diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13545 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.9964444444444442,
+  "global_step": 2248,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0,
+      "loss": 1.5195,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.701961097713179e-06,
+      "loss": 1.502,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2207319521888542e-05,
+      "loss": 1.5273,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5403922195426358e-05,
+      "loss": 1.4043,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.788339985850974e-05,
+      "loss": 1.5195,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.990928061960172e-05,
+      "loss": 1.2734,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.162213839716128e-05,
+      "loss": 1.248,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3105883293139535e-05,
+      "loss": 1.5273,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4414639043777084e-05,
+      "loss": 1.2754,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.558536095622292e-05,
+      "loss": 1.4668,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6644407746943402e-05,
+      "loss": 1.4023,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.76112417173149e-05,
+      "loss": 1.375,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8500642753555412e-05,
+      "loss": 1.4219,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9324099494874456e-05,
+      "loss": 1.3672,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.009071938039828e-05,
+      "loss": 1.5488,
+      "step": 15
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0807844390852715e-05,
+      "loss": 1.3418,
+      "step": 16
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.14814797916583e-05,
+      "loss": 1.3652,
+      "step": 17
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.211660014149026e-05,
+      "loss": 1.4473,
+      "step": 18
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.271737245444797e-05,
+      "loss": 1.3379,
+      "step": 19
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.32873220539361e-05,
+      "loss": 1.2031,
+      "step": 20
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.3829457919049824e-05,
+      "loss": 1.207,
+      "step": 21
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.434636884465658e-05,
+      "loss": 1.3066,
+      "step": 22
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.484029820864644e-05,
+      "loss": 1.3633,
+      "step": 23
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.5313202815028076e-05,
+      "loss": 1.2363,
+      "step": 24
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.576679971701948e-05,
+      "loss": 1.2246,
+      "step": 25
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.620260385126859e-05,
+      "loss": 1.1367,
+      "step": 26
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.6621958565665625e-05,
+      "loss": 1.166,
+      "step": 27
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.7026060592587636e-05,
+      "loss": 1.0498,
+      "step": 28
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.741598063790252e-05,
+      "loss": 0.9873,
+      "step": 29
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.779268047811146e-05,
+      "loss": 1.0381,
+      "step": 30
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.8157027253033876e-05,
+      "loss": 0.9355,
+      "step": 31
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.850980548856589e-05,
+      "loss": 1.0645,
+      "step": 32
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.885172726883194e-05,
+      "loss": 0.8994,
+      "step": 33
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.9183440889371473e-05,
+      "loss": 0.8838,
+      "step": 34
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.950553825567102e-05,
+      "loss": 0.7305,
+      "step": 35
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.981856123920344e-05,
+      "loss": 0.9619,
+      "step": 36
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.0123007162425155e-05,
+      "loss": 1.002,
+      "step": 37
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.0419333552161145e-05,
+      "loss": 0.9434,
+      "step": 38
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.070796227544395e-05,
+      "loss": 0.7383,
+      "step": 39
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.098928315164928e-05,
+      "loss": 0.8408,
+      "step": 40
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.126365711854374e-05,
+      "loss": 0.7803,
+      "step": 41
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.1531419016763004e-05,
+      "loss": 0.7197,
+      "step": 42
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.17928800466077e-05,
+      "loss": 0.8086,
+      "step": 43
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.204832994236975e-05,
+      "loss": 0.7451,
+      "step": 44
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.2298038902286815e-05,
+      "loss": 0.7715,
+      "step": 45
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.254225930635963e-05,
+      "loss": 0.8271,
+      "step": 46
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.278122724941248e-05,
+      "loss": 0.6108,
+      "step": 47
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.301516391274126e-05,
+      "loss": 0.5938,
+      "step": 48
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.324427679432256e-05,
+      "loss": 0.7217,
+      "step": 49
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.346876081473266e-05,
+      "loss": 0.6826,
+      "step": 50
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.3688799313546835e-05,
+      "loss": 0.5654,
+      "step": 51
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.390456494898177e-05,
+      "loss": 0.6006,
+      "step": 52
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.411622051184135e-05,
+      "loss": 0.6699,
+      "step": 53
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4323919663378806e-05,
+      "loss": 0.5049,
+      "step": 54
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4527807605453143e-05,
+      "loss": 0.6201,
+      "step": 55
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.472802169030082e-05,
+      "loss": 0.563,
+      "step": 56
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.4924691976336514e-05,
+      "loss": 0.4858,
+      "step": 57
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.511794173561569e-05,
+      "loss": 0.6768,
+      "step": 58
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.530788791791773e-05,
+      "loss": 0.5908,
+      "step": 59
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.5494641575824634e-05,
+      "loss": 0.4951,
+      "step": 60
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.567830825466439e-05,
+      "loss": 0.5967,
+      "step": 61
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.5858988350747056e-05,
+      "loss": 0.5864,
+      "step": 62
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.603677744093836e-05,
+      "loss": 0.6055,
+      "step": 63
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.621176658627907e-05,
+      "loss": 0.6143,
+      "step": 64
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.638404261206515e-05,
+      "loss": 0.3906,
+      "step": 65
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.6553688366545114e-05,
+      "loss": 0.542,
+      "step": 66
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.672078296016419e-05,
+      "loss": 0.4946,
+      "step": 67
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.6885401987084654e-05,
+      "loss": 0.4854,
+      "step": 68
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.704761773053499e-05,
+      "loss": 0.4053,
+      "step": 69
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.720749935338421e-05,
+      "loss": 0.498,
+      "step": 70
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.736511307519874e-05,
+      "loss": 0.4761,
+      "step": 71
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.752052233691662e-05,
+      "loss": 0.3892,
+      "step": 72
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.767378795416353e-05,
+      "loss": 0.3999,
+      "step": 73
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.782496826013834e-05,
+      "loss": 0.4722,
+      "step": 74
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.797411923890802e-05,
+      "loss": 0.4014,
+      "step": 75
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.812129464987433e-05,
+      "loss": 0.5249,
+      "step": 76
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.826654614410468e-05,
+      "loss": 0.3853,
+      "step": 77
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.8409923373157134e-05,
+      "loss": 0.3936,
+      "step": 78
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.8551474090973324e-05,
+      "loss": 0.4209,
+      "step": 79
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.869124424936245e-05,
+      "loss": 0.436,
+      "step": 80
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.882927808755417e-05,
+      "loss": 0.3669,
+      "step": 81
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.896561821625692e-05,
+      "loss": 0.3828,
+      "step": 82
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.910030569662156e-05,
+      "loss": 0.3237,
+      "step": 83
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.923338011447617e-05,
+      "loss": 0.4204,
+      "step": 84
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.936487965016803e-05,
+      "loss": 0.4482,
+      "step": 85
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.949484114432088e-05,
+      "loss": 0.4429,
+      "step": 86
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.962330015979106e-05,
+      "loss": 0.3867,
+      "step": 87
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.975029104008293e-05,
+      "loss": 0.4443,
+      "step": 88
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.987584696446389e-05,
+      "loss": 0.4077,
+      "step": 89
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5771,
+      "step": 90
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.325,
+      "step": 91
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 0.5229,
+      "step": 92
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.5234,
+      "step": 93
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.4365,
+      "step": 94
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.3828,
+      "step": 95
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.3745,
+      "step": 96
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.3064,
+      "step": 97
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 0.3901,
+      "step": 98
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.3442,
+      "step": 99
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.3555,
+      "step": 100
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4023,
+      "step": 101
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.3381,
+      "step": 102
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.3882,
+      "step": 103
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 0.4014,
+      "step": 104
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4297,
+      "step": 105
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.364,
+      "step": 106
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.4448,
+      "step": 107
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.3142,
+      "step": 108
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.3579,
+      "step": 109
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.3108,
+      "step": 110
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.3408,
+      "step": 111
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.3599,
+      "step": 112
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.3154,
+      "step": 113
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.4016,
+      "step": 114
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 0.417,
+      "step": 115
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.396,
+      "step": 116
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.3428,
+      "step": 117
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.3833,
+      "step": 118
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.3999,
+      "step": 119
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 0.3472,
+      "step": 120
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4067,
+      "step": 121
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.3135,
+      "step": 122
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.3809,
+      "step": 123
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.342,
+      "step": 124
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2346,
+      "step": 125
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 0.4409,
+      "step": 126
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.3081,
+      "step": 127
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.3303,
+      "step": 128
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.312,
+      "step": 129
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.3452,
+      "step": 130
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.3801,
+      "step": 131
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 0.4082,
+      "step": 132
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.3208,
+      "step": 133
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.3867,
+      "step": 134
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.3696,
+      "step": 135
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.3391,
+      "step": 136
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 0.3354,
+      "step": 137
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2949,
+      "step": 138
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2646,
+      "step": 139
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.3843,
+      "step": 140
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2932,
+      "step": 141
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2725,
+      "step": 142
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 0.3149,
+      "step": 143
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.3003,
+      "step": 144
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.3496,
+      "step": 145
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.3403,
+      "step": 146
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.3091,
+      "step": 147
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.387,
+      "step": 148
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2751,
+      "step": 149
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.3945,
+      "step": 150
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.3252,
+      "step": 151
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.3074,
+      "step": 152
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.397,
+      "step": 153
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 0.3457,
+      "step": 154
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.3293,
+      "step": 155
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.304,
+      "step": 156
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2878,
+      "step": 157
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.3916,
+      "step": 158
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2749,
+      "step": 159
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2869,
+      "step": 160
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.3491,
+      "step": 161
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.3176,
+      "step": 162
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.3,
+      "step": 163
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.4165,
+      "step": 164
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 0.3486,
+      "step": 165
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2925,
+      "step": 166
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.3184,
+      "step": 167
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.3279,
+      "step": 168
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2905,
+      "step": 169
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.3477,
+      "step": 170
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2546,
+      "step": 171
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.3025,
+      "step": 172
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2842,
+      "step": 173
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2761,
+      "step": 174
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.3,
+      "step": 175
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.3391,
+      "step": 176
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2793,
+      "step": 177
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.3684,
+      "step": 178
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2462,
+      "step": 179
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2881,
+      "step": 180
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2947,
+      "step": 181
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 0.3442,
+      "step": 182
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2499,
+      "step": 183
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.3582,
+      "step": 184
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2688,
+      "step": 185
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.3193,
+      "step": 186
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.3379,
+      "step": 187
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 0.3413,
+      "step": 188
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.3105,
+      "step": 189
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.3014,
+      "step": 190
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.3027,
+      "step": 191
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.1904,
+      "step": 192
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.3711,
+      "step": 193
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 0.3167,
+      "step": 194
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.208,
+      "step": 195
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2896,
+      "step": 196
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.3091,
+      "step": 197
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.3496,
+      "step": 198
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2656,
+      "step": 199
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2463,
+      "step": 200
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.311,
+      "step": 201
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.3857,
+      "step": 202
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.3091,
+      "step": 203
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2571,
+      "step": 204
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 0.374,
+      "step": 205
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2903,
+      "step": 206
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.457,
+      "step": 207
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2822,
+      "step": 208
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2981,
+      "step": 209
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2429,
+      "step": 210
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2549,
+      "step": 211
