diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6025 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0761421319796955,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0005,
+      "loss": 0.799,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.001,
+      "loss": 0.6675,
+      "step": 2
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000999997522701653,
+      "loss": 0.999,
+      "step": 3
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0009999900908311602,
+      "loss": 0.7497,
+      "step": 4
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0009999777044621652,
+      "loss": 0.5151,
+      "step": 5
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0009999603637174071,
+      "loss": 0.5689,
+      "step": 6
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0009999380687687187,
+      "loss": 1.0888,
+      "step": 7
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0009999108198370248,
+      "loss": 0.7286,
+      "step": 8
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0009998786171923407,
+      "loss": 0.7832,
+      "step": 9
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.000999841461153768,
+      "loss": 0.5225,
+      "step": 10
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0009997993520894936,
+      "loss": 0.6154,
+      "step": 11
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0009997522904167844,
+      "loss": 0.822,
+      "step": 12
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0009997002766019831,
+      "loss": 0.5985,
+      "step": 13
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0009996433111605053,
+      "loss": 0.6488,
+      "step": 14
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.000999581394656832,
+      "loss": 0.5519,
+      "step": 15
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0009995145277045061,
+      "loss": 0.8489,
+      "step": 16
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0009994427109661253,
+      "loss": 0.622,
+      "step": 17
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0009993659451533353,
+      "loss": 0.7171,
+      "step": 18
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0009992842310268233,
+      "loss": 0.6803,
+      "step": 19
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0009991975693963108,
+      "loss": 0.7087,
+      "step": 20
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.000999105961120544,
+      "loss": 0.7689,
+      "step": 21
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0009990094071072877,
+      "loss": 0.5209,
+      "step": 22
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000998907908313314,
+      "loss": 0.6299,
+      "step": 23
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000998801465744394,
+      "loss": 0.9556,
+      "step": 24
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0009986900804552878,
+      "loss": 0.5427,
+      "step": 25
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0009985737535497337,
+      "loss": 0.7997,
+      "step": 26
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0009984524861804376,
+      "loss": 0.7101,
+      "step": 27
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0009983262795490613,
+      "loss": 0.5568,
+      "step": 28
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0009981951349062107,
+      "loss": 0.7616,
+      "step": 29
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0009980590535514234,
+      "loss": 0.5599,
+      "step": 30
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0009979180368331559,
+      "loss": 0.5306,
+      "step": 31
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00099777208614877,
+      "loss": 0.7418,
+      "step": 32
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0009976212029445194,
+      "loss": 0.7747,
+      "step": 33
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000997465388715535,
+      "loss": 0.7886,
+      "step": 34
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00099730464500581,
+      "loss": 0.7207,
+      "step": 35
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0009971389734081848,
+      "loss": 0.7,
+      "step": 36
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0009969683755643318,
+      "loss": 0.5443,
+      "step": 37
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0009967928531647372,
+      "loss": 0.6659,
+      "step": 38
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0009966124079486872,
+      "loss": 0.6102,
+      "step": 39
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.000996427041704248,
+      "loss": 0.6723,
+      "step": 40
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0009962367562682496,
+      "loss": 0.5818,
+      "step": 41
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.000996041553526267,
+      "loss": 0.6673,
+      "step": 42
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0009958414354126022,
+      "loss": 0.7536,
+      "step": 43
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.000995636403910264,
+      "loss": 0.4539,
+      "step": 44
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0009954264610509496,
+      "loss": 0.7296,
+      "step": 45
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0009952116089150232,
+      "loss": 0.5989,
+      "step": 46
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000994991849631496,
+      "loss": 0.5712,
+      "step": 47
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0009947671853780054,
+      "loss": 0.872,
+      "step": 48
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.000994537618380793,
+      "loss": 0.7327,
+      "step": 49
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0009943031509146824,
+      "loss": 0.7512,
+      "step": 50
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0009940637853030573,
+      "loss": 0.6572,
+      "step": 51
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0009938195239178375,
+      "loss": 0.6821,
+      "step": 52
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0009935703691794564,
+      "loss": 0.6785,
+      "step": 53
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0009933163235568369,
+      "loss": 0.9043,
+      "step": 54
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0009930573895673657,
+      "loss": 0.6856,
+      "step": 55
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0009927935697768698,
+      "loss": 0.4684,
+      "step": 56
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0009925248667995907,
+      "loss": 0.5579,
+      "step": 57
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0009922512832981584,
+      "loss": 0.4323,
+      "step": 58
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0009919728219835644,
+      "loss": 1.0761,
+      "step": 59
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0009916894856151356,
+      "loss": 0.5668,
+      "step": 60
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0009914012770005072,
+      "loss": 0.6059,
+      "step": 61
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000991108198995594,
+      "loss": 0.5635,
+      "step": 62
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0009908102545045623,
+      "loss": 0.7705,
+      "step": 63
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0009905074464798022,
+      "loss": 0.8812,
+      "step": 64
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0009901997779218968,
+      "loss": 0.6145,
+      "step": 65
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0009898872518795932,
+      "loss": 0.5681,
+      "step": 66
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0009895698714497724,
+      "loss": 0.658,
+      "step": 67
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0009892476397774185,
+      "loss": 0.5969,
+      "step": 68
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0009889205600555875,
+      "loss": 0.5991,
+      "step": 69
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0009885886355253757,
+      "loss": 0.762,
+      "step": 70
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0009882518694758874,
+      "loss": 0.5969,
+      "step": 71
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0009879102652442023,
+      "loss": 0.5668,
+      "step": 72
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0009875638262153432,
+      "loss": 0.6104,
+      "step": 73
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.000987212555822241,
+      "loss": 0.5512,
+      "step": 74
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0009868564575457022,
+      "loss": 0.5551,
+      "step": 75
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0009864955349143734,
+      "loss": 0.6269,
+      "step": 76
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.000986129791504707,
+      "loss": 0.5629,
+      "step": 77
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0009857592309409247,
+      "loss": 0.7589,
+      "step": 78
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0009853838568949832,
+      "loss": 0.5747,
+      "step": 79
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0009850036730865363,
+      "loss": 0.7552,
+      "step": 80
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.000984618683282899,
+      "loss": 0.6728,
+      "step": 81
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0009842288912990096,
+      "loss": 0.6057,
+      "step": 82
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0009838343009973924,
+      "loss": 0.545,
+      "step": 83
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0009834349162881188,
+      "loss": 0.7007,
+      "step": 84
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0009830307411287697,
+      "loss": 0.5392,
+      "step": 85
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000982621779524394,
+      "loss": 0.8032,
+      "step": 86
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0009822080355274718,
+      "loss": 0.5957,
+      "step": 87
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0009817895132378724,
+      "loss": 0.5639,
+      "step": 88
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0009813662168028144,
+      "loss": 0.6488,
+      "step": 89
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0009809381504168233,
+      "loss": 0.8626,
+      "step": 90
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0009805053183216924,
+      "loss": 0.6872,
+      "step": 91
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0009800677248064382,
+      "loss": 0.6908,
+      "step": 92
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0009796253742072596,
+      "loss": 0.699,
+      "step": 93
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0009791782709074944,
+      "loss": 0.6305,
+      "step": 94
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0009787264193375754,
+      "loss": 0.8407,
+      "step": 95
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0009782698239749871,
+      "loss": 0.7651,
+      "step": 96
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0009778084893442218,
+      "loss": 0.5649,
+      "step": 97
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.000977342420016733,
+      "loss": 0.6298,
+      "step": 98
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0009768716206108921,
+      "loss": 0.6444,
+      "step": 99
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0009763960957919414,
+      "loss": 0.5879,
+      "step": 100
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0009759158502719481,
+      "loss": 0.6006,
+      "step": 101
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0009754308888097583,
+      "loss": 0.6416,
+      "step": 102
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0009749412162109486,
+      "loss": 0.8301,
+      "step": 103
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0009744468373277796,
+      "loss": 0.8659,
+      "step": 104
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0009739477570591473,
+      "loss": 0.7777,
+      "step": 105
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0009734439803505345,
+      "loss": 0.5722,
+      "step": 106
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.000972935512193962,
+      "loss": 0.6249,
+      "step": 107
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0009724223576279394,
+      "loss": 1.1485,
+      "step": 108
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0009719045217374141,
+      "loss": 0.6837,
+      "step": 109
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0009713820096537225,
+      "loss": 0.5921,
+      "step": 110
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0009708548265545375,
+      "loss": 0.8619,
+      "step": 111
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0009703229776638185,
+      "loss": 0.6175,
+      "step": 112
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0009697864682517592,
+      "loss": 0.6915,
+      "step": 113
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0009692453036347351,
+      "loss": 0.5714,
+      "step": 114
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0009686994891752507,
+      "loss": 0.5019,
+      "step": 115
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0009681490302818874,
+      "loss": 1.1504,
+      "step": 116
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0009675939324092487,
+      "loss": 0.6648,
+      "step": 117
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0009670342010579065,
+      "loss": 0.6379,
+      "step": 118
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0009664698417743475,
+      "loss": 0.6071,
+      "step": 119
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0009659008601509168,
+      "loss": 0.675,
+      "step": 120
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0009653272618257631,
+      "loss": 0.5167,
+      "step": 121
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0009647490524827833,
+      "loss": 0.7457,
+      "step": 122
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0009641662378515658,
+      "loss": 0.836,
+      "step": 123
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0009635788237073333,
+      "loss": 0.759,
+      "step": 124
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0009629868158708861,
+      "loss": 0.885,
+      "step": 125
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0009623902202085444,
+      "loss": 0.7311,
+      "step": 126
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0009617890426320899,
+      "loss": 0.7275,
+      "step": 127
