      "learning_rate": 1.7089791231459235e-07,
+      "loss": 0.3107,
+      "step": 1129
+    },
+    {
+      "epoch": 5.3060623896409655,
+      "grad_norm": 0.36861780285835266,
+      "learning_rate": 1.6854322226222103e-07,
+      "loss": 0.2903,
+      "step": 1130
+    },
+    {
+      "epoch": 5.3107710417892875,
+      "grad_norm": 0.381045401096344,
+      "learning_rate": 1.6620430080491217e-07,
+      "loss": 0.2782,
+      "step": 1131
+    },
+    {
+      "epoch": 5.31547969393761,
+      "grad_norm": 0.35286715626716614,
+      "learning_rate": 1.6388116376174768e-07,
+      "loss": 0.2961,
+      "step": 1132
+    },
+    {
+      "epoch": 5.320188346085933,
+      "grad_norm": 0.3706744313240051,
+      "learning_rate": 1.6157382684505258e-07,
+      "loss": 0.3087,
+      "step": 1133
+    },
+    {
+      "epoch": 5.324896998234255,
+      "grad_norm": 0.3698272705078125,
+      "learning_rate": 1.5928230566028934e-07,
+      "loss": 0.2909,
+      "step": 1134
+    },
+    {
+      "epoch": 5.329605650382578,
+      "grad_norm": 0.3476864993572235,
+      "learning_rate": 1.5700661570595326e-07,
+      "loss": 0.299,
+      "step": 1135
+    },
+    {
+      "epoch": 5.334314302530901,
+      "grad_norm": 0.3664033114910126,
+      "learning_rate": 1.5474677237346468e-07,
+      "loss": 0.2937,
+      "step": 1136
+    },
+    {
+      "epoch": 5.339022954679223,
+      "grad_norm": 0.3502042889595032,
+      "learning_rate": 1.5250279094706778e-07,
+      "loss": 0.2941,
+      "step": 1137
+    },
+    {
+      "epoch": 5.3437316068275456,
+      "grad_norm": 0.3417883515357971,
+      "learning_rate": 1.5027468660372608e-07,
+      "loss": 0.2731,
+      "step": 1138
+    },
+    {
+      "epoch": 5.348440258975868,
+      "grad_norm": 0.32379576563835144,
+      "learning_rate": 1.4806247441302046e-07,
+      "loss": 0.2948,
+      "step": 1139
+    },
+    {
+      "epoch": 5.35314891112419,
+      "grad_norm": 0.3409654498100281,
+      "learning_rate": 1.4586616933704528e-07,
+      "loss": 0.2758,
+      "step": 1140
+    },
+    {
+      "epoch": 5.357857563272513,
+      "grad_norm": 0.3374674320220947,
+      "learning_rate": 1.4368578623030982e-07,
+      "loss": 0.3127,
+      "step": 1141
+    },
+    {
+      "epoch": 5.362566215420836,
+      "grad_norm": 0.3632641136646271,
+      "learning_rate": 1.4152133983963644e-07,
+      "loss": 0.2971,
+      "step": 1142
+    },
+    {
+      "epoch": 5.367274867569158,
+      "grad_norm": 0.3576749861240387,
+      "learning_rate": 1.3937284480405987e-07,
+      "loss": 0.2947,
+      "step": 1143
+    },
+    {
+      "epoch": 5.371983519717481,
+      "grad_norm": 0.3590938448905945,
+      "learning_rate": 1.3724031565473112e-07,
+      "loss": 0.2895,
+      "step": 1144
+    },
+    {
+      "epoch": 5.376692171865804,
+      "grad_norm": 0.3656778037548065,
+      "learning_rate": 1.3512376681481558e-07,
+      "loss": 0.2841,
+      "step": 1145
+    },
+    {
+      "epoch": 5.381400824014126,
+      "grad_norm": 0.35744982957839966,
+      "learning_rate": 1.3302321259939883e-07,
+      "loss": 0.289,
+      "step": 1146
+    },
+    {
+      "epoch": 5.386109476162448,
+      "grad_norm": 0.3551943004131317,
+      "learning_rate": 1.3093866721538617e-07,