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.3428,
+      "step": 212
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2129,
+      "step": 213
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2756,
+      "step": 214
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.4487,
+      "step": 215
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2869,
+      "step": 216
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.3398,
+      "step": 217
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2717,
+      "step": 218
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.282,
+      "step": 219
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2878,
+      "step": 220
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.3132,
+      "step": 221
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 0.3247,
+      "step": 222
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.3601,
+      "step": 223
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.3164,
+      "step": 224
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.3389,
+      "step": 225
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.3003,
+      "step": 226
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 0.3525,
+      "step": 227
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.3115,
+      "step": 228
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.3149,
+      "step": 229
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2671,
+      "step": 230
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2777,
+      "step": 231
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2283,
+      "step": 232
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2981,
+      "step": 233
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.3242,
+      "step": 234
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.3425,
+      "step": 235
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.3411,
+      "step": 236
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.3115,
+      "step": 237
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.26,
+      "step": 238
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2959,
+      "step": 239
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.3398,
+      "step": 240
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2734,
+      "step": 241
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2476,
+      "step": 242
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2651,
+      "step": 243
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 0.3562,
+      "step": 244
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3682,
+      "step": 245
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3081,
+      "step": 246
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3379,
+      "step": 247
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2805,
+      "step": 248
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3101,
+      "step": 249
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2241,
+      "step": 250
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2996,
+      "step": 251
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3271,
+      "step": 252
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3394,
+      "step": 253
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2415,
+      "step": 254
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 0.3794,
+      "step": 255
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2944,
+      "step": 256
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2708,
+      "step": 257
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2327,
+      "step": 258
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2859,
+      "step": 259
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3003,
+      "step": 260
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2429,
+      "step": 261
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3281,
+      "step": 262
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2925,
+      "step": 263
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3008,
+      "step": 264
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2939,
+      "step": 265
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3083,
+      "step": 266
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2524,
+      "step": 267
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3853,
+      "step": 268
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2405,
+      "step": 269
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3696,
+      "step": 270
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3223,
+      "step": 271
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2278,
+      "step": 272
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2832,
+      "step": 273
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.304,
+      "step": 274
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3313,
+      "step": 275
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3154,
+      "step": 276
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3328,
+      "step": 277
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 0.1874,
+      "step": 278
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.5305,
+      "step": 279
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3442,
+      "step": 280
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2893,
+      "step": 281
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3176,
+      "step": 282
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3167,
+      "step": 283
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3354,
+      "step": 284
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2253,
+      "step": 285
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.344,
+      "step": 286
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2195,
+      "step": 287
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2417,
+      "step": 288
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3108,
+      "step": 289
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2468,
+      "step": 290
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2959,
+      "step": 291
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2554,
+      "step": 292
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3276,
+      "step": 293
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2742,
+      "step": 294
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2915,
+      "step": 295
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.4048,
+      "step": 296
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3318,
+      "step": 297
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3201,
+      "step": 298
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.291,
+      "step": 299
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3521,
+      "step": 300
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.3735,
+      "step": 301
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2754,
+      "step": 302
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.3044,
+      "step": 303
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2571,
+      "step": 304
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2285,
+      "step": 305
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 0.3037,
+      "step": 306
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2817,
+      "step": 307
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3921,
+      "step": 308
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3376,
+      "step": 309
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2488,
+      "step": 310
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.377,
+      "step": 311
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 0.303,
+      "step": 312
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.3618,
+      "step": 313
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2891,
+      "step": 314
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2224,
+      "step": 315
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2393,
+      "step": 316
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2827,
+      "step": 317
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2126,
+      "step": 318
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2271,
+      "step": 319
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2766,
+      "step": 320
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2394,
+      "step": 321
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2229,
+      "step": 322
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 0.3623,
+      "step": 323
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2905,
+      "step": 324
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3286,
+      "step": 325
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.5757,
+      "step": 326
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2117,
+      "step": 327
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3179,
+      "step": 328
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3667,
+      "step": 329
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2468,
+      "step": 330
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2932,
+      "step": 331
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3369,
+      "step": 332
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3025,
+      "step": 333
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2263,
+      "step": 334
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2488,
+      "step": 335
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2935,
+      "step": 336
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2559,
+      "step": 337
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2358,
+      "step": 338
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2869,
+      "step": 339
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 0.3567,
+      "step": 340
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2905,
+      "step": 341
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2756,
+      "step": 342
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2539,
+      "step": 343
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2013,
+      "step": 344
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3057,
+      "step": 345
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2461,
+      "step": 346
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2261,
+      "step": 347
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3125,
+      "step": 348
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.293,
+      "step": 349
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.3044,
+      "step": 350
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2385,
+      "step": 351
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2705,
+      "step": 352
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2217,
+      "step": 353
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2971,
+      "step": 354
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2616,
+      "step": 355
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2343,
+      "step": 356
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2461,
+      "step": 357
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2319,
+      "step": 358
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2427,
+      "step": 359
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.1735,
+      "step": 360
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.321,
+      "step": 361
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2817,
+      "step": 362
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2502,
+      "step": 363
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2646,
+      "step": 364
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2231,
+      "step": 365
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.3367,
+      "step": 366
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.365,
+      "step": 367
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 0.314,
+      "step": 368
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2644,
+      "step": 369
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2529,
+      "step": 370
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2313,
+      "step": 371
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.324,
+      "step": 372
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.3394,
+      "step": 373
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2722,
+      "step": 374
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2446,
+      "step": 375
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2537,
+      "step": 376
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2375,
+      "step": 377
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2986,
+      "step": 378
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2808,
+      "step": 379
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2825,
+      "step": 380
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2583,
+      "step": 381
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3726,
+      "step": 382
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2913,
+      "step": 383
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.1598,
+      "step": 384
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3726,
+      "step": 385
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2786,
+      "step": 386
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.3188,
+      "step": 387
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2422,
+      "step": 388
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2737,
+      "step": 389
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2396,
+      "step": 390
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2551,
+      "step": 391
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2893,
+      "step": 392
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2603,
+      "step": 393
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.6025,
+      "step": 394
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2722,
+      "step": 395
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2069,
+      "step": 396
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2405,
+      "step": 397
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2246,
+      "step": 398
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2346,
+      "step": 399
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.321,
+      "step": 400
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2439,
+      "step": 401
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2087,
+      "step": 402
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3594,
+      "step": 403
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.4214,
+      "step": 404
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2803,
+      "step": 405
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2615,
+      "step": 406
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2371,
+      "step": 407
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.3376,
+      "step": 408
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2866,
+      "step": 409
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2783,
+      "step": 410
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2644,
+      "step": 411
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2986,
+      "step": 412
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2788,
+      "step": 413
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2975,
+      "step": 414
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.267,
+      "step": 415
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2026,
+      "step": 416
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2786,
+      "step": 417
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2695,
+      "step": 418
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3201,
+      "step": 419
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2192,
+      "step": 420
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2456,
+      "step": 421
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3184,
+      "step": 422
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3279,
+      "step": 423
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 0.183,
+      "step": 424
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2971,
+      "step": 425
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2378,
+      "step": 426
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2505,
+      "step": 427
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.3057,
+      "step": 428
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2253,
+      "step": 429
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 0.4541,
+      "step": 430
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2058,
+      "step": 431
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3694,
+      "step": 432
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2131,
+      "step": 433
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.28,
+      "step": 434
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3867,
+      "step": 435
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2058,
+      "step": 436
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2812,
+      "step": 437
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2317,
+      "step": 438
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2427,
+      "step": 439
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2358,
+      "step": 440
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2471,
+      "step": 441
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2534,
+      "step": 442
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.395,
+      "step": 443
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2522,
+      "step": 444
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.1809,
+      "step": 445
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2834,
+      "step": 446
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2461,
+      "step": 447
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3264,
+      "step": 448
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.1843,
+      "step": 449
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3237,
+      "step": 450
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3799,
+      "step": 451
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2661,
+      "step": 452
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2417,
+      "step": 453
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2366,
+      "step": 454
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2729,
+      "step": 455
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3276,
+      "step": 456
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2593,
+      "step": 457
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 0.3035,
+      "step": 458
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2554,
+      "step": 459
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2991,
+      "step": 460
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.4092,
+      "step": 461
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2971,
+      "step": 462
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2922,
+      "step": 463
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2795,
+      "step": 464