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0009611832890987075,
+      "loss": 0.6473,
+      "step": 128
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0009605729656109264,
+      "loss": 0.6269,
+      "step": 129
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0009599580782165598,
+      "loss": 0.9734,
+      "step": 130
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0009593386330086457,
+      "loss": 0.6377,
+      "step": 131
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0009587146361253867,
+      "loss": 0.7208,
+      "step": 132
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0009580860937500883,
+      "loss": 0.5472,
+      "step": 133
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0009574530121110989,
+      "loss": 0.8343,
+      "step": 134
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0009568153974817464,
+      "loss": 0.6048,
+      "step": 135
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0009561732561802779,
+      "loss": 0.5346,
+      "step": 136
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0009555265945697953,
+      "loss": 0.6295,
+      "step": 137
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0009548754190581938,
+      "loss": 0.6079,
+      "step": 138
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0009542197360980978,
+      "loss": 0.5289,
+      "step": 139
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0009535595521867959,
+      "loss": 0.7579,
+      "step": 140
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0009528948738661784,
+      "loss": 0.8158,
+      "step": 141
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0009522257077226717,
+      "loss": 0.6023,
+      "step": 142
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0009515520603871719,
+      "loss": 0.6647,
+      "step": 143
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0009508739385349812,
+      "loss": 1.1108,
+      "step": 144
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0009501913488857399,
+      "loss": 0.9834,
+      "step": 145
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0009495042982033611,
+      "loss": 0.9049,
+      "step": 146
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0009488127932959625,
+      "loss": 0.6604,
+      "step": 147
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0009481168410158003,
+      "loss": 0.8768,
+      "step": 148
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0009474164482592001,
+      "loss": 0.6579,
+      "step": 149
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0009467116219664893,
+      "loss": 0.7038,
+      "step": 150
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0009460023691219276,
+      "loss": 0.6726,
+      "step": 151
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0009452886967536389,
+      "loss": 1.1224,
+      "step": 152
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0009445706119335407,
+      "loss": 0.7254,
+      "step": 153
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0009438481217772743,
+      "loss": 0.6464,
+      "step": 154
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0009431212334441342,
+      "loss": 1.0653,
+      "step": 155
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0009423899541369978,
+      "loss": 0.7072,
+      "step": 156
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.000941654291102253,
+      "loss": 0.9562,
+      "step": 157
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0009409142516297269,
+      "loss": 0.5401,
+      "step": 158
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0009401698430526141,
+      "loss": 0.628,
+      "step": 159
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0009394210727474029,
+      "loss": 0.6226,
+      "step": 160
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0009386679481338031,
+      "loss": 0.8939,
+      "step": 161
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0009379104766746722,
+      "loss": 0.614,
+      "step": 162
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0009371486658759415,
+      "loss": 0.6249,
+      "step": 163
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0009363825232865413,
+      "loss": 0.5156,
+      "step": 164
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0009356120564983266,
+      "loss": 0.6214,
+      "step": 165
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0009348372731460022,
+      "loss": 0.6466,
+      "step": 166
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0009340581809070458,
+      "loss": 1.3654,
+      "step": 167
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0009332747875016332,
+      "loss": 0.7355,
+      "step": 168
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0009324871006925612,
+      "loss": 0.6488,
+      "step": 169
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0009316951282851706,
+      "loss": 0.6619,
+      "step": 170
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0009308988781272693,
+      "loss": 0.7481,
+      "step": 171
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0009300983581090541,
+      "loss": 0.7286,
+      "step": 172
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0009292935761630325,
+      "loss": 0.6915,
+      "step": 173
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0009284845402639446,
+      "loss": 0.7665,
+      "step": 174
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0009276712584286833,
+      "loss": 0.5691,
+      "step": 175
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.000926853738716216,
+      "loss": 0.9606,
+      "step": 176
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0009260319892275033,
+      "loss": 0.6955,
+      "step": 177
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00092520601810542,
+      "loss": 0.7575,
+      "step": 178
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0009243758335346734,
+      "loss": 0.644,
+      "step": 179
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0009235414437417234,
+      "loss": 0.6696,
+      "step": 180
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0009227028569946996,
+      "loss": 0.7746,
+      "step": 181
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0009218600816033201,
+      "loss": 0.7244,
+      "step": 182
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0009210131259188095,
+      "loss": 0.6162,
+      "step": 183
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0009201619983338152,
+      "loss": 1.1655,
+      "step": 184
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0009193067072823251,
+      "loss": 0.7024,
+      "step": 185
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.000918447261239584,
+      "loss": 0.8246,
+      "step": 186
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0009175836687220084,
+      "loss": 0.8583,
+      "step": 187
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0009167159382871039,
+      "loss": 0.6455,
+      "step": 188
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.000915844078533379,
+      "loss": 0.7452,
+      "step": 189
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0009149680981002608,
+      "loss": 0.6813,
+      "step": 190
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0009140880056680088,
+      "loss": 0.652,
+      "step": 191
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.000913203809957629,
+      "loss": 0.6686,
+      "step": 192
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0009123155197307875,
+      "loss": 0.7548,
+      "step": 193
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0009114231437897244,
+      "loss": 0.8523,
+      "step": 194
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0009105266909771652,
+      "loss": 0.5544,
+      "step": 195
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0009096261701762343,
+      "loss": 0.6712,
+      "step": 196
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0009087215903103663,
+      "loss": 0.4564,
+      "step": 197
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0009078129603432181,
+      "loss": 0.3499,
+      "step": 198
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0009069002892785796,
+      "loss": 0.6096,
+      "step": 199
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0009059835861602853,
+      "loss": 0.1973,
+      "step": 200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0009050628600721233,
+      "loss": 0.4716,
+      "step": 201
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0009041381201377467,
+      "loss": 0.443,
+      "step": 202
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0009032093755205823,
+      "loss": 0.4729,
+      "step": 203
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0009022766354237399,
+      "loss": 0.373,
+      "step": 204
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0009013399090899217,
+      "loss": 0.4033,
+      "step": 205
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0009003992058013302,
+      "loss": 0.3282,
+      "step": 206
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0008994545348795758,
+      "loss": 0.4702,
+      "step": 207
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0008985059056855857,
+      "loss": 0.5088,
+      "step": 208
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0008975533276195101,
+      "loss": 0.4294,
+      "step": 209
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0008965968101206292,
+      "loss": 0.2792,
+      "step": 210
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0008956363626672594,
+      "loss": 0.3841,
+      "step": 211
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0008946719947766611,
+      "loss": 0.4744,
+      "step": 212
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0008937037160049415,
+      "loss": 0.3686,
+      "step": 213
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0008927315359469625,
+      "loss": 0.6375,
+      "step": 214
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0008917554642362443,
+      "loss": 0.8962,
+      "step": 215
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0008907755105448703,
+      "loss": 0.4171,
+      "step": 216
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.000889791684583391,
+      "loss": 0.4826,
+      "step": 217
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0008888039961007281,
+      "loss": 0.5129,
+      "step": 218
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.000887812454884078,
+      "loss": 0.2824,
+      "step": 219
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0008868170707588142,
+      "loss": 0.5414,
+      "step": 220
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0008858178535883905,
+      "loss": 0.4396,
+      "step": 221
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.000884814813274243,
+      "loss": 0.4208,
+      "step": 222
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0008838079597556925,
+      "loss": 0.3622,
+      "step": 223
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0008827973030098447,
+      "loss": 0.5095,
+      "step": 224
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0008817828530514931,
+      "loss": 0.3444,
+      "step": 225
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0008807646199330186,
+      "loss": 0.3598,
+      "step": 226
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0008797426137442896,
+      "loss": 0.2523,
+      "step": 227
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0008787168446125638,
+      "loss": 0.3462,
+      "step": 228
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0008776873227023851,
+      "loss": 0.4004,
+      "step": 229
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0008766540582154859,
+      "loss": 0.4201,
+      "step": 230
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0008756170613906833,
+      "loss": 0.4267,
+      "step": 231
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0008745763425037796,
+      "loss": 0.4423,
+      "step": 232
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0008735319118674597,
+      "loss": 0.5584,
+      "step": 233
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0008724837798311882,
+      "loss": 0.2896,
+      "step": 234
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0008714319567811089,
+      "loss": 0.4205,
+      "step": 235
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0008703764531399392,
+      "loss": 0.38,
+      "step": 236
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.000869317279366869,
+      "loss": 0.4093,
+      "step": 237
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0008682544459574561,
+      "loss": 0.4539,
+      "step": 238
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0008671879634435224,
+      "loss": 0.4635,
+      "step": 239
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0008661178423930491,
+      "loss": 0.3336,
+      "step": 240
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0008650440934100728,
+      "loss": 0.3342,
+      "step": 241
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0008639667271345798,
+      "loss": 0.4935,
+      "step": 242
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0008628857542424009,
+      "loss": 0.499,
+      "step": 243
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0008618011854451056,
+      "loss": 0.6852,
+      "step": 244
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0008607130314898956,
+      "loss": 0.5297,
+      "step": 245
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.000859621303159499,
+      "loss": 0.3527,
+      "step": 246
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0008585260112720631,
+      "loss": 0.4286,
+      "step": 247
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0008574271666810469,
+      "loss": 0.4928,
+      "step": 248
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0008563247802751139,
+      "loss": 0.5073,
+      "step": 249
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0008552188629780245,
+      "loss": 0.4953,
+      "step": 250
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0008541094257485265,
+      "loss": 0.4353,
+      "step": 251
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0008529964795802485,
+      "loss": 0.4549,
+      "step": 252
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0008518800355015891,
+      "loss": 0.5555,
+      "step": 253
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0008507601045756085,
+      "loss": 0.3896,