+      "loss": 0.293,
+      "step": 1147
+    },
+    {
+      "epoch": 5.390818128310771,
+      "grad_norm": 0.3526786267757416,
+      "learning_rate": 1.2887014476141214e-07,
+      "loss": 0.3119,
+      "step": 1148
+    },
+    {
+      "epoch": 5.395526780459093,
+      "grad_norm": 0.37751346826553345,
+      "learning_rate": 1.268176592277376e-07,
+      "loss": 0.3053,
+      "step": 1149
+    },
+    {
+      "epoch": 5.400235432607416,
+      "grad_norm": 0.3662261962890625,
+      "learning_rate": 1.2478122449616215e-07,
+      "loss": 0.2832,
+      "step": 1150
+    },
+    {
+      "epoch": 5.404944084755739,
+      "grad_norm": 0.3833887279033661,
+      "learning_rate": 1.2276085433992557e-07,
+      "loss": 0.2971,
+      "step": 1151
+    },
+    {
+      "epoch": 5.409652736904061,
+      "grad_norm": 0.36077573895454407,
+      "learning_rate": 1.2075656242361732e-07,
+      "loss": 0.2592,
+      "step": 1152
+    },
+    {
+      "epoch": 5.414361389052384,
+      "grad_norm": 0.32590293884277344,
+      "learning_rate": 1.1876836230308275e-07,
+      "loss": 0.2997,
+      "step": 1153
+    },
+    {
+      "epoch": 5.4190700412007065,
+      "grad_norm": 0.3507887125015259,
+      "learning_rate": 1.1679626742533201e-07,
+      "loss": 0.2917,
+      "step": 1154
+    },
+    {
+      "epoch": 5.423778693349028,
+      "grad_norm": 0.35047176480293274,
+      "learning_rate": 1.1484029112844797e-07,
+      "loss": 0.3099,
+      "step": 1155
+    },
+    {
+      "epoch": 5.428487345497351,
+      "grad_norm": 0.40482771396636963,
+      "learning_rate": 1.1290044664149874e-07,
+      "loss": 0.2895,
+      "step": 1156
+    },
+    {
+      "epoch": 5.433195997645674,
+      "grad_norm": 0.3553638160228729,
+      "learning_rate": 1.1097674708444472e-07,
+      "loss": 0.2957,
+      "step": 1157
+    },
+    {
+      "epoch": 5.437904649793996,
+      "grad_norm": 0.355226069688797,
+      "learning_rate": 1.0906920546805255e-07,
+      "loss": 0.3006,
+      "step": 1158
+    },
+    {
+      "epoch": 5.442613301942319,
+      "grad_norm": 0.37443095445632935,
+      "learning_rate": 1.07177834693806e-07,
+      "loss": 0.2902,
+      "step": 1159
+    },
+    {
+      "epoch": 5.447321954090642,
+      "grad_norm": 0.3441007435321808,
+      "learning_rate": 1.0530264755381826e-07,
+      "loss": 0.2611,
+      "step": 1160
+    },
+    {
+      "epoch": 5.452030606238964,
+      "grad_norm": 0.33422932028770447,
+      "learning_rate": 1.0344365673074674e-07,
+      "loss": 0.283,
+      "step": 1161
+    },
+    {
+      "epoch": 5.4567392583872865,
+      "grad_norm": 0.34415027499198914,
+      "learning_rate": 1.0160087479770514e-07,
+      "loss": 0.2868,
+      "step": 1162
+    },
+    {
+      "epoch": 5.461447910535609,
+      "grad_norm": 0.36474141478538513,
+      "learning_rate": 9.977431421818229e-08,
+      "loss": 0.3002,
+      "step": 1163
+    },
+    {
+      "epoch": 5.466156562683932,
+      "grad_norm": 0.4352918565273285,
+      "learning_rate": 9.796398734595285e-08,
+      "loss": 0.3022,
+      "step": 1164
+    },
+    {
+      "epoch": 5.470865214832254,
+      "grad_norm": 0.3702634572982788,
+      "learning_rate": 9.616990642499796e-08,
+      "loss": 0.2976,