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.3157,
+      "step": 465
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2744,
+      "step": 466
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2266,
+      "step": 467
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2969,
+      "step": 468
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2498,
+      "step": 469
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2974,
+      "step": 470
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.1393,
+      "step": 471
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2424,
+      "step": 472
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2659,
+      "step": 473
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2405,
+      "step": 474
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 0.3286,
+      "step": 475
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3237,
+      "step": 476
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.4231,
+      "step": 477
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3303,
+      "step": 478
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2385,
+      "step": 479
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2522,
+      "step": 480
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2441,
+      "step": 481
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2983,
+      "step": 482
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2509,
+      "step": 483
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2422,
+      "step": 484
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2603,
+      "step": 485
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 0.3398,
+      "step": 486
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2466,
+      "step": 487
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2937,
+      "step": 488
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3311,
+      "step": 489
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2673,
+      "step": 490
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2637,
+      "step": 491
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 0.1716,
+      "step": 492
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2305,
+      "step": 493
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2168,
+      "step": 494
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3108,
+      "step": 495
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3105,
+      "step": 496
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 0.1785,
+      "step": 497
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.3538,
+      "step": 498
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2576,
+      "step": 499
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2568,
+      "step": 500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2703,
+      "step": 501
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.416,
+      "step": 502
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2786,
+      "step": 503
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.1975,
+      "step": 504
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2393,
+      "step": 505
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2341,
+      "step": 506
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.23,
+      "step": 507
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2649,
+      "step": 508
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2295,
+      "step": 509
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.3152,
+      "step": 510
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.1573,
+      "step": 511
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2231,
+      "step": 512
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2332,
+      "step": 513
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2383,
+      "step": 514
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.3276,
+      "step": 515
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.3074,
+      "step": 516
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2319,
+      "step": 517
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.3049,
+      "step": 518
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2881,
+      "step": 519
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2583,
+      "step": 520
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.3501,
+      "step": 521
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.3223,
+      "step": 522
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.1954,
+      "step": 523
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2305,
+      "step": 524
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2434,
+      "step": 525
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2756,
+      "step": 526
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.283,
+      "step": 527
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2426,
+      "step": 528
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2808,
+      "step": 529
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2618,
+      "step": 530
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 531
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2585,
+      "step": 532
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2603,
+      "step": 533
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2705,
+      "step": 534
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2913,
+      "step": 535
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.3491,
+      "step": 536
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2267,
+      "step": 537
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2012,
+      "step": 538
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2617,
+      "step": 539
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2534,
+      "step": 540
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.1899,
+      "step": 541
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2751,
+      "step": 542
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.23,
+      "step": 543
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.3792,
+      "step": 544
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2715,
+      "step": 545
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2104,
+      "step": 546
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.174,
+      "step": 547
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2878,
+      "step": 548
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.332,
+      "step": 549
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.1823,
+      "step": 550
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2782,
+      "step": 551
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.3242,
+      "step": 552
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2517,
+      "step": 553
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 0.3254,
+      "step": 554
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2124,
+      "step": 555
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2562,
+      "step": 556
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2321,
+      "step": 557
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2761,
+      "step": 558
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 0.3198,
+      "step": 559
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2412,
+      "step": 560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2361,
+      "step": 561
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2629,
+      "step": 562
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.252667635679245,
+      "eval_runtime": 118.7073,
+      "eval_samples_per_second": 2.527,
+      "eval_steps_per_second": 0.632,
+      "step": 562
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2251,
+      "step": 563
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.3428,
+      "step": 564
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 0.3271,
+      "step": 565
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.1775,
+      "step": 566
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.1478,
+      "step": 567
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.3059,
+      "step": 568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2644,
+      "step": 569
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 0.1985,
+      "step": 570
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2598,
+      "step": 571
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2168,
+      "step": 572
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.1897,
+      "step": 573
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.22,
+      "step": 574
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2847,
+      "step": 575
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2323,
+      "step": 576
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.3604,
+      "step": 577
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2505,
+      "step": 578
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2605,
+      "step": 579
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2537,
+      "step": 580
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2673,
+      "step": 581
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2148,
+      "step": 582
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2368,
+      "step": 583
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2312,
+      "step": 584
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2603,
+      "step": 585
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2112,
+      "step": 586
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2041,
+      "step": 587
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2808,
+      "step": 588
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2622,
+      "step": 589
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2649,
+      "step": 590
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2859,
+      "step": 591
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2219,
+      "step": 592
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2751,
+      "step": 593
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.3069,
+      "step": 594
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.208,
+      "step": 595
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2712,
+      "step": 596
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2417,
+      "step": 597
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2085,
+      "step": 598
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 0.265,
+      "step": 599
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2522,
+      "step": 600
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.217,
+      "step": 601
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2488,
+      "step": 602
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2441,
+      "step": 603
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2773,
+      "step": 604
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.3657,
+      "step": 605
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2677,
+      "step": 606
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2405,
+      "step": 607
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2578,
+      "step": 608
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.3755,
+      "step": 609
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2346,
+      "step": 610
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2319,
+      "step": 611
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.304,
+      "step": 612
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.3589,
+      "step": 613
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.1447,
+      "step": 614
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 0.3394,
+      "step": 615
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2563,
+      "step": 616
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.1664,
+      "step": 617
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2742,
+      "step": 618
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2474,
+      "step": 619
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2517,
+      "step": 620
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2202,
+      "step": 621
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.252,
+      "step": 622
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.1925,
+      "step": 623
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2517,
+      "step": 624
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.3547,
+      "step": 625
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.1815,
+      "step": 626
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2129,
+      "step": 627
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.1929,
+      "step": 628
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2888,
+      "step": 629
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2593,
+      "step": 630
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2756,
+      "step": 631
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 0.3179,
+      "step": 632
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2407,
+      "step": 633
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2981,
+      "step": 634
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2087,
+      "step": 635
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.239,
+      "step": 636
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2695,
+      "step": 637
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 0.3452,
+      "step": 638
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.3323,
+      "step": 639
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.3218,
+      "step": 640
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2683,
+      "step": 641
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2158,
+      "step": 642
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.1976,
+      "step": 643
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2288,
+      "step": 644
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2637,
+      "step": 645
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.29,
+      "step": 646
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2764,
+      "step": 647
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2288,
+      "step": 648
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 0.3267,
+      "step": 649
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.241,
+      "step": 650
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2559,
+      "step": 651
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2683,
+      "step": 652
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2244,
+      "step": 653
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2771,
+      "step": 654
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2693,
+      "step": 655
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2366,
+      "step": 656
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.3186,
+      "step": 657
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2789,
+      "step": 658
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2346,
+      "step": 659
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 0.3081,
+      "step": 660
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2053,
+      "step": 661
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2751,
+      "step": 662
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2314,
+      "step": 663
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.3174,
+      "step": 664
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2659,
+      "step": 665
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2443,
+      "step": 666
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.17,
+      "step": 667
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2405,
+      "step": 668
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.1887,
+      "step": 669
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2256,
+      "step": 670
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2905,
+      "step": 671
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2919,
+      "step": 672
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.3403,
+      "step": 673
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.3018,
+      "step": 674
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.1644,
+      "step": 675
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.1875,
+      "step": 676
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 0.2817,
+      "step": 677
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2349,
+      "step": 678
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2778,
+      "step": 679
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.1979,
+      "step": 680
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2546,
+      "step": 681
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.238,
+      "step": 682
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2163,
+      "step": 683
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2927,
+      "step": 684
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 685
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2346,
+      "step": 686
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2329,
+      "step": 687
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2854,
+      "step": 688
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2664,
+      "step": 689
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.207,
+      "step": 690
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.238,
+      "step": 691
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.3291,
+      "step": 692
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2395,
+      "step": 693
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2649,
+      "step": 694
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.3059,
+      "step": 695
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2266,
+      "step": 696
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2177,
+      "step": 697
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.29,
+      "step": 698
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2351,
+      "step": 699
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 0.3337,
+      "step": 700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2314,
+      "step": 701
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2651,
+      "step": 702
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.321,
+      "step": 703
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2241,
+      "step": 704
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2482,
+      "step": 705
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2288,
+      "step": 706
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2471,
+      "step": 707
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.3274,
+      "step": 708
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2505,
+      "step": 709
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.271,
+      "step": 710
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2202,
+      "step": 711
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2186,
+      "step": 712
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.334,
+      "step": 713
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2351,
+      "step": 714
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2527,
+      "step": 715