+      "step": 254
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0008496366978999189,
+      "loss": 0.3865,
+      "step": 255
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0008485098266065744,
+      "loss": 0.5022,
+      "step": 256
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0008473795018619604,
+      "loss": 0.4086,
+      "step": 257
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0008462457348666835,
+      "loss": 0.4797,
+      "step": 258
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0008451085368554601,
+      "loss": 0.4812,
+      "step": 259
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0008439679190970051,
+      "loss": 0.6112,
+      "step": 260
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0008428238928939207,
+      "loss": 0.3997,
+      "step": 261
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0008416764695825835,
+      "loss": 0.5654,
+      "step": 262
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0008405256605330332,
+      "loss": 0.3926,
+      "step": 263
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0008393714771488589,
+      "loss": 0.4882,
+      "step": 264
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0008382139308670875,
+      "loss": 0.5966,
+      "step": 265
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0008370530331580686,
+      "loss": 0.7363,
+      "step": 266
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.000835888795525362,
+      "loss": 0.4962,
+      "step": 267
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0008347212295056239,
+      "loss": 0.4113,
+      "step": 268
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0008335503466684915,
+      "loss": 0.6778,
+      "step": 269
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0008323761586164695,
+      "loss": 0.4551,
+      "step": 270
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0008311986769848141,
+      "loss": 0.3871,
+      "step": 271
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0008300179134414187,
+      "loss": 0.521,
+      "step": 272
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0008288338796866977,
+      "loss": 0.6278,
+      "step": 273
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0008276465874534701,
+      "loss": 0.3758,
+      "step": 274
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0008264560485068446,
+      "loss": 0.5227,
+      "step": 275
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0008252622746441021,
+      "loss": 0.3918,
+      "step": 276
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0008240652776945781,
+      "loss": 0.377,
+      "step": 277
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0008228650695195472,
+      "loss": 0.5587,
+      "step": 278
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0008216616620121042,
+      "loss": 0.4371,
+      "step": 279
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0008204550670970469,
+      "loss": 0.3701,
+      "step": 280
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0008192452967307575,
+      "loss": 0.3613,
+      "step": 281
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0008180323629010848,
+      "loss": 0.5874,
+      "step": 282
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0008168162776272244,
+      "loss": 0.4882,
+      "step": 283
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0008155970529596006,
+      "loss": 0.3851,
+      "step": 284
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0008143747009797464,
+      "loss": 0.49,
+      "step": 285
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.000813149233800184,
+      "loss": 0.584,
+      "step": 286
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0008119206635643044,
+      "loss": 0.4602,
+      "step": 287
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0008106890024462481,
+      "loss": 0.5223,
+      "step": 288
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0008094542626507828,
+      "loss": 0.4446,
+      "step": 289
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0008082164564131844,
+      "loss": 0.4818,
+      "step": 290
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.000806975595999114,
+      "loss": 0.4525,
+      "step": 291
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0008057316937044977,
+      "loss": 0.5049,
+      "step": 292
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0008044847618554035,
+      "loss": 0.4057,
+      "step": 293
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0008032348128079204,
+      "loss": 0.4718,
+      "step": 294
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0008019818589480352,
+      "loss": 0.43,
+      "step": 295
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0008007259126915101,
+      "loss": 0.4611,
+      "step": 296
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0007994669864837594,
+      "loss": 0.5587,
+      "step": 297
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0007982050927997263,
+      "loss": 0.7014,
+      "step": 298
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0007969402441437595,
+      "loss": 0.4874,
+      "step": 299
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0007956724530494887,
+      "loss": 0.6031,
+      "step": 300
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0007944017320797012,
+      "loss": 0.4922,
+      "step": 301
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0007931280938262169,
+      "loss": 0.5301,
+      "step": 302
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0007918515509097633,
+      "loss": 0.5914,
+      "step": 303
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0007905721159798514,
+      "loss": 0.5355,
+      "step": 304
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.000789289801714649,
+      "loss": 0.9378,
+      "step": 305
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.0007880046208208562,
+      "loss": 0.5175,
+      "step": 306
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0007867165860335792,
+      "loss": 0.4921,
+      "step": 307
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0007854257101162037,
+      "loss": 0.3571,
+      "step": 308
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0007841320058602688,
+      "loss": 0.4316,
+      "step": 309
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0007828354860853399,
+      "loss": 0.7104,
+      "step": 310
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0007815361636388827,
+      "loss": 0.5402,
+      "step": 311
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0007802340513961342,
+      "loss": 0.451,
+      "step": 312
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0007789291622599765,
+      "loss": 0.4272,
+      "step": 313
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0007776215091608085,
+      "loss": 0.6552,
+      "step": 314
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0007763111050564178,
+      "loss": 0.5025,
+      "step": 315
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0007749979629318516,
+      "loss": 0.4279,
+      "step": 316
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0007736820957992895,
+      "loss": 0.5845,
+      "step": 317
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0007723635166979133,
+      "loss": 0.4912,
+      "step": 318
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0007710422386937784,
+      "loss": 0.4654,
+      "step": 319
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0007697182748796841,
+      "loss": 0.3999,
+      "step": 320
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0007683916383750436,
+      "loss": 0.584,
+      "step": 321
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0007670623423257548,
+      "loss": 0.4966,
+      "step": 322
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0007657303999040692,
+      "loss": 0.3419,
+      "step": 323
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0007643958243084619,
+      "loss": 0.5106,
+      "step": 324
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0007630586287635007,
+      "loss": 0.5535,
+      "step": 325
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0007617188265197149,
+      "loss": 0.4968,
+      "step": 326
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.0007603764308534636,
+      "loss": 0.4221,
+      "step": 327
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.0007590314550668054,
+      "loss": 0.5194,
+      "step": 328
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0007576839124873654,
+      "loss": 0.4809,
+      "step": 329
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0007563338164682036,
+      "loss": 0.4809,
+      "step": 330
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0007549811803876825,
+      "loss": 0.4812,
+      "step": 331
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.0007536260176493348,
+      "loss": 0.4273,
+      "step": 332
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00075226834168173,
+      "loss": 0.4738,
+      "step": 333
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0007509081659383416,
+      "loss": 0.448,
+      "step": 334
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0007495455038974146,
+      "loss": 0.4226,
+      "step": 335
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0007481803690618304,
+      "loss": 0.4618,
+      "step": 336
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.000746812774958974,
+      "loss": 0.4488,
+      "step": 337
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.0007454427351405999,
+      "loss": 0.4695,
+      "step": 338
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.0007440702631826976,
+      "loss": 0.3598,
+      "step": 339
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0007426953726853573,
+      "loss": 0.4688,
+      "step": 340
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0007413180772726348,
+      "loss": 0.3983,
+      "step": 341
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.0007399383905924165,
+      "loss": 0.5252,
+      "step": 342
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.0007385563263162847,
+      "loss": 0.5176,
+      "step": 343
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.0007371718981393814,
+      "loss": 0.6202,
+      "step": 344
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.0007357851197802735,
+      "loss": 0.3895,
+      "step": 345
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0007343960049808155,
+      "loss": 0.4626,
+      "step": 346
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0007330045675060148,
+      "loss": 0.4189,
+      "step": 347
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0007316108211438946,
+      "loss": 0.3928,
+      "step": 348
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0007302147797053569,
+      "loss": 0.6127,
+      "step": 349
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0007288164570240462,
+      "loss": 0.4124,
+      "step": 350
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0007274158669562126,
+      "loss": 0.4845,
+      "step": 351
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.0007260130233805739,
+      "loss": 0.5578,
+      "step": 352
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.0007246079401981784,
+      "loss": 0.9514,
+      "step": 353
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0007232006313322668,
+      "loss": 0.4722,
+      "step": 354
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0007217911107281352,
+      "loss": 0.462,
+      "step": 355
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.0007203793923529956,
+      "loss": 0.5369,
+      "step": 356
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.0007189654901958386,
+      "loss": 0.5005,
+      "step": 357
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.0007175494182672939,
+      "loss": 0.5964,
+      "step": 358
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.0007161311905994922,
+      "loss": 0.4215,
+      "step": 359
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.0007147108212459257,
+      "loss": 0.5255,
+      "step": 360
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.000713288324281309,
+      "loss": 0.4609,
+      "step": 361
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.0007118637138014396,
+      "loss": 0.5004,
+      "step": 362
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.0007104370039230583,
+      "loss": 0.5112,
+      "step": 363
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.0007090082087837092,
+      "loss": 0.572,
+      "step": 364
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.0007075773425415994,
+      "loss": 0.3962,
+      "step": 365
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.0007061444193754596,
+      "loss": 0.5081,
+      "step": 366
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.0007047094534844022,
+      "loss": 0.4317,
+      "step": 367
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.0007032724590877821,
+      "loss": 0.434,
+      "step": 368
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.0007018334504250545,
+      "loss": 0.424,
+      "step": 369
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.0007003924417556343,
+      "loss": 0.5263,
+      "step": 370
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.0006989494473587554,
+      "loss": 0.3515,
+      "step": 371
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.0006975044815333281,
+      "loss": 0.5854,
+      "step": 372
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.0006960575585977984,
+      "loss": 0.5951,
+      "step": 373
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0006946086928900053,
+      "loss": 0.5526,
+      "step": 374
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0006931578987670395,
+      "loss": 0.4395,
+      "step": 375
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.0006917051906051005,
+      "loss": 0.4951,
+      "step": 376
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.0006902505827993541,
+      "loss": 0.5379,
+      "step": 377
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.0006887940897637908,
+      "loss": 0.4688,
+      "step": 378
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.0006873357259310815,
+      "loss": 0.422,
+      "step": 379
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.0006858755057524354,
+      "loss": 0.5442,
+      "step": 380