+      "step": 1165
+    },
+    {
+      "epoch": 5.475573866980577,
+      "grad_norm": 0.3480238616466522,
+      "learning_rate": 9.43920835894191e-08,
+      "loss": 0.2856,
+      "step": 1166
+    },
+    {
+      "epoch": 5.4802825191289,
+      "grad_norm": 0.378632515668869,
+      "learning_rate": 9.263053086336044e-08,
+      "loss": 0.2947,
+      "step": 1167
+    },
+    {
+      "epoch": 5.484991171277222,
+      "grad_norm": 0.39319896697998047,
+      "learning_rate": 9.088526016092142e-08,
+      "loss": 0.2951,
+      "step": 1168
+    },
+    {
+      "epoch": 5.4896998234255445,
+      "grad_norm": 0.3514487147331238,
+      "learning_rate": 8.915628328608206e-08,
+      "loss": 0.2995,
+      "step": 1169
+    },
+    {
+      "epoch": 5.494408475573867,
+      "grad_norm": 0.38079124689102173,
+      "learning_rate": 8.744361193261913e-08,
+      "loss": 0.3163,
+      "step": 1170
+    },
+    {
+      "epoch": 5.499117127722189,
+      "grad_norm": 0.36984172463417053,
+      "learning_rate": 8.574725768402903e-08,
+      "loss": 0.2783,
+      "step": 1171
+    },
+    {
+      "epoch": 5.503825779870512,
+      "grad_norm": 0.36924082040786743,
+      "learning_rate": 8.406723201344891e-08,
+      "loss": 0.2848,
+      "step": 1172
+    },
+    {
+      "epoch": 5.508534432018835,
+      "grad_norm": 0.3524254262447357,
+      "learning_rate": 8.240354628357899e-08,
+      "loss": 0.3126,
+      "step": 1173
+    },
+    {
+      "epoch": 5.513243084167157,
+      "grad_norm": 0.37437206506729126,
+      "learning_rate": 8.075621174660626e-08,
+      "loss": 0.2956,
+      "step": 1174
+    },
+    {
+      "epoch": 5.51795173631548,
+      "grad_norm": 0.3524854779243469,
+      "learning_rate": 7.912523954412693e-08,
+      "loss": 0.2885,
+      "step": 1175
+    },
+    {
+      "epoch": 5.522660388463803,
+      "grad_norm": 0.41334500908851624,
+      "learning_rate": 7.751064070707248e-08,
+      "loss": 0.2838,
+      "step": 1176
+    },
+    {
+      "epoch": 5.5273690406121245,
+      "grad_norm": 0.33389919996261597,
+      "learning_rate": 7.591242615563487e-08,
+      "loss": 0.2923,
+      "step": 1177
+    },
+    {
+      "epoch": 5.532077692760447,
+      "grad_norm": 0.349431574344635,
+      "learning_rate": 7.433060669919307e-08,
+      "loss": 0.2814,
+      "step": 1178
+    },
+    {
+      "epoch": 5.53678634490877,
+      "grad_norm": 0.34167221188545227,
+      "learning_rate": 7.27651930362372e-08,
+      "loss": 0.3036,
+      "step": 1179
+    },
+    {
+      "epoch": 5.541494997057092,
+      "grad_norm": 0.36053913831710815,
+      "learning_rate": 7.121619575430061e-08,
+      "loss": 0.295,
+      "step": 1180
+    },
+    {
+      "epoch": 5.546203649205415,
+      "grad_norm": 0.34517520666122437,
+      "learning_rate": 6.968362532988465e-08,
+      "loss": 0.2877,
+      "step": 1181
+    },
+    {
+      "epoch": 5.550912301353738,
+      "grad_norm": 0.3514353036880493,
+      "learning_rate": 6.816749212839008e-08,
+      "loss": 0.2846,
+      "step": 1182
+    },
+    {
+      "epoch": 5.55562095350206,
+      "grad_norm": 0.4496132731437683,
+      "learning_rate": 6.666780640404436e-08,
+      "loss": 0.2776,
+      "step": 1183
+    },
+    {