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2307,
+      "step": 716
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 0.3252,
+      "step": 717
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.28,
+      "step": 718
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.281,
+      "step": 719
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2346,
+      "step": 720
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2642,
+      "step": 721
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2114,
+      "step": 722
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2365,
+      "step": 723
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.3428,
+      "step": 724
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.3337,
+      "step": 725
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2742,
+      "step": 726
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2197,
+      "step": 727
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2722,
+      "step": 728
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.1876,
+      "step": 729
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2906,
+      "step": 730
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.207,
+      "step": 731
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2355,
+      "step": 732
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2496,
+      "step": 733
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 0.3579,
+      "step": 734
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2181,
+      "step": 735
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.301,
+      "step": 736
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.1458,
+      "step": 737
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.1726,
+      "step": 738
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2053,
+      "step": 739
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2908,
+      "step": 740
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 741
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.3445,
+      "step": 742
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.1598,
+      "step": 743
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.3391,
+      "step": 744
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2283,
+      "step": 745
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2419,
+      "step": 746
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2888,
+      "step": 747
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.1726,
+      "step": 748
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.1996,
+      "step": 749
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 0.197,
+      "step": 750
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2402,
+      "step": 751
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2611,
+      "step": 752
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.3293,
+      "step": 753
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2844,
+      "step": 754
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.3562,
+      "step": 755
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 0.3521,
+      "step": 756
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.1232,
+      "step": 757
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.26,
+      "step": 758
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2673,
+      "step": 759
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.3384,
+      "step": 760
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2058,
+      "step": 761
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 0.239,
+      "step": 762
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2744,
+      "step": 763
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2866,
+      "step": 764
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2305,
+      "step": 765
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2578,
+      "step": 766
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2256,
+      "step": 767
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2419,
+      "step": 768
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2539,
+      "step": 769
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.3115,
+      "step": 770
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.248,
+      "step": 771
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2004,
+      "step": 772
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 0.292,
+      "step": 773
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2422,
+      "step": 774
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.3032,
+      "step": 775
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.1383,
+      "step": 776
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2864,
+      "step": 777
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2998,
+      "step": 778
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2903,
+      "step": 779
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.1334,
+      "step": 780
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.3093,
+      "step": 781
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.3149,
+      "step": 782
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2334,
+      "step": 783
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2705,
+      "step": 784
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2061,
+      "step": 785
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2332,
+      "step": 786
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.3235,
+      "step": 787
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2852,
+      "step": 788
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2859,
+      "step": 789
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2808,
+      "step": 790
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2532,
+      "step": 791
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2461,
+      "step": 792
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2583,
+      "step": 793
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2236,
+      "step": 794
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2415,
+      "step": 795
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2224,
+      "step": 796
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.3125,
+      "step": 797
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2505,
+      "step": 798
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2385,
+      "step": 799
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2542,
+      "step": 800
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2688,
+      "step": 801
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.1743,
+      "step": 802
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2317,
+      "step": 803
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2729,
+      "step": 804
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2898,
+      "step": 805
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2925,
+      "step": 806
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2473,
+      "step": 807
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2319,
+      "step": 808
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2958,
+      "step": 809
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2415,
+      "step": 810
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2373,
+      "step": 811
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2834,
+      "step": 812
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2908,
+      "step": 813
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.239,
+      "step": 814
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2136,
+      "step": 815
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2247,
+      "step": 816
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2666,
+      "step": 817
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2405,
+      "step": 818
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2349,
+      "step": 819
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.3616,
+      "step": 820
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2981,
+      "step": 821
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2571,
+      "step": 822
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.5605,
+      "step": 823
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2502,
+      "step": 824
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.3228,
+      "step": 825
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.5198,
+      "step": 826
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.293,
+      "step": 827
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2013,
+      "step": 828
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2502,
+      "step": 829
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.1412,
+      "step": 830
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2676,
+      "step": 831
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2012,
+      "step": 832
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3557,
+      "step": 833
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2876,
+      "step": 834
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2185,
+      "step": 835
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2261,
+      "step": 836
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3462,
+      "step": 837
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2522,
+      "step": 838
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2676,
+      "step": 839
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2078,
+      "step": 840
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3359,
+      "step": 841
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2235,
+      "step": 842
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.1982,
+      "step": 843
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2051,
+      "step": 844
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2493,
+      "step": 845
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2607,
+      "step": 846
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.1639,
+      "step": 847
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.24,
+      "step": 848
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2837,
+      "step": 849
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2568,
+      "step": 850
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2168,
+      "step": 851
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 0.3147,
+      "step": 852
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2869,
+      "step": 853
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2592,
+      "step": 854
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2859,
+      "step": 855
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2661,
+      "step": 856
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 0.29,
+      "step": 857
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2595,
+      "step": 858
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3188,
+      "step": 859
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3855,
+      "step": 860
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2156,
+      "step": 861
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2834,
+      "step": 862
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2771,
+      "step": 863
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.3206,
+      "step": 864
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2817,
+      "step": 865
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2113,
+      "step": 866
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2043,
+      "step": 867
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2375,
+      "step": 868
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2734,
+      "step": 869
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3062,
+      "step": 870
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.1943,
+      "step": 871
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3521,
+      "step": 872
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2212,
+      "step": 873
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2603,
+      "step": 874
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2871,
+      "step": 875
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2307,
+      "step": 876
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2175,
+      "step": 877
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2654,
+      "step": 878
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.1907,
+      "step": 879
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 0.354,
+      "step": 880
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.1775,
+      "step": 881
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.29,
+      "step": 882
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2231,
+      "step": 883
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2202,
+      "step": 884
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 0.23,
+      "step": 885
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2668,
+      "step": 886
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.1831,
+      "step": 887
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2472,
+      "step": 888
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.274,
+      "step": 889
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2554,
+      "step": 890
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2974,
+      "step": 891
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2466,
+      "step": 892
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2177,
+      "step": 893
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2119,
+      "step": 894
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2666,
+      "step": 895
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2368,
+      "step": 896
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2529,
+      "step": 897
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2598,
+      "step": 898
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2855,
+      "step": 899
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2825,
+      "step": 900
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2898,
+      "step": 901
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2556,
+      "step": 902
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2356,
+      "step": 903
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2505,
+      "step": 904
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.3186,
+      "step": 905
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.1616,
+      "step": 906
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.1885,
+      "step": 907
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2517,
+      "step": 908
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2391,
+      "step": 909
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2292,
+      "step": 910
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2659,
+      "step": 911
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.311,
+      "step": 912
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2112,
+      "step": 913
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2122,
+      "step": 914
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2683,
+      "step": 915
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2566,
+      "step": 916
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.1743,
+      "step": 917
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2571,
+      "step": 918
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2205,
+      "step": 919
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.293,
+      "step": 920
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2373,
+      "step": 921
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2478,
+      "step": 922
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.3425,
+      "step": 923
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2256,
+      "step": 924
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2461,
+      "step": 925
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2231,
+      "step": 926
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2042,
+      "step": 927
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2292,
+      "step": 928
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2625,
+      "step": 929
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 0.241,
+      "step": 930
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2212,
+      "step": 931
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.3372,
+      "step": 932
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.3584,
+      "step": 933
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.3555,
+      "step": 934
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2239,
+      "step": 935
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2751,
+      "step": 936
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2139,
+      "step": 937
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2513,
+      "step": 938
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2139,
+      "step": 939
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.3093,
+      "step": 940
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.3103,
+      "step": 941
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2585,
+      "step": 942
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2324,
+      "step": 943
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3438,
+      "step": 944
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2161,
+      "step": 945
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.281,
+      "step": 946
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2839,
+      "step": 947
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2761,
+      "step": 948
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.271,
+      "step": 949
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2375,
+      "step": 950
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2281,
+      "step": 951
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2542,
+      "step": 952
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2595,
+      "step": 953
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3069,
+      "step": 954
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2729,
+      "step": 955
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2313,
+      "step": 956
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2544,
+      "step": 957
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2078,
+      "step": 958
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2317,
+      "step": 959
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2083,
+      "step": 960
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2042,
+      "step": 961
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2202,
+      "step": 962
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.1926,
+      "step": 963
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 0.229,
+      "step": 964
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.313,
+      "step": 965
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2089,
+      "step": 966
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2734,
+      "step": 967
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2246,
+      "step": 968
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2144,