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.0006844134436974567,
+      "loss": 0.5796,
+      "step": 381
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0006829495542540013,
+      "loss": 0.3116,
+      "step": 382
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0006814838519280324,
+      "loss": 0.5884,
+      "step": 383
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.000680016351243478,
+      "loss": 0.4409,
+      "step": 384
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.0006785470667420861,
+      "loss": 0.5072,
+      "step": 385
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.000677076012983281,
+      "loss": 0.6952,
+      "step": 386
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.000675603204544019,
+      "loss": 0.5757,
+      "step": 387
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.0006741286560186436,
+      "loss": 0.4874,
+      "step": 388
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.0006726523820187413,
+      "loss": 0.6527,
+      "step": 389
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.0006711743971729967,
+      "loss": 0.3801,
+      "step": 390
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.0006696947161270476,
+      "loss": 0.5924,
+      "step": 391
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.0006682133535433393,
+      "loss": 0.4358,
+      "step": 392
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.0006667303241009804,
+      "loss": 0.3506,
+      "step": 393
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.0006652456424955964,
+      "loss": 0.4905,
+      "step": 394
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.0006637593234391843,
+      "loss": 0.207,
+      "step": 395
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.0006622713816599673,
+      "loss": 0.2183,
+      "step": 396
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.0006607818319022481,
+      "loss": 0.2478,
+      "step": 397
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.0006592906889262632,
+      "loss": 0.2421,
+      "step": 398
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0006577979675080369,
+      "loss": 0.2596,
+      "step": 399
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0006563036824392345,
+      "loss": 0.1455,
+      "step": 400
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.0006548078485270152,
+      "loss": 0.2416,
+      "step": 401
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.0006533104805938873,
+      "loss": 0.1597,
+      "step": 402
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.0006518115934775586,
+      "loss": 0.2274,
+      "step": 403
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.0006503112020307916,
+      "loss": 0.2043,
+      "step": 404
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.0006488093211212554,
+      "loss": 0.1718,
+      "step": 405
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.0006473059656313782,
+      "loss": 0.2003,
+      "step": 406
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.0006458011504582005,
+      "loss": 0.2682,
+      "step": 407
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.0006442948905132266,
+      "loss": 0.3159,
+      "step": 408
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0006427872007222777,
+      "loss": 0.1926,
+      "step": 409
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0006412780960253436,
+      "loss": 0.1705,
+      "step": 410
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.0006397675913764345,
+      "loss": 0.244,
+      "step": 411
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.0006382557017434331,
+      "loss": 0.263,
+      "step": 412
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.0006367424421079462,
+      "loss": 0.1458,
+      "step": 413
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.0006352278274651562,
+      "loss": 0.2241,
+      "step": 414
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.0006337118728236722,
+      "loss": 0.1758,
+      "step": 415
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.0006321945932053822,
+      "loss": 0.2477,
+      "step": 416
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.0006306760036453034,
+      "loss": 0.2506,
+      "step": 417
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.0006291561191914333,
+      "loss": 0.1463,
+      "step": 418
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.0006276349549046008,
+      "loss": 0.2307,
+      "step": 419
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.0006261125258583171,
+      "loss": 0.2191,
+      "step": 420
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.0006245888471386262,
+      "loss": 0.168,
+      "step": 421
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.0006230639338439548,
+      "loss": 0.2035,
+      "step": 422
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.000621537801084964,
+      "loss": 0.1725,
+      "step": 423
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.0006200104639843984,
+      "loss": 0.2434,
+      "step": 424
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.0006184819376769364,
+      "loss": 0.156,
+      "step": 425
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.0006169522373090413,
+      "loss": 0.2614,
+      "step": 426
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.0006154213780388092,
+      "loss": 0.1728,
+      "step": 427
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.0006138893750358212,
+      "loss": 0.4142,
+      "step": 428
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.0006123562434809912,
+      "loss": 0.2358,
+      "step": 429
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.0006108219985664161,
+      "loss": 0.1511,
+      "step": 430
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.0006092866554952256,
+      "loss": 0.2079,
+      "step": 431
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.0006077502294814311,
+      "loss": 0.2778,
+      "step": 432
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.000606212735749775,
+      "loss": 0.1745,
+      "step": 433
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.0006046741895355802,
+      "loss": 0.2087,
+      "step": 434
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.0006031346060845986,
+      "loss": 0.3661,
+      "step": 435
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.0006015940006528601,
+      "loss": 0.2497,
+      "step": 436
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.0006000523885065223,
+      "loss": 0.1748,
+      "step": 437
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.0005985097849217179,
+      "loss": 0.2426,
+      "step": 438
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.0005969662051844041,
+      "loss": 0.2053,
+      "step": 439
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.0005954216645902108,
+      "loss": 0.2444,
+      "step": 440
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00059387617844429,
+      "loss": 0.2316,
+      "step": 441
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.0005923297620611623,
+      "loss": 0.3341,
+      "step": 442
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0005907824307645668,
+      "loss": 0.2375,
+      "step": 443
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0005892341998873089,
+      "loss": 0.1812,
+      "step": 444
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.0005876850847711072,
+      "loss": 0.1778,
+      "step": 445
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.0005861351007664434,
+      "loss": 0.2015,
+      "step": 446
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.0005845842632324087,
+      "loss": 0.2272,
+      "step": 447
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.000583032587536552,
+      "loss": 0.2465,
+      "step": 448
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.0005814800890547278,
+      "loss": 0.2765,
+      "step": 449
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.0005799267831709441,
+      "loss": 0.2826,
+      "step": 450
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.000578372685277209,
+      "loss": 0.2576,
+      "step": 451
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.0005768178107733791,
+      "loss": 0.3645,
+      "step": 452
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.0005752621750670068,
+      "loss": 0.1209,
+      "step": 453
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.0005737057935731867,
+      "loss": 0.2813,
+      "step": 454
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.0005721486817144044,
+      "loss": 0.2002,
+      "step": 455
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.0005705908549203824,
+      "loss": 0.249,
+      "step": 456
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.0005690323286279274,
+      "loss": 0.2097,
+      "step": 457
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.0005674731182807781,
+      "loss": 0.2456,
+      "step": 458
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.0005659132393294514,
+      "loss": 0.2035,
+      "step": 459
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.0005643527072310891,
+      "loss": 0.1656,
+      "step": 460
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.0005627915374493061,
+      "loss": 0.334,
+      "step": 461
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.0005612297454540352,
+      "loss": 0.171,
+      "step": 462
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.0005596673467213756,
+      "loss": 0.1982,
+      "step": 463
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.0005581043567334383,
+      "loss": 0.1921,
+      "step": 464
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.0005565407909781934,
+      "loss": 0.2609,
+      "step": 465
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.0005549766649493165,
+      "loss": 0.2503,
+      "step": 466
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.000553411994146035,
+      "loss": 0.2269,
+      "step": 467
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0005518467940729739,
+      "loss": 0.1349,
+      "step": 468
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0005502810802400039,
+      "loss": 0.2037,
+      "step": 469
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.0005487148681620861,
+      "loss": 0.1903,
+      "step": 470
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.000547148173359119,
+      "loss": 0.187,
+      "step": 471
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.0005455810113557839,
+      "loss": 0.211,
+      "step": 472
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.0005440133976813926,
+      "loss": 0.2496,
+      "step": 473
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0005424453478697321,
+      "loss": 0.2057,
+      "step": 474
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.000540876877458911,
+      "loss": 0.1973,
+      "step": 475
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.0005393080019912061,
+      "loss": 0.232,
+      "step": 476
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.0005377387370129079,
+      "loss": 0.1924,
+      "step": 477
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.0005361690980741663,
+      "loss": 0.2347,
+      "step": 478
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.0005345991007288371,
+      "loss": 0.2134,
+      "step": 479
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.000533028760534328,
+      "loss": 0.2552,
+      "step": 480
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.0005314580930514431,
+      "loss": 0.3036,
+      "step": 481
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.0005298871138442307,
+      "loss": 0.1862,
+      "step": 482
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.0005283158384798275,
+      "loss": 0.3097,
+      "step": 483
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.0005267442825283048,
+      "loss": 0.2764,
+      "step": 484
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.0005251724615625145,
+      "loss": 0.2332,
+      "step": 485
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0005236003911579344,
+      "loss": 0.1589,
+      "step": 486
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0005220280868925145,
+      "loss": 0.1943,
+      "step": 487
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.0005204555643465215,
+      "loss": 0.1713,
+      "step": 488
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.0005188828391023856,
+      "loss": 0.2151,
+      "step": 489
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.0005173099267445451,
+      "loss": 0.1708,
+      "step": 490
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.0005157368428592932,
+      "loss": 0.1367,
+      "step": 491
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.000514163603034622,
+      "loss": 0.2488,
+      "step": 492
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.0005125902228600693,
+      "loss": 0.1392,
+      "step": 493
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0005110167179265636,
+      "loss": 0.2407,
+      "step": 494
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0005094431038262692,
+      "loss": 0.2323,
+      "step": 495
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.0005078693961524329,
+      "loss": 0.236,
+      "step": 496
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.0005062956104992285,
+      "loss": 0.1665,
+      "step": 497
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.0005047217624616019,
+      "loss": 0.2289,
+      "step": 498
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.0005031478676351178,
+      "loss": 0.6418,
+      "step": 499
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.0005015739416158049,
+      "loss": 0.1473,
+      "step": 500
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.0005,
+      "loss": 0.2873,
+      "step": 501
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.0004984260583841952,
+      "loss": 0.1745,
+      "step": 502
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.0004968521323648822,
+      "loss": 0.2458,
+      "step": 503
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.0004952782375383984,
+      "loss": 0.2808,
+      "step": 504
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.0004937043895007717,
+      "loss": 0.1909,
+      "step": 505
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.0004921306038475671,
+      "loss": 0.2185,
+      "step": 506