+      "epoch": 5.560329605650383,
+      "grad_norm": 0.3367858827114105,
+      "learning_rate": 6.51845782998356e-08,
+      "loss": 0.3124,
+      "step": 1184
+    },
+    {
+      "epoch": 5.565038257798705,
+      "grad_norm": 0.3680709898471832,
+      "learning_rate": 6.371781784744124e-08,
+      "loss": 0.2965,
+      "step": 1185
+    },
+    {
+      "epoch": 5.569746909947027,
+      "grad_norm": 0.349960595369339,
+      "learning_rate": 6.226753496716254e-08,
+      "loss": 0.29,
+      "step": 1186
+    },
+    {
+      "epoch": 5.57445556209535,
+      "grad_norm": 0.35704490542411804,
+      "learning_rate": 6.083373946785459e-08,
+      "loss": 0.2893,
+      "step": 1187
+    },
+    {
+      "epoch": 5.579164214243673,
+      "grad_norm": 0.36921223998069763,
+      "learning_rate": 5.941644104686256e-08,
+      "loss": 0.301,
+      "step": 1188
+    },
+    {
+      "epoch": 5.583872866391995,
+      "grad_norm": 0.3648601174354553,
+      "learning_rate": 5.8015649289955014e-08,
+      "loss": 0.2765,
+      "step": 1189
+    },
+    {
+      "epoch": 5.588581518540318,
+      "grad_norm": 0.33216673135757446,
+      "learning_rate": 5.663137367125898e-08,
+      "loss": 0.2764,
+      "step": 1190
+    },
+    {
+      "epoch": 5.593290170688641,
+      "grad_norm": 0.34482541680336,
+      "learning_rate": 5.5263623553196424e-08,
+      "loss": 0.2959,
+      "step": 1191
+    },
+    {
+      "epoch": 5.597998822836963,
+      "grad_norm": 0.3605605959892273,
+      "learning_rate": 5.391240818642007e-08,
+      "loss": 0.2871,
+      "step": 1192
+    },
+    {
+      "epoch": 5.602707474985285,
+      "grad_norm": 0.36288997530937195,
+      "learning_rate": 5.257773670975214e-08,
+      "loss": 0.3013,
+      "step": 1193
+    },
+    {
+      "epoch": 5.607416127133608,
+      "grad_norm": 0.37322282791137695,
+      "learning_rate": 5.12596181501207e-08,
+      "loss": 0.2878,
+      "step": 1194
+    },
+    {
+      "epoch": 5.61212477928193,
+      "grad_norm": 0.343702495098114,
+      "learning_rate": 4.995806142250065e-08,
+      "loss": 0.2656,
+      "step": 1195
+    },
+    {
+      "epoch": 5.616833431430253,
+      "grad_norm": 0.3344155550003052,
+      "learning_rate": 4.867307532985227e-08,
+      "loss": 0.3142,
+      "step": 1196
+    },
+    {
+      "epoch": 5.621542083578576,
+      "grad_norm": 0.3713268041610718,
+      "learning_rate": 4.740466856306164e-08,
+      "loss": 0.2864,
+      "step": 1197
+    },
+    {
+      "epoch": 5.626250735726898,
+      "grad_norm": 0.34247687458992004,
+      "learning_rate": 4.615284970088174e-08,
+      "loss": 0.2865,
+      "step": 1198
+    },
+    {
+      "epoch": 5.630959387875221,
+      "grad_norm": 0.38114747405052185,
+      "learning_rate": 4.491762720987558e-08,
+      "loss": 0.3027,
+      "step": 1199
+    },
+    {
+      "epoch": 5.6356680400235435,
+      "grad_norm": 0.3579918146133423,
+      "learning_rate": 4.369900944435734e-08,
+      "loss": 0.3007,
+      "step": 1200
+    },
+    {
+      "epoch": 5.640376692171865,
+      "grad_norm": 0.35347315669059753,
+      "learning_rate": 4.249700464633744e-08,
+      "loss": 0.285,
+      "step": 1201
+    },
+    {
+      "epoch": 5.645085344320188,
+      "grad_norm": 0.35810258984565735,