+      "step": 969
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2256,
+      "step": 970
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.1729,
+      "step": 971
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2241,
+      "step": 972
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2123,
+      "step": 973
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2693,
+      "step": 974
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2322,
+      "step": 975
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.187,
+      "step": 976
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2073,
+      "step": 977
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3552,
+      "step": 978
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2458,
+      "step": 979
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2288,
+      "step": 980
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 0.24,
+      "step": 981
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2336,
+      "step": 982
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2236,
+      "step": 983
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2137,
+      "step": 984
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2581,
+      "step": 985
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2776,
+      "step": 986
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3352,
+      "step": 987
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2625,
+      "step": 988
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2166,
+      "step": 989
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2676,
+      "step": 990
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2988,
+      "step": 991
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2056,
+      "step": 992
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2859,
+      "step": 993
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3035,
+      "step": 994
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2964,
+      "step": 995
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.249,
+      "step": 996
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2417,
+      "step": 997
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2686,
+      "step": 998
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2852,
+      "step": 999
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.1995,
+      "step": 1000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2571,
+      "step": 1001
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2341,
+      "step": 1002
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2952,
+      "step": 1003
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 0.3223,
+      "step": 1004
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.1619,
+      "step": 1005
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2419,
+      "step": 1006
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.1934,
+      "step": 1007
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2788,
+      "step": 1008
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2148,
+      "step": 1009
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2361,
+      "step": 1010
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2439,
+      "step": 1011
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2241,
+      "step": 1012
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2212,
+      "step": 1013
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.22,
+      "step": 1014
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2827,
+      "step": 1015
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2261,
+      "step": 1016
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2598,
+      "step": 1017
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2732,
+      "step": 1018
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2439,
+      "step": 1019
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 0.1848,
+      "step": 1020
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.1942,
+      "step": 1021
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2727,
+      "step": 1022
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.3223,
+      "step": 1023
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.163,
+      "step": 1024
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2322,
+      "step": 1025
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2832,
+      "step": 1026
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2678,
+      "step": 1027
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2168,
+      "step": 1028
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2285,
+      "step": 1029
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2065,
+      "step": 1030
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2605,
+      "step": 1031
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2139,
+      "step": 1032
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2922,
+      "step": 1033
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.3105,
+      "step": 1034
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2598,
+      "step": 1035
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2954,
+      "step": 1036
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2847,
+      "step": 1037
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2488,
+      "step": 1038
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2163,
+      "step": 1039
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3206,
+      "step": 1040
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2505,
+      "step": 1041
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3101,
+      "step": 1042
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 0.3098,
+      "step": 1043
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2517,
+      "step": 1044
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2554,
+      "step": 1045
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2314,
+      "step": 1046
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2441,
+      "step": 1047
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.3342,
+      "step": 1048
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2163,
+      "step": 1049
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2705,
+      "step": 1050
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2864,
+      "step": 1051
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2495,
+      "step": 1052
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2009,
+      "step": 1053
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 0.293,
+      "step": 1054
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2314,
+      "step": 1055
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2141,
+      "step": 1056
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2144,
+      "step": 1057
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2231,
+      "step": 1058
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.25,
+      "step": 1059
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2847,
+      "step": 1060
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2725,
+      "step": 1061
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2695,
+      "step": 1062
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2949,
+      "step": 1063
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2693,
+      "step": 1064
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 0.1851,
+      "step": 1065
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.241,
+      "step": 1066
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2302,
+      "step": 1067
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.3584,
+      "step": 1068
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.3057,
+      "step": 1069
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.3167,
+      "step": 1070
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2935,
+      "step": 1071
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2458,
+      "step": 1072
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.233,
+      "step": 1073
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2834,
+      "step": 1074
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2854,
+      "step": 1075
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2375,
+      "step": 1076
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 0.1857,
+      "step": 1077
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2356,
+      "step": 1078
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.1979,
+      "step": 1079
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.229,
+      "step": 1080
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.5784,
+      "step": 1081
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2988,
+      "step": 1082
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2188,
+      "step": 1083
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2106,
+      "step": 1084
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.1919,
+      "step": 1085
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2097,
+      "step": 1086
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2764,
+      "step": 1087
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2679,
+      "step": 1088
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.4344,
+      "step": 1089
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2561,
+      "step": 1090
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2908,
+      "step": 1091
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2886,
+      "step": 1092
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2639,
+      "step": 1093
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2101,
+      "step": 1094
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2529,
+      "step": 1095
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2314,
+      "step": 1096
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2065,
+      "step": 1097
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2383,
+      "step": 1098
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2209,
+      "step": 1099
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.168,
+      "step": 1100
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2383,
+      "step": 1101
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2756,
+      "step": 1102
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2355,
+      "step": 1103
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2319,
+      "step": 1104
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 0.285,
+      "step": 1105
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.3091,
+      "step": 1106
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2324,
+      "step": 1107
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.209,
+      "step": 1108
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2629,
+      "step": 1109
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2019,
+      "step": 1110
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2178,
+      "step": 1111
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2805,
+      "step": 1112
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2764,
+      "step": 1113
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.3508,
+      "step": 1114
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.302,
+      "step": 1115
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2246,
+      "step": 1116
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2211,
+      "step": 1117
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2013,
+      "step": 1118
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.23,
+      "step": 1119
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.1913,
+      "step": 1120
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.1936,
+      "step": 1121
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 0.1804,
+      "step": 1122
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.186,
+      "step": 1123
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.3025,
+      "step": 1124
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.3154,
+      "step": 1125
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.23581379652023315,
+      "eval_runtime": 118.6461,
+      "eval_samples_per_second": 2.529,
+      "eval_steps_per_second": 0.632,
+      "step": 1125
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2822,
+      "step": 1126
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 0.277,
+      "step": 1127
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.1865,
+      "step": 1128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.1844,
+      "step": 1129
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2039,
+      "step": 1130
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2429,
+      "step": 1131
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.1929,
+      "step": 1132
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5e-05,
+      "loss": 0.269,
+      "step": 1133
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2842,
+      "step": 1134
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2139,
+      "step": 1135
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.1885,
+      "step": 1136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2375,
+      "step": 1137
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2405,
+      "step": 1138
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2959,
+      "step": 1139
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2449,
+      "step": 1140
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2913,
+      "step": 1141
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.23,
+      "step": 1142
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2971,
+      "step": 1143
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2568,
+      "step": 1144
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2871,
+      "step": 1145
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2424,
+      "step": 1146
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.3442,
+      "step": 1147
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.3494,
+      "step": 1148
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2274,
+      "step": 1149
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2473,
+      "step": 1150
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.1959,
+      "step": 1151
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2699,
+      "step": 1152
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2465,
+      "step": 1153
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2185,
+      "step": 1154
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5e-05,
+      "loss": 0.1722,
+      "step": 1155
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.3369,
+      "step": 1156
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.1997,
+      "step": 1157
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2209,
+      "step": 1158
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.1896,
+      "step": 1159
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2385,
+      "step": 1160
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5e-05,
+      "loss": 0.1995,
+      "step": 1161
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.29,
+      "step": 1162
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.353,
+      "step": 1163
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2576,
+      "step": 1164
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2144,
+      "step": 1165
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2305,
+      "step": 1166
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2056,
+      "step": 1167
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2234,
+      "step": 1168
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2893,
+      "step": 1169
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.1848,
+      "step": 1170
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2356,
+      "step": 1171
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2554,
+      "step": 1172
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2124,
+      "step": 1173
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.228,
+      "step": 1174
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2339,
+      "step": 1175
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.3523,
+      "step": 1176
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2516,
+      "step": 1177
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2192,
+      "step": 1178
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2004,
+      "step": 1179
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2,
+      "step": 1180
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2434,
+      "step": 1181
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.1611,
+      "step": 1182
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2432,
+      "step": 1183
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2786,
+      "step": 1184
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2231,
+      "step": 1185
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2727,
+      "step": 1186
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.1921,
+      "step": 1187
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2773,
+      "step": 1188
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2527,
+      "step": 1189
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2261,
+      "step": 1190
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2693,
+      "step": 1191
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.3198,
+      "step": 1192
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.1754,
+      "step": 1193
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.3186,
+      "step": 1194
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2144,
+      "step": 1195
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2341,
+      "step": 1196
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2874,
+      "step": 1197
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.301,
+      "step": 1198
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2622,
+      "step": 1199
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2083,
+      "step": 1200
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2827,
+      "step": 1201
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2483,
+      "step": 1202
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2302,
+      "step": 1203
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2274,
+      "step": 1204
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.254,
+      "step": 1205
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5e-05,
+      "loss": 0.1959,
+      "step": 1206
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.1366,
+      "step": 1207
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2732,
+      "step": 1208
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.3005,
+      "step": 1209
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2654,
+      "step": 1210
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2366,
+      "step": 1211
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2047,
+      "step": 1212
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2416,
+      "step": 1213
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2502,
+      "step": 1214
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2158,
+      "step": 1215