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.0004905568961737309,
+      "loss": 0.236,
+      "step": 507
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0004889832820734367,
+      "loss": 0.2238,
+      "step": 508
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0004874097771399308,
+      "loss": 0.207,
+      "step": 509
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.0004858363969653781,
+      "loss": 0.2601,
+      "step": 510
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.00048426315714070683,
+      "loss": 0.2049,
+      "step": 511
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.00048269007325545506,
+      "loss": 0.2763,
+      "step": 512
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.00048111716089761455,
+      "loss": 0.2787,
+      "step": 513
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.00047954443565347867,
+      "loss": 0.2045,
+      "step": 514
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.00047797191310748567,
+      "loss": 0.2062,
+      "step": 515
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.00047639960884206576,
+      "loss": 0.1595,
+      "step": 516
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.0004748275384374856,
+      "loss": 0.1786,
+      "step": 517
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.00047325571747169545,
+      "loss": 0.2422,
+      "step": 518
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.00047168416152017255,
+      "loss": 0.1506,
+      "step": 519
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.0004701128861557694,
+      "loss": 0.1656,
+      "step": 520
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.0004685419069485569,
+      "loss": 0.26,
+      "step": 521
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.0004669712394656723,
+      "loss": 0.3045,
+      "step": 522
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.00046540089927116305,
+      "loss": 0.2051,
+      "step": 523
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00046383090192583397,
+      "loss": 0.198,
+      "step": 524
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.0004622612629870923,
+      "loss": 0.2119,
+      "step": 525
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.00046069199800879404,
+      "loss": 0.2543,
+      "step": 526
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00045912312254108907,
+      "loss": 0.2439,
+      "step": 527
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00045755465213026807,
+      "loss": 0.1628,
+      "step": 528
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.00045598660231860745,
+      "loss": 0.2464,
+      "step": 529
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.0004544189886442162,
+      "loss": 0.1534,
+      "step": 530
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.0004528518266408811,
+      "loss": 0.2754,
+      "step": 531
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.00045128513183791385,
+      "loss": 0.2381,
+      "step": 532
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.000449718919759996,
+      "loss": 0.3008,
+      "step": 533
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.0004481532059270262,
+      "loss": 0.3641,
+      "step": 534
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.0004465880058539652,
+      "loss": 0.3051,
+      "step": 535
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00044502333505068355,
+      "loss": 0.2982,
+      "step": 536
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.0004434592090218065,
+      "loss": 0.1558,
+      "step": 537
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.0004418956432665618,
+      "loss": 0.2677,
+      "step": 538
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.0004403326532786245,
+      "loss": 0.2461,
+      "step": 539
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.00043877025454596493,
+      "loss": 0.2472,
+      "step": 540
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.000437208462550694,
+      "loss": 0.2212,
+      "step": 541
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.0004356472927689109,
+      "loss": 0.198,
+      "step": 542
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.00043408676067054866,
+      "loss": 0.2927,
+      "step": 543
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.000432526881719222,
+      "loss": 0.2636,
+      "step": 544
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.00043096767137207256,
+      "loss": 0.359,
+      "step": 545
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.0004294091450796177,
+      "loss": 0.2385,
+      "step": 546
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.0004278513182855956,
+      "loss": 0.136,
+      "step": 547
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.0004262942064268134,
+      "loss": 0.1927,
+      "step": 548
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.00042473782493299324,
+      "loss": 0.2469,
+      "step": 549
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.000423182189226621,
+      "loss": 0.252,
+      "step": 550
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.00042162731472279095,
+      "loss": 0.2133,
+      "step": 551
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.000420073216829056,
+      "loss": 0.2453,
+      "step": 552
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.0004185199109452721,
+      "loss": 0.274,
+      "step": 553
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.00041696741246344813,
+      "loss": 0.2342,
+      "step": 554
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.00041541573676759125,
+      "loss": 0.17,
+      "step": 555
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.00041386489923355666,
+      "loss": 0.234,
+      "step": 556
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.000412314915228893,
+      "loss": 0.1674,
+      "step": 557
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.00041076580011269127,
+      "loss": 0.2432,
+      "step": 558
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.0004092175692354333,
+      "loss": 0.2297,
+      "step": 559
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.00040767023793883785,
+      "loss": 0.2347,
+      "step": 560
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.0004061238215557103,
+      "loss": 0.2065,
+      "step": 561
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.0004045783354097893,
+      "loss": 0.1912,
+      "step": 562
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.0004030337948155962,
+      "loss": 0.2439,
+      "step": 563
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.00040149021507828223,
+      "loss": 0.1611,
+      "step": 564
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.0003999476114934778,
+      "loss": 0.2625,
+      "step": 565
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.0003984059993471399,
+      "loss": 0.1553,
+      "step": 566
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.0003968653939154016,
+      "loss": 0.2644,
+      "step": 567
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00039532581046441994,
+      "loss": 0.1631,
+      "step": 568
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.00039378726425022515,
+      "loss": 0.2319,
+      "step": 569
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.00039224977051856904,
+      "loss": 0.1823,
+      "step": 570
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.0003907133445047747,
+      "loss": 0.1732,
+      "step": 571
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.000389178001433584,
+      "loss": 0.1973,
+      "step": 572
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.000387643756519009,
+      "loss": 0.2714,
+      "step": 573
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00038611062496417893,
+      "loss": 0.2124,
+      "step": 574
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.000384578621961191,
+      "loss": 0.3089,
+      "step": 575
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.00038304776269095886,
+      "loss": 0.2345,
+      "step": 576
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.0003815180623230637,
+      "loss": 0.1838,
+      "step": 577
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.0003799895360156017,
+      "loss": 0.1751,
+      "step": 578
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00037846219891503606,
+      "loss": 0.1985,
+      "step": 579
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00037693606615604524,
+      "loss": 0.2054,
+      "step": 580
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.00037541115286137403,
+      "loss": 0.3287,
+      "step": 581
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.00037388747414168294,
+      "loss": 0.1547,
+      "step": 582
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.00037236504509539936,
+      "loss": 0.2142,
+      "step": 583
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.0003708438808085668,
+      "loss": 0.2749,
+      "step": 584
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.00036932399635469673,
+      "loss": 0.251,
+      "step": 585
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.0003678054067946178,
+      "loss": 0.3888,
+      "step": 586
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00036628812717632794,
+      "loss": 0.1405,
+      "step": 587
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00036477217253484397,
+      "loss": 0.2749,
+      "step": 588
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.000363257557892054,
+      "loss": 0.1246,
+      "step": 589
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.00036174429825656685,
+      "loss": 0.2626,
+      "step": 590
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0003602324086235655,
+      "loss": 0.1572,
+      "step": 591
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.00035872190397465636,
+      "loss": 0.0793,
+      "step": 592
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.00035721279927772233,
+      "loss": 0.0366,
+      "step": 593
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.0003557051094867735,
+      "loss": 0.0661,
+      "step": 594
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.0003541988495417997,
+      "loss": 0.0447,
+      "step": 595
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00035269403436862174,
+      "loss": 0.0356,
+      "step": 596
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.0003511906788787447,
+      "loss": 0.0507,
+      "step": 597
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.00034968879796920836,
+      "loss": 0.0461,
+      "step": 598
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.0003481884065224415,
+      "loss": 0.0855,
+      "step": 599
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00034668951940611275,
+      "loss": 0.0666,
+      "step": 600
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.0003451921514729848,
+      "loss": 0.0708,
+      "step": 601
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.0003436963175607656,
+      "loss": 0.0662,
+      "step": 602
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00034220203249196317,
+      "loss": 0.0593,
+      "step": 603
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.0003407093110737368,
+      "loss": 0.0654,
+      "step": 604
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00033921816809775204,
+      "loss": 0.0445,
+      "step": 605
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.00033772861834003274,
+      "loss": 0.0355,
+      "step": 606
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.00033624067656081575,
+      "loss": 0.0426,
+      "step": 607
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.0003347543575044035,
+      "loss": 0.0438,
+      "step": 608
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00033326967589901967,
+      "loss": 0.0836,
+      "step": 609
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.0003317866464566607,
+      "loss": 0.0479,
+      "step": 610
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.0003303052838729525,
+      "loss": 0.1101,
+      "step": 611
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00032882560282700336,
+      "loss": 0.0661,
+      "step": 612
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.0003273476179812588,
+      "loss": 0.0452,
+      "step": 613
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.0003258713439813565,
+      "loss": 0.0338,
+      "step": 614
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.0003243967954559811,
+      "loss": 0.0277,
+      "step": 615
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.0003229239870167191,
+      "loss": 0.0559,
+      "step": 616
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.00032145293325791394,
+      "loss": 0.0732,
+      "step": 617
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.0003199836487565222,
+      "loss": 0.123,
+      "step": 618
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.0003185161480719677,
+      "loss": 0.0501,
+      "step": 619
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.0003170504457459989,
+      "loss": 0.0836,
+      "step": 620
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.0003155865563025433,
+      "loss": 0.0464,
+      "step": 621
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.00031412449424756474,
+      "loss": 0.0927,
+      "step": 622
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.0003126642740689186,
+      "loss": 0.0291,
+      "step": 623
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00031120591023620927,
+      "loss": 0.0793,
+      "step": 624
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.0003097494172006459,
+      "loss": 0.0812,
+      "step": 625
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.00030829480939489965,
+      "loss": 0.0852,
+      "step": 626
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.00030684210123296053,
+      "loss": 0.1087,
+      "step": 627
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00030539130710999473,
+      "loss": 0.074,
+      "step": 628
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.0003039424414022016,
+      "loss": 0.0655,
+      "step": 629
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.00030249551846667206,
+      "loss": 0.0622,
+      "step": 630
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.0003010505526412447,
+      "loss": 0.0373,
+      "step": 631
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.0002996075582443658,