+      "learning_rate": 4.1311620945465314e-08,
+      "loss": 0.287,
+      "step": 1202
+    },
+    {
+      "epoch": 5.649793996468511,
+      "grad_norm": 0.3517029881477356,
+      "learning_rate": 4.014286635897535e-08,
+      "loss": 0.3004,
+      "step": 1203
+    },
+    {
+      "epoch": 5.654502648616833,
+      "grad_norm": 0.3566516935825348,
+      "learning_rate": 3.899074879163245e-08,
+      "loss": 0.2794,
+      "step": 1204
+    },
+    {
+      "epoch": 5.659211300765156,
+      "grad_norm": 0.4844449758529663,
+      "learning_rate": 3.785527603567929e-08,
+      "loss": 0.2724,
+      "step": 1205
+    },
+    {
+      "epoch": 5.663919952913479,
+      "grad_norm": 0.3332138955593109,
+      "learning_rate": 3.673645577078111e-08,
+      "loss": 0.3026,
+      "step": 1206
+    },
+    {
+      "epoch": 5.6686286050618016,
+      "grad_norm": 0.35666659474372864,
+      "learning_rate": 3.5634295563977126e-08,
+      "loss": 0.2889,
+      "step": 1207
+    },
+    {
+      "epoch": 5.6733372572101235,
+      "grad_norm": 0.43679553270339966,
+      "learning_rate": 3.4548802869627806e-08,
+      "loss": 0.3002,
+      "step": 1208
+    },
+    {
+      "epoch": 5.678045909358446,
+      "grad_norm": 0.354524165391922,
+      "learning_rate": 3.347998502936267e-08,
+      "loss": 0.302,
+      "step": 1209
+    },
+    {
+      "epoch": 5.682754561506769,
+      "grad_norm": 0.3709380328655243,
+      "learning_rate": 3.242784927203507e-08,
+      "loss": 0.3057,
+      "step": 1210
+    },
+    {
+      "epoch": 5.687463213655091,
+      "grad_norm": 0.35389119386672974,
+      "learning_rate": 3.1392402713668056e-08,
+      "loss": 0.2975,
+      "step": 1211
+    },
+    {
+      "epoch": 5.692171865803414,
+      "grad_norm": 0.36993980407714844,
+      "learning_rate": 3.0373652357410243e-08,
+      "loss": 0.2738,
+      "step": 1212
+    },
+    {
+      "epoch": 5.696880517951737,
+      "grad_norm": 0.3319277763366699,
+      "learning_rate": 2.937160509348641e-08,
+      "loss": 0.2864,
+      "step": 1213
+    },
+    {
+      "epoch": 5.701589170100059,
+      "grad_norm": 0.3565705418586731,
+      "learning_rate": 2.838626769915226e-08,
+      "loss": 0.2783,
+      "step": 1214
+    },
+    {
+      "epoch": 5.706297822248382,
+      "grad_norm": 0.34223315119743347,
+      "learning_rate": 2.741764683864695e-08,
+      "loss": 0.2773,
+      "step": 1215
+    },
+    {
+      "epoch": 5.711006474396704,
+      "grad_norm": 0.33122357726097107,
+      "learning_rate": 2.6465749063149247e-08,
+      "loss": 0.2791,
+      "step": 1216
+    },
+    {
+      "epoch": 5.715715126545026,
+      "grad_norm": 0.3491283357143402,
+      "learning_rate": 2.553058081073312e-08,
+      "loss": 0.2573,
+      "step": 1217
+    },
+    {
+      "epoch": 5.720423778693349,
+      "grad_norm": 0.3793807923793793,
+      "learning_rate": 2.461214840632331e-08,
+      "loss": 0.3066,
+      "step": 1218
+    },
+    {
+      "epoch": 5.725132430841672,
+      "grad_norm": 0.3676687777042389,
+      "learning_rate": 2.3710458061653453e-08,
+      "loss": 0.2955,
+      "step": 1219
+    },
+    {
+      "epoch": 5.729841082989994,
+      "grad_norm": 0.3596877455711365,
+      "learning_rate": 2.2825515875224414e-08,