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2371,
+      "step": 1216
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2917,
+      "step": 1217
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2334,
+      "step": 1218
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2556,
+      "step": 1219
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2209,
+      "step": 1220
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2976,
+      "step": 1221
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2686,
+      "step": 1222
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5e-05,
+      "loss": 0.1953,
+      "step": 1223
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2783,
+      "step": 1224
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2423,
+      "step": 1225
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.324,
+      "step": 1226
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.1836,
+      "step": 1227
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2856,
+      "step": 1228
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2083,
+      "step": 1229
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2234,
+      "step": 1230
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.3203,
+      "step": 1231
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2043,
+      "step": 1232
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.1584,
+      "step": 1233
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2634,
+      "step": 1234
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.3062,
+      "step": 1235
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.2971,
+      "step": 1236
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.186,
+      "step": 1237
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.1873,
+      "step": 1238
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.1968,
+      "step": 1239
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5e-05,
+      "loss": 0.2444,
+      "step": 1240
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.3411,
+      "step": 1241
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.3262,
+      "step": 1242
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2202,
+      "step": 1243
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2336,
+      "step": 1244
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2626,
+      "step": 1245
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.1648,
+      "step": 1246
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.1711,
+      "step": 1247
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.271,
+      "step": 1248
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2949,
+      "step": 1249
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2979,
+      "step": 1250
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2751,
+      "step": 1251
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2886,
+      "step": 1252
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.261,
+      "step": 1253
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2622,
+      "step": 1254
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.1836,
+      "step": 1255
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2122,
+      "step": 1256
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5e-05,
+      "loss": 0.3105,
+      "step": 1257
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.1407,
+      "step": 1258
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2787,
+      "step": 1259
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2252,
+      "step": 1260
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2001,
+      "step": 1261
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2393,
+      "step": 1262
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2988,
+      "step": 1263
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2954,
+      "step": 1264
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2367,
+      "step": 1265
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2512,
+      "step": 1266
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2256,
+      "step": 1267
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2712,
+      "step": 1268
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2939,
+      "step": 1269
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2303,
+      "step": 1270
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2273,
+      "step": 1271
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2349,
+      "step": 1272
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.3071,
+      "step": 1273
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5e-05,
+      "loss": 0.1632,
+      "step": 1274
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2573,
+      "step": 1275
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.1812,
+      "step": 1276
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.1543,
+      "step": 1277
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.1907,
+      "step": 1278
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5e-05,
+      "loss": 0.231,
+      "step": 1279
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.301,
+      "step": 1280
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2163,
+      "step": 1281
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.302,
+      "step": 1282
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2256,
+      "step": 1283
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2424,
+      "step": 1284
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2393,
+      "step": 1285
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2336,
+      "step": 1286
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2109,
+      "step": 1287
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2173,
+      "step": 1288
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.1963,
+      "step": 1289
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2134,
+      "step": 1290
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.1893,
+      "step": 1291
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2251,
+      "step": 1292
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2607,
+      "step": 1293
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2053,
+      "step": 1294
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2637,
+      "step": 1295
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2573,
+      "step": 1296
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2488,
+      "step": 1297
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.1818,
+      "step": 1298
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2854,
+      "step": 1299
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2488,
+      "step": 1300
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.1667,
+      "step": 1301
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5e-05,
+      "loss": 0.167,
+      "step": 1302
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2495,
+      "step": 1303
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2991,
+      "step": 1304
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2263,
+      "step": 1305
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2737,
+      "step": 1306
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5e-05,
+      "loss": 0.1915,
+      "step": 1307
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2341,
+      "step": 1308
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2354,
+      "step": 1309
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2612,
+      "step": 1310
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2766,
+      "step": 1311
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.1638,
+      "step": 1312
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5e-05,
+      "loss": 0.3032,
+      "step": 1313
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2794,
+      "step": 1314
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.1854,
+      "step": 1315
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2705,
+      "step": 1316
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.1953,
+      "step": 1317
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.24,
+      "step": 1318
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2229,
+      "step": 1319
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.1958,
+      "step": 1320
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2686,
+      "step": 1321
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2415,
+      "step": 1322
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2131,
+      "step": 1323
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5e-05,
+      "loss": 0.1895,
+      "step": 1324
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.3933,
+      "step": 1325
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2073,
+      "step": 1326
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2078,
+      "step": 1327
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.3374,
+      "step": 1328
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2722,
+      "step": 1329
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5e-05,
+      "loss": 0.1884,
+      "step": 1330
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2141,
+      "step": 1331
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2571,
+      "step": 1332
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2246,
+      "step": 1333
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.3091,
+      "step": 1334
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2375,
+      "step": 1335
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.3208,
+      "step": 1336
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2229,
+      "step": 1337
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2262,
+      "step": 1338
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2153,
+      "step": 1339
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.163,
+      "step": 1340
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5e-05,
+      "loss": 0.543,
+      "step": 1341
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2397,
+      "step": 1342
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.239,
+      "step": 1343
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2749,
+      "step": 1344
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2043,
+      "step": 1345
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2449,
+      "step": 1346
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2561,
+      "step": 1347
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2017,
+      "step": 1348
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.281,
+      "step": 1349
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2817,
+      "step": 1350
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2317,
+      "step": 1351
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2922,
+      "step": 1352
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.1699,
+      "step": 1353
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2554,
+      "step": 1354
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2156,
+      "step": 1355
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2302,
+      "step": 1356
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2063,
+      "step": 1357
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2244,
+      "step": 1358
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2437,
+      "step": 1359
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2107,
+      "step": 1360
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.1963,
+      "step": 1361
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2256,
+      "step": 1362
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2102,
+      "step": 1363
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5e-05,
+      "loss": 0.248,
+      "step": 1364
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.3167,
+      "step": 1365
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2146,
+      "step": 1366
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2454,
+      "step": 1367
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.1787,
+      "step": 1368
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2672,
+      "step": 1369
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2239,
+      "step": 1370
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.1742,
+      "step": 1371
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2056,
+      "step": 1372
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.1946,
+      "step": 1373
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2505,
+      "step": 1374
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2242,
+      "step": 1375
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2131,
+      "step": 1376
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.1797,
+      "step": 1377
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.1912,
+      "step": 1378
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2788,
+      "step": 1379
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2222,
+      "step": 1380
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2893,
+      "step": 1381
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2424,
+      "step": 1382
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2073,
+      "step": 1383
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2532,
+      "step": 1384
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2025,
+      "step": 1385
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 5e-05,
+      "loss": 0.1541,
+      "step": 1386
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2307,
+      "step": 1387
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.251,
+      "step": 1388
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2368,
+      "step": 1389
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2188,
+      "step": 1390
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.1624,
+      "step": 1391
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2229,
+      "step": 1392
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2988,
+      "step": 1393
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3169,
+      "step": 1394
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2808,
+      "step": 1395
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.1835,
+      "step": 1396
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2257,
+      "step": 1397
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2266,
+      "step": 1398
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2092,
+      "step": 1399
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.1908,
+      "step": 1400
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2698,
+      "step": 1401
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.238,
+      "step": 1402
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3022,
+      "step": 1403
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.1851,
+      "step": 1404
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2263,
+      "step": 1405
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.3325,
+      "step": 1406
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2332,
+      "step": 1407
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2812,
+      "step": 1408
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 5e-05,
+      "loss": 0.1727,
+      "step": 1409
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2661,
+      "step": 1410
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2542,
+      "step": 1411
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2412,
+      "step": 1412
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2562,
+      "step": 1413
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 5e-05,
+      "loss": 0.1968,
+      "step": 1414
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.313,
+      "step": 1415
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2864,
+      "step": 1416
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2292,
+      "step": 1417
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.4523,
+      "step": 1418
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2764,
+      "step": 1419
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2803,
+      "step": 1420
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2952,
+      "step": 1421
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2959,
+      "step": 1422
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2649,
+      "step": 1423
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2598,
+      "step": 1424
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2769,
+      "step": 1425
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.27,
+      "step": 1426
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.238,
+      "step": 1427
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2435,
+      "step": 1428
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.1917,
+      "step": 1429
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.1946,
+      "step": 1430
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5e-05,
+      "loss": 0.1556,
+      "step": 1431
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2463,
+      "step": 1432
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2234,
+      "step": 1433
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.1826,
+      "step": 1434
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2136,
+      "step": 1435
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.3103,
+      "step": 1436
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2031,
+      "step": 1437
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.1917,
+      "step": 1438
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2139,
+      "step": 1439
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2722,
+      "step": 1440
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2527,
+      "step": 1441
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2466,
+      "step": 1442
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2485,
+      "step": 1443
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2231,
+      "step": 1444
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2678,
+      "step": 1445
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.1929,
+      "step": 1446
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.28,
+      "step": 1447
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2419,
+      "step": 1448
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2463,
+      "step": 1449
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2273,
+      "step": 1450
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.1821,
+      "step": 1451
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.3016,
+      "step": 1452
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2458,
+      "step": 1453
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2281,
+      "step": 1454
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.222,
+      "step": 1455
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2341,
+      "step": 1456
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2644,
+      "step": 1457
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.3013,
+      "step": 1458
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2031,
+      "step": 1459
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2314,
+      "step": 1460
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.1729,