+      "loss": 0.054,
+      "step": 632
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.0002981665495749457,
+      "loss": 0.0741,
+      "step": 633
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.00029672754091221806,
+      "loss": 0.0931,
+      "step": 634
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.0002952905465155977,
+      "loss": 0.0934,
+      "step": 635
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.0002938555806245406,
+      "loss": 0.089,
+      "step": 636
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00029242265745840063,
+      "loss": 0.0873,
+      "step": 637
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.00029099179121629115,
+      "loss": 0.0719,
+      "step": 638
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.0002895629960769417,
+      "loss": 0.1076,
+      "step": 639
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.0002881362861985606,
+      "loss": 0.0738,
+      "step": 640
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.0002867116757186911,
+      "loss": 0.0744,
+      "step": 641
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.00028528917875407435,
+      "loss": 0.0468,
+      "step": 642
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.0002838688094005078,
+      "loss": 0.0523,
+      "step": 643
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.00028245058173270623,
+      "loss": 0.0498,
+      "step": 644
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.00028103450980416134,
+      "loss": 0.0621,
+      "step": 645
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.0002796206076470044,
+      "loss": 0.0792,
+      "step": 646
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.0002782088892718648,
+      "loss": 0.0536,
+      "step": 647
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00027679936866773316,
+      "loss": 0.0689,
+      "step": 648
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.0002753920598018217,
+      "loss": 0.0561,
+      "step": 649
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00027398697661942627,
+      "loss": 0.0681,
+      "step": 650
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00027258413304378734,
+      "loss": 0.0969,
+      "step": 651
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00027118354297595395,
+      "loss": 0.0473,
+      "step": 652
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00026978522029464324,
+      "loss": 0.0433,
+      "step": 653
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.0002683891788561055,
+      "loss": 0.0564,
+      "step": 654
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.0002669954324939852,
+      "loss": 0.1075,
+      "step": 655
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.00026560399501918467,
+      "loss": 0.0576,
+      "step": 656
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.0002642148802197267,
+      "loss": 0.0872,
+      "step": 657
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.0002628281018606186,
+      "loss": 0.1002,
+      "step": 658
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 0.00026144367368371535,
+      "loss": 0.0549,
+      "step": 659
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 0.0002600616094075835,
+      "loss": 0.0571,
+      "step": 660
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 0.00025868192272736514,
+      "loss": 0.0875,
+      "step": 661
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 0.0002573046273146427,
+      "loss": 0.0623,
+      "step": 662
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 0.0002559297368173024,
+      "loss": 0.1347,
+      "step": 663
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 0.00025455726485940013,
+      "loss": 0.0528,
+      "step": 664
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 0.00025318722504102606,
+      "loss": 0.0508,
+      "step": 665
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 0.0002518196309381696,
+      "loss": 0.0697,
+      "step": 666
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 0.0002504544961025853,
+      "loss": 0.0655,
+      "step": 667
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 0.00024909183406165834,
+      "loss": 0.0566,
+      "step": 668
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.0002477316583182702,
+      "loss": 0.0714,
+      "step": 669
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.00024637398235066526,
+      "loss": 0.0747,
+      "step": 670
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 0.0002450188196123177,
+      "loss": 0.0617,
+      "step": 671
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 0.00024366618353179642,
+      "loss": 0.0488,
+      "step": 672
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 0.0002423160875126348,
+      "loss": 0.0622,
+      "step": 673
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 0.00024096854493319475,
+      "loss": 0.0697,
+      "step": 674
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 0.00023962356914653655,
+      "loss": 0.0897,
+      "step": 675
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 0.00023828117348028528,
+      "loss": 0.078,
+      "step": 676
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 0.00023694137123649945,
+      "loss": 0.0958,
+      "step": 677
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 0.00023560417569153797,
+      "loss": 0.1041,
+      "step": 678
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.0002342696000959309,
+      "loss": 0.0466,
+      "step": 679
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.00023293765767424534,
+      "loss": 0.0474,
+      "step": 680
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 0.00023160836162495652,
+      "loss": 0.0809,
+      "step": 681
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 0.00023028172512031605,
+      "loss": 0.0506,
+      "step": 682
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.00022895776130622182,
+      "loss": 0.0766,
+      "step": 683
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.0002276364833020868,
+      "loss": 0.0307,
+      "step": 684
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 0.00022631790420071064,
+      "loss": 0.0586,
+      "step": 685
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 0.00022500203706814854,
+      "loss": 0.0805,
+      "step": 686
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 0.00022368889494358237,
+      "loss": 0.0766,
+      "step": 687
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 0.00022237849083919144,
+      "loss": 0.0723,
+      "step": 688
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.00022107083774002361,
+      "loss": 0.0643,
+      "step": 689
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.00021976594860386594,
+      "loss": 0.0543,
+      "step": 690
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 0.00021846383636111745,
+      "loss": 0.0654,
+      "step": 691
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 0.0002171645139146601,
+      "loss": 0.066,
+      "step": 692
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 0.00021586799413973135,
+      "loss": 0.0478,
+      "step": 693
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 0.00021457428988379634,
+      "loss": 0.1587,
+      "step": 694
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 0.00021328341396642091,
+      "loss": 0.0895,
+      "step": 695
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 0.00021199537917914385,
+      "loss": 0.0559,
+      "step": 696
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.0002107101982853511,
+      "loss": 0.0638,
+      "step": 697
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.00020942788402014868,
+      "loss": 0.0335,
+      "step": 698
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 0.0002081484490902366,
+      "loss": 0.2609,
+      "step": 699
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 0.0002068719061737831,
+      "loss": 0.0565,
+      "step": 700
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 0.00020559826792029885,
+      "loss": 0.0791,
+      "step": 701
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 0.00020432754695051135,
+      "loss": 0.0252,
+      "step": 702
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.0002030597558562406,
+      "loss": 0.0817,
+      "step": 703
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.0002017949072002737,
+      "loss": 0.014,
+      "step": 704
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 0.0002005330135162408,
+      "loss": 0.0377,
+      "step": 705
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 0.00019927408730848988,
+      "loss": 0.0633,
+      "step": 706
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 0.00019801814105196498,
+      "loss": 0.0667,
+      "step": 707
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 0.00019676518719207977,
+      "loss": 0.0506,
+      "step": 708
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 0.00019551523814459665,
+      "loss": 0.0661,
+      "step": 709
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 0.0001942683062955024,
+      "loss": 0.0808,
+      "step": 710
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 0.00019302440400088605,
+      "loss": 0.0402,
+      "step": 711
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 0.00019178354358681548,
+      "loss": 0.0523,
+      "step": 712
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 0.00019054573734921716,
+      "loss": 0.0338,
+      "step": 713
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 0.00018931099755375202,
+      "loss": 0.0693,
+      "step": 714
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 0.00018807933643569559,
+      "loss": 0.0384,
+      "step": 715
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 0.00018685076619981607,
+      "loss": 0.0587,
+      "step": 716
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 0.0001856252990202537,
+      "loss": 0.0533,
+      "step": 717
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 0.0001844029470403993,
+      "loss": 0.0493,
+      "step": 718
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 0.00018318372237277565,
+      "loss": 0.0897,
+      "step": 719
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 0.00018196763709891524,
+      "loss": 0.0726,
+      "step": 720
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 0.0001807547032692424,
+      "loss": 0.0668,
+      "step": 721
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 0.0001795449329029531,
+      "loss": 0.0753,
+      "step": 722
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 0.00017833833798789594,
+      "loss": 0.071,
+      "step": 723
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.00017713493048045293,
+      "loss": 0.0453,
+      "step": 724
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.000175934722305422,
+      "loss": 0.0406,
+      "step": 725
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 0.0001747377253558982,
+      "loss": 0.0635,
+      "step": 726
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 0.00017354395149315533,
+      "loss": 0.0737,
+      "step": 727
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 0.00017235341254653004,
+      "loss": 0.0618,
+      "step": 728
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 0.00017116612031330253,
+      "loss": 0.0542,
+      "step": 729
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 0.00016998208655858137,
+      "loss": 0.081,
+      "step": 730
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 0.00016880132301518597,
+      "loss": 0.0573,
+      "step": 731
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 0.00016762384138353076,
+      "loss": 0.0335,
+      "step": 732
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 0.00016644965333150846,
+      "loss": 0.0498,
+      "step": 733
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 0.00016527877049437624,
+      "loss": 0.0684,
+      "step": 734
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 0.00016411120447463806,
+      "loss": 0.0448,
+      "step": 735
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 0.00016294696684193154,
+      "loss": 0.0779,
+      "step": 736
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 0.0001617860691329126,
+      "loss": 0.0652,
+      "step": 737
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.00016062852285114122,
+      "loss": 0.0356,
+      "step": 738
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.00015947433946696693,
+      "loss": 0.1102,
+      "step": 739
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 0.0001583235304174167,
+      "loss": 0.0407,
+      "step": 740
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 0.00015717610710607949,
+      "loss": 0.0249,
+      "step": 741
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 0.00015603208090299498,
+      "loss": 0.053,
+      "step": 742
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 0.00015489146314454,
+      "loss": 0.0357,
+      "step": 743
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 0.0001537542651333167,
+      "loss": 0.0729,
+      "step": 744
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 0.00015262049813803956,
+      "loss": 0.0398,
+      "step": 745
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 0.00015149017339342576,
+      "loss": 0.0674,
+      "step": 746
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 0.00015036330210008115,
+      "loss": 0.0381,
+      "step": 747
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 0.0001492398954243916,
+      "loss": 0.0765,
+      "step": 748
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 0.00014811996449841098,
+      "loss": 0.0245,
+      "step": 749
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 0.0001470035204197517,
+      "loss": 0.037,
+      "step": 750
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 0.0001458905742514734,
+      "loss": 0.0521,
+      "step": 751
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 0.0001447811370219757,
+      "loss": 0.0429,
+      "step": 752
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 0.0001436752197248861,
+      "loss": 0.0589,
+      "step": 753
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.00014257283331895317,
+      "loss": 0.0922,
+      "step": 754
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.00014147398872793693,
+      "loss": 0.0742,
+      "step": 755
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 0.00014037869684050113,