+      "loss": 0.3021,
+      "step": 1220
+    },
+    {
+      "epoch": 5.734549735138317,
+      "grad_norm": 0.3660774528980255,
+      "learning_rate": 2.1957327832261567e-08,
+      "loss": 0.2806,
+      "step": 1221
+    },
+    {
+      "epoch": 5.73925838728664,
+      "grad_norm": 0.3432364761829376,
+      "learning_rate": 2.110589980467537e-08,
+      "loss": 0.2792,
+      "step": 1222
+    },
+    {
+      "epoch": 5.743967039434962,
+      "grad_norm": 0.3409741520881653,
+      "learning_rate": 2.0271237551021395e-08,
+      "loss": 0.2828,
+      "step": 1223
+    },
+    {
+      "epoch": 5.748675691583284,
+      "grad_norm": 0.3487401604652405,
+      "learning_rate": 1.945334671646232e-08,
+      "loss": 0.294,
+      "step": 1224
+    },
+    {
+      "epoch": 5.753384343731607,
+      "grad_norm": 0.3659566640853882,
+      "learning_rate": 1.865223283272738e-08,
+      "loss": 0.2897,
+      "step": 1225
+    },
+    {
+      "epoch": 5.758092995879929,
+      "grad_norm": 0.3651147186756134,
+      "learning_rate": 1.7867901318077696e-08,
+      "loss": 0.3139,
+      "step": 1226
+    },
+    {
+      "epoch": 5.762801648028252,
+      "grad_norm": 0.38366273045539856,
+      "learning_rate": 1.710035747726768e-08,
+      "loss": 0.2798,
+      "step": 1227
+    },
+    {
+      "epoch": 5.767510300176575,
+      "grad_norm": 0.3514769375324249,
+      "learning_rate": 1.6349606501509796e-08,
+      "loss": 0.2884,
+      "step": 1228
+    },
+    {
+      "epoch": 5.772218952324897,
+      "grad_norm": 0.329897940158844,
+      "learning_rate": 1.561565346843985e-08,
+      "loss": 0.2697,
+      "step": 1229
+    },
+    {
+      "epoch": 5.77692760447322,
+      "grad_norm": 0.3363187909126282,
+      "learning_rate": 1.4898503342082592e-08,
+      "loss": 0.297,
+      "step": 1230
+    },
+    {
+      "epoch": 5.7816362566215425,
+      "grad_norm": 0.36480292677879333,
+      "learning_rate": 1.4198160972816732e-08,
+      "loss": 0.2828,
+      "step": 1231
+    },
+    {
+      "epoch": 5.786344908769864,
+      "grad_norm": 0.352155864238739,
+      "learning_rate": 1.351463109734441e-08,
+      "loss": 0.2747,
+      "step": 1232
+    },
+    {
+      "epoch": 5.791053560918187,
+      "grad_norm": 0.3681313693523407,
+      "learning_rate": 1.2847918338657617e-08,
+      "loss": 0.2956,
+      "step": 1233
+    },
+    {
+      "epoch": 5.79576221306651,
+      "grad_norm": 0.3519361913204193,
+      "learning_rate": 1.2198027206006823e-08,
+      "loss": 0.2964,
+      "step": 1234
+    },
+    {
+      "epoch": 5.800470865214832,
+      "grad_norm": 0.3671719431877136,
+      "learning_rate": 1.1564962094871569e-08,
+      "loss": 0.2981,
+      "step": 1235
+    },
+    {
+      "epoch": 5.805179517363155,
+      "grad_norm": 0.36352333426475525,
+      "learning_rate": 1.0948727286930194e-08,
+      "loss": 0.2731,
+      "step": 1236
+    },
+    {
+      "epoch": 5.809888169511478,
+      "grad_norm": 0.3341374695301056,
+      "learning_rate": 1.0349326950030159e-08,
+      "loss": 0.293,
+      "step": 1237
+    },
+    {
+      "epoch": 5.8145968216598,
+      "grad_norm": 0.3493800759315491,
+      "learning_rate": 9.766765138160828e-09,
+      "loss": 0.2873,
+      "step": 1238