+      "step": 1461
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2275,
+      "step": 1462
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2407,
+      "step": 1463
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2288,
+      "step": 1464
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5e-05,
+      "loss": 0.1841,
+      "step": 1465
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.218,
+      "step": 1466
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2395,
+      "step": 1467
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2029,
+      "step": 1468
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2881,
+      "step": 1469
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2478,
+      "step": 1470
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2463,
+      "step": 1471
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2344,
+      "step": 1472
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.229,
+      "step": 1473
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2341,
+      "step": 1474
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.1992,
+      "step": 1475
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2271,
+      "step": 1476
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2903,
+      "step": 1477
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2739,
+      "step": 1478
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2603,
+      "step": 1479
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2247,
+      "step": 1480
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.1833,
+      "step": 1481
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2468,
+      "step": 1482
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2615,
+      "step": 1483
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2737,
+      "step": 1484
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2529,
+      "step": 1485
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.269,
+      "step": 1486
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2507,
+      "step": 1487
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.1719,
+      "step": 1488
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4253,
+      "step": 1489
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.1656,
+      "step": 1490
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2341,
+      "step": 1491
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2153,
+      "step": 1492
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5e-05,
+      "loss": 0.1577,
+      "step": 1493
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2427,
+      "step": 1494
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2693,
+      "step": 1495
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2061,
+      "step": 1496
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2322,
+      "step": 1497
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.217,
+      "step": 1498
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2568,
+      "step": 1499
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2983,
+      "step": 1500
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2944,
+      "step": 1501
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.1931,
+      "step": 1502
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2357,
+      "step": 1503
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2568,
+      "step": 1504
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.1809,
+      "step": 1505
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.24,
+      "step": 1506
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.239,
+      "step": 1507
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2622,
+      "step": 1508
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2959,
+      "step": 1509
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2136,
+      "step": 1510
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2446,
+      "step": 1511
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2563,
+      "step": 1512
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2128,
+      "step": 1513
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.157,
+      "step": 1514
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5e-05,
+      "loss": 0.1681,
+      "step": 1515
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2512,
+      "step": 1516
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2473,
+      "step": 1517
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2722,
+      "step": 1518
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2549,
+      "step": 1519
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.1853,
+      "step": 1520
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5e-05,
+      "loss": 0.3586,
+      "step": 1521
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.1963,
+      "step": 1522
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.1868,
+      "step": 1523
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.3066,
+      "step": 1524
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.1772,
+      "step": 1525
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2487,
+      "step": 1526
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2112,
+      "step": 1527
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 0.3005,
+      "step": 1528
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2592,
+      "step": 1529
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 0.1902,
+      "step": 1530
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2495,
+      "step": 1531
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2434,
+      "step": 1532
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.1985,
+      "step": 1533
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2024,
+      "step": 1534
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.231,
+      "step": 1535
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2303,
+      "step": 1536
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2629,
+      "step": 1537
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2434,
+      "step": 1538
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2578,
+      "step": 1539
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2048,
+      "step": 1540
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2222,
+      "step": 1541
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2456,
+      "step": 1542
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.1775,
+      "step": 1543
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 5e-05,
+      "loss": 0.241,
+      "step": 1544
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2349,
+      "step": 1545
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 0.3438,
+      "step": 1546
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2634,
+      "step": 1547
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2866,
+      "step": 1548
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2229,
+      "step": 1549
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2502,
+      "step": 1550
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.1836,
+      "step": 1551
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2079,
+      "step": 1552
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2363,
+      "step": 1553
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2341,
+      "step": 1554
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 5e-05,
+      "loss": 0.3193,
+      "step": 1555
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 0.1658,
+      "step": 1556
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2202,
+      "step": 1557
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3162,
+      "step": 1558
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2229,
+      "step": 1559
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2095,
+      "step": 1560
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2188,
+      "step": 1561
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2878,
+      "step": 1562
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2178,
+      "step": 1563
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2246,
+      "step": 1564
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2457,
+      "step": 1565
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2156,
+      "step": 1566
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.1982,
+      "step": 1567
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3054,
+      "step": 1568
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2345,
+      "step": 1569
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.1741,
+      "step": 1570
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2351,
+      "step": 1571
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2148,
+      "step": 1572
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2922,
+      "step": 1573
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2402,
+      "step": 1574
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 0.3232,
+      "step": 1575
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2361,
+      "step": 1576
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 5e-05,
+      "loss": 0.196,
+      "step": 1577
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2422,
+      "step": 1578
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.238,
+      "step": 1579
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2461,
+      "step": 1580
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.4678,
+      "step": 1581
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2581,
+      "step": 1582
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 5e-05,
+      "loss": 0.1812,
+      "step": 1583
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2581,
+      "step": 1584
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.231,
+      "step": 1585
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.3105,
+      "step": 1586
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2144,
+      "step": 1587
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2839,
+      "step": 1588
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 5e-05,
+      "loss": 0.3125,
+      "step": 1589
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2125,
+      "step": 1590
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 0.1978,
+      "step": 1591
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2935,
+      "step": 1592
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2288,
+      "step": 1593
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2354,
+      "step": 1594
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.218,
+      "step": 1595
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2822,
+      "step": 1596
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2534,
+      "step": 1597
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.1609,
+      "step": 1598
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.1973,
+      "step": 1599
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5e-05,
+      "loss": 0.1998,
+      "step": 1600
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2307,
+      "step": 1601
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2467,
+      "step": 1602
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2805,
+      "step": 1603
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2676,
+      "step": 1604
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5e-05,
+      "loss": 0.1936,
+      "step": 1605
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2539,
+      "step": 1606
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.1743,
+      "step": 1607
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2468,
+      "step": 1608
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.213,
+      "step": 1609
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.293,
+      "step": 1610
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2153,
+      "step": 1611
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.5334,
+      "step": 1612
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2527,
+      "step": 1613
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2307,
+      "step": 1614
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.1698,
+      "step": 1615
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2345,
+      "step": 1616
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2014,
+      "step": 1617
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2908,
+      "step": 1618
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2322,
+      "step": 1619
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3005,
+      "step": 1620
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2979,
+      "step": 1621
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2289,
+      "step": 1622
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2639,
+      "step": 1623
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.1782,
+      "step": 1624
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2271,
+      "step": 1625
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.1671,
+      "step": 1626
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2131,
+      "step": 1627
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5e-05,
+      "loss": 0.1765,
+      "step": 1628
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.3276,
+      "step": 1629
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2712,
+      "step": 1630
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.3064,
+      "step": 1631
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2727,
+      "step": 1632
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2869,
+      "step": 1633
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5e-05,
+      "loss": 0.1658,
+      "step": 1634
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 0.1802,
+      "step": 1635
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 0.1499,
+      "step": 1636
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 0.1692,
+      "step": 1637
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2272,
+      "step": 1638
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2417,
+      "step": 1639
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2032,
+      "step": 1640
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2649,
+      "step": 1641
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2656,
+      "step": 1642
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2104,
+      "step": 1643
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.3284,
+      "step": 1644
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2351,
+      "step": 1645
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2219,
+      "step": 1646
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2048,
+      "step": 1647
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2423,
+      "step": 1648
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2371,
+      "step": 1649
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5e-05,
+      "loss": 0.3093,
+      "step": 1650
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2585,
+      "step": 1651
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2906,
+      "step": 1652
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2749,
+      "step": 1653
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2979,
+      "step": 1654
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2891,
+      "step": 1655
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2263,
+      "step": 1656
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.266,
+      "step": 1657
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2346,
+      "step": 1658
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2461,
+      "step": 1659
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.1484,
+      "step": 1660
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.1934,
+      "step": 1661
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5e-05,
+      "loss": 0.1975,
+      "step": 1662
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2313,
+      "step": 1663
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 0.1802,
+      "step": 1664
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 0.1941,
+      "step": 1665
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 0.281,
+      "step": 1666
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5e-05,
+      "loss": 0.1815,
+      "step": 1667
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.1326,
+      "step": 1668
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.1619,
+      "step": 1669
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2898,
+      "step": 1670
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2554,
+      "step": 1671
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2147,
+      "step": 1672
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2424,
+      "step": 1673
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2722,
+      "step": 1674
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2437,
+      "step": 1675
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2344,
+      "step": 1676
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.172,
+      "step": 1677
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2163,
+      "step": 1678
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2555,
+      "step": 1679
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 0.1765,
+      "step": 1680
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2476,
+      "step": 1681
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2815,
+      "step": 1682
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2935,
+      "step": 1683
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2576,
+      "step": 1684
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 0.1968,
+      "step": 1685
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 0.3438,
+      "step": 1686
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2065,
+      "step": 1687
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.22996826469898224,
+      "eval_runtime": 118.5591,
+      "eval_samples_per_second": 2.53,
+      "eval_steps_per_second": 0.633,
+      "step": 1687
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2205,
+      "step": 1688
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2266,
+      "step": 1689
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2649,
+      "step": 1690
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5e-05,
+      "loss": 0.1531,
+      "step": 1691
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5e-05,
+      "loss": 0.146,
+      "step": 1692
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2344,
+      "step": 1693
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2935,
+      "step": 1694
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5e-05,
+      "loss": 0.2887,
+      "step": 1695
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5e-05,
+      "loss": 0.3352,
+      "step": 1696
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5e-05,
+      "loss": 0.1514,
+      "step": 1697
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5e-05,
+      "loss": 0.196,
+      "step": 1698
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5e-05,
+      "loss": 0.1816,
+      "step": 1699
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2502,
+      "step": 1700
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5e-05,