+      "loss": 0.0461,
+      "step": 756
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 0.0001392869685101044,
+      "loss": 0.0312,
+      "step": 757
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 0.00013819881455489458,
+      "loss": 0.0173,
+      "step": 758
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 0.00013711424575759911,
+      "loss": 0.0386,
+      "step": 759
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 0.00013603327286542023,
+      "loss": 0.0292,
+      "step": 760
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 0.00013495590658992718,
+      "loss": 0.0973,
+      "step": 761
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 0.000133882157606951,
+      "loss": 0.0646,
+      "step": 762
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 0.00013281203655647755,
+      "loss": 0.1611,
+      "step": 763
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 0.0001317455540425439,
+      "loss": 0.0415,
+      "step": 764
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 0.00013068272063313102,
+      "loss": 0.0521,
+      "step": 765
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 0.00012962354686006083,
+      "loss": 0.035,
+      "step": 766
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 0.00012856804321889116,
+      "loss": 0.0574,
+      "step": 767
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 0.0001275162201688118,
+      "loss": 0.0436,
+      "step": 768
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 0.00012646808813254035,
+      "loss": 0.0405,
+      "step": 769
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 0.00012542365749622047,
+      "loss": 0.0526,
+      "step": 770
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 0.00012438293860931676,
+      "loss": 0.0275,
+      "step": 771
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 0.00012334594178451424,
+      "loss": 0.0757,
+      "step": 772
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 0.00012231267729761485,
+      "loss": 0.0317,
+      "step": 773
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 0.00012128315538743645,
+      "loss": 0.0306,
+      "step": 774
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 0.00012025738625571026,
+      "loss": 0.0453,
+      "step": 775
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 0.00011923538006698154,
+      "loss": 0.044,
+      "step": 776
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 0.00011821714694850688,
+      "loss": 0.0719,
+      "step": 777
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.00011720269699015528,
+      "loss": 0.056,
+      "step": 778
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.00011619204024430768,
+      "loss": 0.0287,
+      "step": 779
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 0.00011518518672575701,
+      "loss": 0.0672,
+      "step": 780
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 0.00011418214641160956,
+      "loss": 0.0477,
+      "step": 781
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.00011318292924118584,
+      "loss": 0.0515,
+      "step": 782
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.0001121875451159221,
+      "loss": 0.0937,
+      "step": 783
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 0.00011119600389927181,
+      "loss": 0.0534,
+      "step": 784
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 0.00011020831541660914,
+      "loss": 0.0238,
+      "step": 785
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.00010922448945512981,
+      "loss": 0.0686,
+      "step": 786
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.00010824453576375576,
+      "loss": 0.0209,
+      "step": 787
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 0.00010726846405303752,
+      "loss": 0.066,
+      "step": 788
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 0.0001062962839950587,
+      "loss": 0.0097,
+      "step": 789
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 0.00010532800522333902,
+      "loss": 0.0172,
+      "step": 790
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 0.00010436363733274057,
+      "loss": 0.0185,
+      "step": 791
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 0.00010340318987937097,
+      "loss": 0.0164,
+      "step": 792
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 0.00010244667238048988,
+      "loss": 0.0069,
+      "step": 793
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 0.0001014940943144142,
+      "loss": 0.018,
+      "step": 794
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 0.00010054546512042422,
+      "loss": 0.0065,
+      "step": 795
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.960079419866985e-05,
+      "loss": 0.0177,
+      "step": 796
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 9.866009091007833e-05,
+      "loss": 0.0105,
+      "step": 797
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 9.772336457626013e-05,
+      "loss": 0.0127,
+      "step": 798
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 9.679062447941778e-05,
+      "loss": 0.0111,
+      "step": 799
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 9.586187986225326e-05,
+      "loss": 0.0114,
+      "step": 800
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 9.493713992787673e-05,
+      "loss": 0.006,
+      "step": 801
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 9.401641383971476e-05,
+      "loss": 0.0041,
+      "step": 802
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 9.309971072142037e-05,
+      "loss": 0.0126,
+      "step": 803
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 9.218703965678204e-05,
+      "loss": 0.0107,
+      "step": 804
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 9.12784096896338e-05,
+      "loss": 0.0045,
+      "step": 805
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 9.03738298237658e-05,
+      "loss": 0.0039,
+      "step": 806
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.94733090228349e-05,
+      "loss": 0.0067,
+      "step": 807
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.857685621027567e-05,
+      "loss": 0.0114,
+      "step": 808
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.768448026921244e-05,
+      "loss": 0.0121,
+      "step": 809
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.679619004237111e-05,
+      "loss": 0.0055,
+      "step": 810
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.591199433199127e-05,
+      "loss": 0.0422,
+      "step": 811
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.503190189973914e-05,
+      "loss": 0.0126,
+      "step": 812
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.415592146662105e-05,
+      "loss": 0.0071,
+      "step": 813
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.328406171289621e-05,
+      "loss": 0.0035,
+      "step": 814
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.24163312779917e-05,
+      "loss": 0.0216,
+      "step": 815
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.155273876041613e-05,
+      "loss": 0.0069,
+      "step": 816
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.069329271767484e-05,
+      "loss": 0.0219,
+      "step": 817
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 7.983800166618483e-05,
+      "loss": 0.0105,
+      "step": 818
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 7.898687408119065e-05,
+      "loss": 0.0114,
+      "step": 819
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 7.813991839667995e-05,
+      "loss": 0.0269,
+      "step": 820
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 7.72971430053005e-05,
+      "loss": 0.0235,
+      "step": 821
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 7.645855625827657e-05,
+      "loss": 0.0289,
+      "step": 822
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 7.562416646532661e-05,
+      "loss": 0.0042,
+      "step": 823
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 7.479398189458003e-05,
+      "loss": 0.0193,
+      "step": 824
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 7.396801077249677e-05,
+      "loss": 0.024,
+      "step": 825
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 7.31462612837841e-05,
+      "loss": 0.006,
+      "step": 826
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 7.232874157131669e-05,
+      "loss": 0.0372,
+      "step": 827
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 7.15154597360555e-05,
+      "loss": 0.0196,
+      "step": 828
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 7.070642383696762e-05,
+      "loss": 0.0154,
+      "step": 829
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 6.99016418909459e-05,
+      "loss": 0.0152,
+      "step": 830
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 6.910112187273066e-05,
+      "loss": 0.0112,
+      "step": 831
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 6.830487171482935e-05,
+      "loss": 0.0072,
+      "step": 832
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 6.75128993074388e-05,
+      "loss": 0.0038,
+      "step": 833
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 6.672521249836688e-05,
+      "loss": 0.0065,
+      "step": 834
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 6.594181909295427e-05,
+      "loss": 0.0224,
+      "step": 835
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 6.516272685399794e-05,
+      "loss": 0.0035,
+      "step": 836
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 6.438794350167337e-05,
+      "loss": 0.0209,
+      "step": 837
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 6.36174767134588e-05,
+      "loss": 0.0052,
+      "step": 838
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 6.285133412405858e-05,
+      "loss": 0.013,
+      "step": 839
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 6.208952332532785e-05,
+      "loss": 0.0097,
+      "step": 840
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 6.133205186619695e-05,
+      "loss": 0.0178,
+      "step": 841
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 6.057892725259717e-05,
+      "loss": 0.0089,
+      "step": 842
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 5.983015694738597e-05,
+      "loss": 0.0112,
+      "step": 843
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 5.908574837027309e-05,
+      "loss": 0.0057,
+      "step": 844
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 5.8345708897747095e-05,
+      "loss": 0.003,
+      "step": 845
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 5.761004586300234e-05,
+      "loss": 0.0048,
+      "step": 846
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 5.687876655586582e-05,
+      "loss": 0.0136,
+      "step": 847
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 5.615187822272583e-05,
+      "loss": 0.0277,
+      "step": 848
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 5.5429388066459305e-05,
+      "loss": 0.0309,
+      "step": 849
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 5.4711303246361146e-05,
+      "loss": 0.0192,
+      "step": 850
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 5.3997630878072356e-05,
+      "loss": 0.0061,
+      "step": 851
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 5.328837803351083e-05,
+      "loss": 0.0031,
+      "step": 852
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 5.258355174079993e-05,
+      "loss": 0.0031,
+      "step": 853
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 5.1883158984199704e-05,
+      "loss": 0.0063,
+      "step": 854
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 5.118720670403748e-05,
+      "loss": 0.0113,
+      "step": 855
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 5.0495701796639096e-05,
+      "loss": 0.0044,
+      "step": 856
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 4.980865111426003e-05,
+      "loss": 0.0022,
+      "step": 857
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 4.9126061465018856e-05,
+      "loss": 0.0045,
+      "step": 858
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 4.844793961282812e-05,
+      "loss": 0.0099,
+      "step": 859
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 4.777429227732843e-05,
+      "loss": 0.0075,
+      "step": 860
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 4.710512613382151e-05,
+      "loss": 0.014,
+      "step": 861
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 4.6440447813204225e-05,
+      "loss": 0.005,
+      "step": 862
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 4.578026390190232e-05,
+      "loss": 0.0102,
+      "step": 863
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 4.512458094180616e-05,
+      "loss": 0.0027,
+      "step": 864
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 4.447340543020473e-05,
+      "loss": 0.0066,
+      "step": 865
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 4.3826743819722236e-05,
+      "loss": 0.0089,
+      "step": 866
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 4.318460251825357e-05,
+      "loss": 0.008,
+      "step": 867
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 4.2546987888901266e-05,
+      "loss": 0.0097,
+      "step": 868
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 4.191390624991159e-05,
+      "loss": 0.0037,
+      "step": 869
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 4.12853638746134e-05,
+      "loss": 0.0416,
+      "step": 870
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 4.066136699135436e-05,
+      "loss": 0.0102,
+      "step": 871
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 4.004192178344029e-05,
+      "loss": 0.0159,
+      "step": 872
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.942703438907358e-05,
+      "loss": 0.0045,
+      "step": 873
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.881671090129246e-05,
+      "loss": 0.0112,
+      "step": 874
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.821095736791008e-05,
+      "loss": 0.0061,
+      "step": 875
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.7609779791455744e-05,
+      "loss": 0.0139,
+      "step": 876
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.7013184129113974e-05,
+      "loss": 0.0227,
+      "step": 877
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.642117629266678e-05,
+      "loss": 0.0184,
+      "step": 878
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.58337621484342e-05,
+      "loss": 0.0063,
+      "step": 879