+    },
+    {
+      "epoch": 5.8193054738081225,
+      "grad_norm": 0.3361739218235016,
+      "learning_rate": 9.201045791426555e-09,
+      "loss": 0.2787,
+      "step": 1239
+    },
+    {
+      "epoch": 5.824014125956445,
+      "grad_norm": 0.33776283264160156,
+      "learning_rate": 8.652172736017817e-09,
+      "loss": 0.3051,
+      "step": 1240
+    },
+    {
+      "epoch": 5.828722778104767,
+      "grad_norm": 0.35557821393013,
+      "learning_rate": 8.120149684187618e-09,
+      "loss": 0.2668,
+      "step": 1241
+    },
+    {
+      "epoch": 5.83343143025309,
+      "grad_norm": 0.3537602424621582,
+      "learning_rate": 7.604980234225124e-09,
+      "loss": 0.3018,
+      "step": 1242
+    },
+    {
+      "epoch": 5.838140082401413,
+      "grad_norm": 0.3534339964389801,
+      "learning_rate": 7.106667870432071e-09,
+      "loss": 0.2764,
+      "step": 1243
+    },
+    {
+      "epoch": 5.842848734549735,
+      "grad_norm": 0.3374592363834381,
+      "learning_rate": 6.625215963098896e-09,
+      "loss": 0.2806,
+      "step": 1244
+    },
+    {
+      "epoch": 5.847557386698058,
+      "grad_norm": 0.3431640565395355,
+      "learning_rate": 6.160627768481142e-09,
+      "loss": 0.3044,
+      "step": 1245
+    },
+    {
+      "epoch": 5.8522660388463805,
+      "grad_norm": 0.36093828082084656,
+      "learning_rate": 5.712906428778919e-09,
+      "loss": 0.2911,
+      "step": 1246
+    },
+    {
+      "epoch": 5.8569746909947025,
+      "grad_norm": 0.3618963658809662,
+      "learning_rate": 5.2820549721144254e-09,
+      "loss": 0.2915,
+      "step": 1247
+    },
+    {
+      "epoch": 5.861683343143025,
+      "grad_norm": 0.34805533289909363,
+      "learning_rate": 4.868076312512515e-09,
+      "loss": 0.2804,
+      "step": 1248
+    },
+    {
+      "epoch": 5.866391995291348,
+      "grad_norm": 0.344169557094574,
+      "learning_rate": 4.470973249879607e-09,
+      "loss": 0.282,
+      "step": 1249
+    },
+    {
+      "epoch": 5.87110064743967,
+      "grad_norm": 0.3552144169807434,
+      "learning_rate": 4.090748469986472e-09,
+      "loss": 0.3176,
+      "step": 1250
+    },
+    {
+      "epoch": 5.875809299587993,
+      "grad_norm": 0.352480947971344,
+      "learning_rate": 3.727404544448254e-09,
+      "loss": 0.2767,
+      "step": 1251
+    },
+    {
+      "epoch": 5.880517951736316,
+      "grad_norm": 0.3459225594997406,
+      "learning_rate": 3.3809439307086465e-09,
+      "loss": 0.2794,
+      "step": 1252
+    },
+    {
+      "epoch": 5.885226603884638,
+      "grad_norm": 0.3352890908718109,
+      "learning_rate": 3.051368972022406e-09,
+      "loss": 0.2913,
+      "step": 1253
+    },
+    {
+      "epoch": 5.8899352560329605,
+      "grad_norm": 0.3477289378643036,
+      "learning_rate": 2.7386818974395324e-09,
+      "loss": 0.3003,
+      "step": 1254
+    },
+    {
+      "epoch": 5.894643908181283,
+      "grad_norm": 0.36927521228790283,
+      "learning_rate": 2.4428848217908362e-09,
+      "loss": 0.3073,
+      "step": 1255
+    },
+    {
+      "epoch": 5.899352560329605,
+      "grad_norm": 0.37141096591949463,
+      "learning_rate": 2.1639797456723955e-09,
+      "loss": 0.2815,
+      "step": 1256
+    },
+    {
+      "epoch": 5.904061212477928,
+      "grad_norm": 0.34543880820274353,
+      "learning_rate": 1.9019685554333422e-09,
+      "loss": 0.2837,
+      "step": 1257
+    },
+    {
+      "epoch": 5.908769864626251,
+      "grad_norm": 0.4301663339138031,
+      "learning_rate": 1.6568530231628189e-09,
+      "loss": 0.2703,
+      "step": 1258
+    },
+    {
+      "epoch": 5.913478516774573,
+      "grad_norm": 0.3248118758201599,
+      "learning_rate": 1.4286348066769317e-09,
+      "loss": 0.2937,
+      "step": 1259
+    },
+    {
+      "epoch": 5.918187168922896,
+      "grad_norm": 0.35874074697494507,
+      "learning_rate": 1.2173154495087603e-09,
+      "loss": 0.2643,
+      "step": 1260
+    },
+    {
+      "epoch": 5.922895821071219,
+      "grad_norm": 0.3497905135154724,
+      "learning_rate": 1.0228963808978087e-09,
+      "loss": 0.2906,
+      "step": 1261
+    },
+    {
+      "epoch": 5.9276044732195405,
+      "grad_norm": 0.3518345057964325,
+      "learning_rate": 8.453789157794601e-10,
+      "loss": 0.3026,
+      "step": 1262
+    },
+    {
+      "epoch": 5.932313125367863,
+      "grad_norm": 0.38623765110969543,
+      "learning_rate": 6.847642547769262e-10,
+      "loss": 0.2787,
+      "step": 1263
+    },
+    {
+      "epoch": 5.937021777516186,
+      "grad_norm": 0.34783104062080383,
+      "learning_rate": 5.41053484192644e-10,
+      "loss": 0.2924,
+      "step": 1264
+    },
+    {
+      "epoch": 5.941730429664508,
+      "grad_norm": 0.34925082325935364,
+      "learning_rate": 4.14247576001614e-10,
+      "loss": 0.2857,
+      "step": 1265
+    },
+    {
+      "epoch": 5.946439081812831,
+      "grad_norm": 0.5177316665649414,
+      "learning_rate": 3.043473878436287e-10,
+      "loss": 0.305,
+      "step": 1266
+    },
+    {
+      "epoch": 5.951147733961154,
+      "grad_norm": 0.3745754063129425,
+      "learning_rate": 2.1135366301910932e-10,
+      "loss": 0.3044,
+      "step": 1267
+    },
+    {
+      "epoch": 5.955856386109476,
+      "grad_norm": 0.36931711435317993,
+      "learning_rate": 1.3526703048216683e-10,
+      "loss": 0.2991,
+      "step": 1268
+    },
+    {
+      "epoch": 5.960565038257799,
+      "grad_norm": 0.36866995692253113,
+      "learning_rate": 7.608800483782652e-11,
+      "loss": 0.3037,
+      "step": 1269
+    },
+    {
+      "epoch": 5.965273690406121,
+      "grad_norm": 0.36637187004089355,
+      "learning_rate": 3.381698633814212e-11,
+      "loss": 0.3177,
+      "step": 1270
+    },
+    {
+      "epoch": 5.969982342554443,
+      "grad_norm": 0.39666616916656494,
+      "learning_rate": 8.45426087942025e-12,
+      "loss": 0.3023,
+      "step": 1271
+    },
+    {
+      "epoch": 5.974690994702766,
+      "grad_norm": 0.3777483403682709,
+      "learning_rate": 0.0,
+      "loss": 0.2891,
+      "step": 1272
+    },
+    {
+      "epoch": 5.974690994702766,
+      "step": 1272,
+      "total_flos": 1.2865326835008799e+19,
+      "train_loss": 0.3775725749866018,
+      "train_runtime": 249298.3546,
+      "train_samples_per_second": 5.233,
+      "train_steps_per_second": 0.005
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1272,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2865326835008799e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}