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.5250947517216634e-05,
+      "loss": 0.0095,
+      "step": 880
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.467273817423688e-05,
+      "loss": 0.0451,
+      "step": 881
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.40991398490833e-05,
+      "loss": 0.0086,
+      "step": 882
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.353015822565253e-05,
+      "loss": 0.0197,
+      "step": 883
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.296579894209345e-05,
+      "loss": 0.0232,
+      "step": 884
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.240606759075143e-05,
+      "loss": 0.004,
+      "step": 885
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.185096971811274e-05,
+      "loss": 0.0071,
+      "step": 886
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.1300510824749276e-05,
+      "loss": 0.0077,
+      "step": 887
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 3.0754696365265064e-05,
+      "loss": 0.0156,
+      "step": 888
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 3.0213531748240763e-05,
+      "loss": 0.0067,
+      "step": 889
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 2.9677022336181414e-05,
+      "loss": 0.0304,
+      "step": 890
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 2.9145173445462582e-05,
+      "loss": 0.006,
+      "step": 891
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 2.8617990346277655e-05,
+      "loss": 0.0188,
+      "step": 892
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 2.8095478262585904e-05,
+      "loss": 0.0145,
+      "step": 893
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.7577642372060676e-05,
+      "loss": 0.0076,
+      "step": 894
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.7064487806037985e-05,
+      "loss": 0.0043,
+      "step": 895
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.6556019649465524e-05,
+      "loss": 0.0237,
+      "step": 896
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.605224294085279e-05,
+      "loss": 0.0117,
+      "step": 897
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.5553162672220465e-05,
+      "loss": 0.0166,
+      "step": 898
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.5058783789051463e-05,
+      "loss": 0.0019,
+      "step": 899
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.45691111902418e-05,
+      "loss": 0.0276,
+      "step": 900
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.4084149728051953e-05,
+      "loss": 0.0064,
+      "step": 901
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.3603904208058692e-05,
+      "loss": 0.0092,
+      "step": 902
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.3128379389108e-05,
+      "loss": 0.0086,
+      "step": 903
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.2657579983267064e-05,
+      "loss": 0.0065,
+      "step": 904
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.219151065577829e-05,
+      "loss": 0.0017,
+      "step": 905
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.1730176025012817e-05,
+      "loss": 0.0046,
+      "step": 906
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.1273580662424797e-05,
+      "loss": 0.0067,
+      "step": 907
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.082172909250568e-05,
+      "loss": 0.0064,
+      "step": 908
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.0374625792740464e-05,
+      "loss": 0.0069,
+      "step": 909
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.993227519356189e-05,
+      "loss": 0.0041,
+      "step": 910
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.94946816783077e-05,
+      "loss": 0.0047,
+      "step": 911
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.906184958317664e-05,
+      "loss": 0.0057,
+      "step": 912
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.863378319718578e-05,
+      "loss": 0.0069,
+      "step": 913
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.82104867621275e-05,
+      "loss": 0.0217,
+      "step": 914
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.7791964472528233e-05,
+      "loss": 0.0219,
+      "step": 915
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.737822047560611e-05,
+      "loss": 0.0068,
+      "step": 916
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.696925887123052e-05,
+      "loss": 0.0136,
+      "step": 917
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1.656508371188109e-05,
+      "loss": 0.0102,
+      "step": 918
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1.616569900260767e-05,
+      "loss": 0.0051,
+      "step": 919
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.577110870099041e-05,
+      "loss": 0.0154,
+      "step": 920
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.538131671710108e-05,
+      "loss": 0.0713,
+      "step": 921
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.4996326913463754e-05,
+      "loss": 0.0157,
+      "step": 922
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1.461614310501691e-05,
+      "loss": 0.019,
+      "step": 923
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1.4240769059075342e-05,
+      "loss": 0.0123,
+      "step": 924
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.387020849529319e-05,
+      "loss": 0.0032,
+      "step": 925
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.3504465085626639e-05,
+      "loss": 0.0044,
+      "step": 926
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.3143542454297885e-05,
+      "loss": 0.0061,
+      "step": 927
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.2787444177759067e-05,
+      "loss": 0.0145,
+      "step": 928
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.243617378465689e-05,
+      "loss": 0.0038,
+      "step": 929
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.208973475579761e-05,
+      "loss": 0.0049,
+      "step": 930
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.1748130524112665e-05,
+      "loss": 0.0022,
+      "step": 931
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.1411364474624264e-05,
+      "loss": 0.0133,
+      "step": 932
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.1079439944412406e-05,
+      "loss": 0.0191,
+      "step": 933
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.0752360222581469e-05,
+      "loss": 0.0045,
+      "step": 934
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.0430128550227624e-05,
+      "loss": 0.008,
+      "step": 935
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.0112748120406857e-05,
+      "loss": 0.0066,
+      "step": 936
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 9.800222078103272e-06,
+      "loss": 0.0041,
+      "step": 937
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 9.492553520197733e-06,
+      "loss": 0.0113,
+      "step": 938
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 9.189745495437607e-06,
+      "loss": 0.0131,
+      "step": 939
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 8.89180100440612e-06,
+      "loss": 0.0025,
+      "step": 940
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 8.59872299949288e-06,
+      "loss": 0.0092,
+      "step": 941
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 8.31051438486441e-06,
+      "loss": 0.0018,
+      "step": 942
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 8.027178016435766e-06,
+      "loss": 0.0059,
+      "step": 943
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 7.748716701841684e-06,
+      "loss": 0.018,
+      "step": 944
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 7.475133200409212e-06,
+      "loss": 0.0161,
+      "step": 945
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 7.206430223130278e-06,
+      "loss": 0.017,
+      "step": 946
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 6.9426104326345e-06,
+      "loss": 0.003,
+      "step": 947
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 6.683676443163312e-06,
+      "loss": 0.0087,
+      "step": 948
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 6.429630820543597e-06,
+      "loss": 0.0113,
+      "step": 949
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 6.180476082162656e-06,
+      "loss": 0.0156,
+      "step": 950
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 5.936214696942887e-06,
+      "loss": 0.005,
+      "step": 951
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 5.696849085317646e-06,
+      "loss": 0.0041,
+      "step": 952
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 5.462381619207091e-06,
+      "loss": 0.0068,
+      "step": 953
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 5.232814621994597e-06,
+      "loss": 0.0027,
+      "step": 954
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 5.008150368503994e-06,
+      "loss": 0.0037,
+      "step": 955
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 4.788391084976862e-06,
+      "loss": 0.0125,
+      "step": 956
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 4.5735389490503265e-06,
+      "loss": 0.0126,
+      "step": 957
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 4.36359608973591e-06,
+      "loss": 0.0088,
+      "step": 958
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 4.158564587397828e-06,
+      "loss": 0.0032,
+      "step": 959
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 3.9584464737330016e-06,
+      "loss": 0.0066,
+      "step": 960
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 3.763243731750521e-06,
+      "loss": 0.0061,
+      "step": 961
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 3.572958295752049e-06,
+      "loss": 0.0032,
+      "step": 962
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 3.387592051312782e-06,
+      "loss": 0.0064,
+      "step": 963
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 3.207146835262742e-06,
+      "loss": 0.0031,
+      "step": 964
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 3.031624435668345e-06,
+      "loss": 0.0152,
+      "step": 965
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 2.861026591815141e-06,
+      "loss": 0.0139,
+      "step": 966
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 2.6953549941900467e-06,
+      "loss": 0.0061,
+      "step": 967
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 2.5346112844650826e-06,
+      "loss": 0.0033,
+      "step": 968
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 2.378797055480608e-06,
+      "loss": 0.0026,
+      "step": 969
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 2.227913851230057e-06,
+      "loss": 0.0067,
+      "step": 970
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.081963166844225e-06,
+      "loss": 0.0045,
+      "step": 971
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.9409464485766747e-06,
+      "loss": 0.0126,
+      "step": 972
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1.804865093789354e-06,
+      "loss": 0.0036,
+      "step": 973
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1.6737204509387205e-06,
+      "loss": 0.0019,
+      "step": 974
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1.5475138195623628e-06,
+      "loss": 0.0085,
+      "step": 975
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1.4262464502663442e-06,
+      "loss": 0.0063,
+      "step": 976
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.309919544712268e-06,
+      "loss": 0.0052,
+      "step": 977
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.1985342556060653e-06,
+      "loss": 0.0166,
+      "step": 978
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.092091686686114e-06,
+      "loss": 0.0158,
+      "step": 979
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 9.905928927123608e-07,
+      "loss": 0.0166,
+      "step": 980
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 8.94038879455994e-07,
+      "loss": 0.0074,
+      "step": 981
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 8.024306036893969e-07,
+      "loss": 0.0084,
+      "step": 982
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 7.157689731767669e-07,
+      "loss": 0.0071,
+      "step": 983
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 6.340548466648444e-07,
+      "loss": 0.0031,
+      "step": 984
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 5.572890338748082e-07,
+      "loss": 0.0131,
+      "step": 985
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.854722954938385e-07,
+      "loss": 0.0016,
+      "step": 986
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 4.1860534316801037e-07,
+      "loss": 0.0054,
+      "step": 987
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 3.566888394948009e-07,
+      "loss": 0.0064,
+      "step": 988
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 2.997233980168157e-07,
+      "loss": 0.0024,
+      "step": 989
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 2.477095832156828e-07,
+      "loss": 0.004,
+      "step": 990
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 2.0064791050633523e-07,
+      "loss": 0.0041,
+      "step": 991
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1.5853884623195924e-07,
+      "loss": 0.0206,
+      "step": 992
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1.2138280765944253e-07,
+      "loss": 0.002,
+      "step": 993
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 8.918016297515542e-08,
+      "loss": 0.005,
+      "step": 994
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 6.193123128134249e-08,
+      "loss": 0.0032,
+      "step": 995
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 3.963628259290308e-08,
+      "loss": 0.0051,
+      "step": 996
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 2.229553783478222e-08,
+      "loss": 0.0031,
+      "step": 997
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 9.90916883986115e-09,
+      "loss": 0.0042,
+      "step": 998
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 2.4772983469589784e-09,
+      "loss": 0.0022,
+      "step": 999
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 0.0,
+      "loss": 0.0035,
+      "step": 1000
+    },
+    {
+      "epoch": 5.08,
+      "step": 1000,
+      "total_flos": 4.507227485768909e+16,
+      "train_loss": 0.2918202262979466,
+      "train_runtime": 2044.5236,
+      "train_samples_per_second": 1.956,
+      "train_steps_per_second": 0.489
+    }
+  ],
+  "max_steps": 1000,
+  "num_train_epochs": 6,
+  "total_flos": 4.507227485768909e+16,
+  "trial_name": null,
+  "trial_params": null
+}