+      "loss": 0.2825,
+      "step": 1701
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2666,
+      "step": 1702
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5e-05,
+      "loss": 0.197,
+      "step": 1703
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2356,
+      "step": 1704
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5e-05,
+      "loss": 0.3115,
+      "step": 1705
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2615,
+      "step": 1706
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 5e-05,
+      "loss": 0.2029,
+      "step": 1707
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2717,
+      "step": 1708
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2566,
+      "step": 1709
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2029,
+      "step": 1710
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2903,
+      "step": 1711
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 5e-05,
+      "loss": 0.2803,
+      "step": 1712
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2363,
+      "step": 1713
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2366,
+      "step": 1714
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2385,
+      "step": 1715
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5e-05,
+      "loss": 0.1902,
+      "step": 1716
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2263,
+      "step": 1717
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 5e-05,
+      "loss": 0.2305,
+      "step": 1718
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2415,
+      "step": 1719
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5e-05,
+      "loss": 0.1956,
+      "step": 1720
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2126,
+      "step": 1721
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2299,
+      "step": 1722
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5e-05,
+      "loss": 0.2324,
+      "step": 1723
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 5e-05,
+      "loss": 0.1958,
+      "step": 1724
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2297,
+      "step": 1725
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2034,
+      "step": 1726
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2522,
+      "step": 1727
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2969,
+      "step": 1728
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 5e-05,
+      "loss": 0.2167,
+      "step": 1729
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2537,
+      "step": 1730
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5e-05,
+      "loss": 0.22,
+      "step": 1731
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5e-05,
+      "loss": 0.1567,
+      "step": 1732
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5e-05,
+      "loss": 0.2407,
+      "step": 1733
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5e-05,
+      "loss": 0.1805,
+      "step": 1734
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 5e-05,
+      "loss": 0.1775,
+      "step": 1735
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2004,
+      "step": 1736
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2229,
+      "step": 1737
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2029,
+      "step": 1738
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2712,
+      "step": 1739
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 5e-05,
+      "loss": 0.2336,
+      "step": 1740
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 5e-05,
+      "loss": 0.304,
+      "step": 1741
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 5e-05,
+      "loss": 0.1453,
+      "step": 1742
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2058,
+      "step": 1743
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2065,
+      "step": 1744
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2534,
+      "step": 1745
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 5e-05,
+      "loss": 0.2031,
+      "step": 1746
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 5e-05,
+      "loss": 0.199,
+      "step": 1747
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2249,
+      "step": 1748
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 5e-05,
+      "loss": 0.1797,
+      "step": 1749
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 5e-05,
+      "loss": 0.2236,
+      "step": 1750
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 5e-05,
+      "loss": 0.196,
+      "step": 1751
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 5e-05,
+      "loss": 0.1738,
+      "step": 1752
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 5e-05,
+      "loss": 0.1853,
+      "step": 1753
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 5e-05,
+      "loss": 0.1926,
+      "step": 1754
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2458,
+      "step": 1755
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 5e-05,
+      "loss": 0.1412,
+      "step": 1756
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 5e-05,
+      "loss": 0.2705,
+      "step": 1757
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2288,
+      "step": 1758
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2476,
+      "step": 1759
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2239,
+      "step": 1760
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 5e-05,
+      "loss": 0.1721,
+      "step": 1761
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2527,
+      "step": 1762
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 5e-05,
+      "loss": 0.2361,
+      "step": 1763
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5e-05,
+      "loss": 0.3032,
+      "step": 1764
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2256,
+      "step": 1765
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5e-05,
+      "loss": 0.1885,
+      "step": 1766
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5e-05,
+      "loss": 0.271,
+      "step": 1767
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5e-05,
+      "loss": 0.2788,
+      "step": 1768
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 5e-05,
+      "loss": 0.1837,
+      "step": 1769
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2393,
+      "step": 1770
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2195,
+      "step": 1771
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2805,
+      "step": 1772
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2771,
+      "step": 1773
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 5e-05,
+      "loss": 0.2922,
+      "step": 1774
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2261,
+      "step": 1775
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 5e-05,
+      "loss": 0.1963,
+      "step": 1776
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 5e-05,
+      "loss": 0.3347,
+      "step": 1777
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 5e-05,
+      "loss": 0.1498,
+      "step": 1778
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 5e-05,
+      "loss": 0.3308,
+      "step": 1779
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 5e-05,
+      "loss": 0.2756,
+      "step": 1780
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2405,
+      "step": 1781
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 5e-05,
+      "loss": 0.1858,
+      "step": 1782
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2083,
+      "step": 1783
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 5e-05,
+      "loss": 0.2502,
+      "step": 1784
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 5e-05,
+      "loss": 0.1987,
+      "step": 1785
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2417,
+      "step": 1786
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5e-05,
+      "loss": 0.325,
+      "step": 1787
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2646,
+      "step": 1788
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2306,
+      "step": 1789
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2148,
+      "step": 1790
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 5e-05,
+      "loss": 0.2319,
+      "step": 1791
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 5e-05,
+      "loss": 0.261,
+      "step": 1792
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 5e-05,
+      "loss": 0.197,
+      "step": 1793
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2029,
+      "step": 1794
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 5e-05,
+      "loss": 0.2966,
+      "step": 1795
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 5e-05,
+      "loss": 0.5317,
+      "step": 1796
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 5e-05,
+      "loss": 0.1909,
+      "step": 1797
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 5e-05,
+      "loss": 0.228,
+      "step": 1798
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 5e-05,
+      "loss": 0.1952,
+      "step": 1799
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 5e-05,
+      "loss": 0.2327,
+      "step": 1800
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 5e-05,
+      "loss": 0.3196,
+      "step": 1801
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 5e-05,
+      "loss": 0.6997,
+      "step": 1802
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2319,
+      "step": 1803
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2705,
+      "step": 1804
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2048,
+      "step": 1805
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2057,
+      "step": 1806
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5e-05,
+      "loss": 0.1721,
+      "step": 1807
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5e-05,
+      "loss": 0.2495,
+      "step": 1808
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2178,
+      "step": 1809
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5e-05,
+      "loss": 0.26,
+      "step": 1810
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2275,
+      "step": 1811
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5e-05,
+      "loss": 0.3105,
+      "step": 1812
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2937,
+      "step": 1813
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5e-05,
+      "loss": 0.2729,
+      "step": 1814
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2427,
+      "step": 1815
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5e-05,
+      "loss": 0.1873,
+      "step": 1816
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5e-05,
+      "loss": 0.1567,
+      "step": 1817
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5e-05,
+      "loss": 0.1774,
+      "step": 1818
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5e-05,
+      "loss": 0.2634,
+      "step": 1819
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2228,
+      "step": 1820
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5e-05,
+      "loss": 0.1777,
+      "step": 1821
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2185,
+      "step": 1822
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5e-05,
+      "loss": 0.1564,
+      "step": 1823
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2217,
+      "step": 1824
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5e-05,
+      "loss": 0.2441,
+      "step": 1825
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5e-05,
+      "loss": 0.1925,
+      "step": 1826
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2383,
+      "step": 1827
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2031,
+      "step": 1828
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2615,
+      "step": 1829
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5e-05,
+      "loss": 0.2686,
+      "step": 1830
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5e-05,
+      "loss": 0.1357,
+      "step": 1831
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5e-05,
+      "loss": 0.1887,
+      "step": 1832
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2251,
+      "step": 1833
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5e-05,
+      "loss": 0.2488,
+      "step": 1834
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5e-05,
+      "loss": 0.1707,
+      "step": 1835
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5e-05,
+      "loss": 0.1957,
+      "step": 1836
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2019,
+      "step": 1837
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5e-05,
+      "loss": 0.1809,
+      "step": 1838
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2124,
+      "step": 1839
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5e-05,
+      "loss": 0.238,
+      "step": 1840
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5e-05,
+      "loss": 0.3088,
+      "step": 1841
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5e-05,
+      "loss": 0.2563,
+      "step": 1842
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2937,
+      "step": 1843
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 5e-05,
+      "loss": 0.229,
+      "step": 1844
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 5e-05,
+      "loss": 0.2,
+      "step": 1845
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 5e-05,
+      "loss": 0.1675,
+      "step": 1846
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 5e-05,
+      "loss": 0.1759,
+      "step": 1847
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5e-05,
+      "loss": 0.1682,
+      "step": 1848
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2273,
+      "step": 1849
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5e-05,
+      "loss": 0.1971,
+      "step": 1850
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2363,
+      "step": 1851
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5e-05,
+      "loss": 0.2805,
+      "step": 1852
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5e-05,
+      "loss": 0.3372,
+      "step": 1853
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5e-05,
+      "loss": 0.166,
+      "step": 1854
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2778,
+      "step": 1855
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5e-05,
+      "loss": 0.215,
+      "step": 1856
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2495,
+      "step": 1857
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2629,
+      "step": 1858
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5e-05,
+      "loss": 0.2681,
+      "step": 1859
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5e-05,
+      "loss": 0.1851,
+      "step": 1860
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2086,
+      "step": 1861
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2378,
+      "step": 1862
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5e-05,
+      "loss": 0.1849,
+      "step": 1863
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5e-05,
+      "loss": 0.2001,
+      "step": 1864
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5e-05,
+      "loss": 0.3037,
+      "step": 1865
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5e-05,
+      "loss": 0.1829,
+      "step": 1866
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2296,
+      "step": 1867
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5e-05,
+      "loss": 0.2646,
+      "step": 1868
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5e-05,
+      "loss": 0.21,
+      "step": 1869
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5e-05,
+      "loss": 0.1831,
+      "step": 1870
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5e-05,
+      "loss": 0.3352,
+      "step": 1871
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2993,
+      "step": 1872
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5e-05,
+      "loss": 0.1306,
+      "step": 1873
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2358,
+      "step": 1874
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5e-05,
+      "loss": 0.2456,
+      "step": 1875
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5e-05,
+      "loss": 0.272,
+      "step": 1876
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2136,
+      "step": 1877
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2334,
+      "step": 1878
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2319,
+      "step": 1879
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5e-05,
+      "loss": 0.1572,
+      "step": 1880
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5e-05,
+      "loss": 0.2726,
+      "step": 1881
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2336,
+      "step": 1882
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5e-05,
+      "loss": 0.226,
+      "step": 1883
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2368,
+      "step": 1884
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5e-05,
+      "loss": 0.2129,
+      "step": 1885
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5e-05,
+      "loss": 0.3213,
+      "step": 1886
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5e-05,
+      "loss": 0.179,
+      "step": 1887
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5e-05,
+      "loss": 0.1337,
+      "step": 1888
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2375,
+      "step": 1889
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5e-05,
+      "loss": 0.1455,
+      "step": 1890
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2849,
+      "step": 1891
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5e-05,
+      "loss": 0.2495,
+      "step": 1892
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5e-05,
+      "loss": 0.241,
+      "step": 1893
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5e-05,
+      "loss": 0.272,
+      "step": 1894
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5e-05,
+      "loss": 0.249,
+      "step": 1895
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5e-05,
+      "loss": 0.1494,
+      "step": 1896
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5e-05,
+      "loss": 0.191,
+      "step": 1897
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5e-05,
+      "loss": 0.2131,
+      "step": 1898
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5e-05,
+      "loss": 0.1887,
+      "step": 1899
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2004,
+      "step": 1900
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2239,
+      "step": 1901
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5e-05,
+      "loss": 0.343,
+      "step": 1902
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2542,
+      "step": 1903
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5e-05,
+      "loss": 0.2512,
+      "step": 1904
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 5e-05,
+      "loss": 0.1362,
+      "step": 1905
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 5e-05,
+      "loss": 0.219,
+      "step": 1906
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2664,
+      "step": 1907
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2197,
+      "step": 1908
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 5e-05,
+      "loss": 0.2727,
+      "step": 1909
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5e-05,
+      "loss": 0.3154,
+      "step": 1910
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2644,
+      "step": 1911
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5e-05,
+      "loss": 0.1865,
+      "step": 1912
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5e-05,
+      "loss": 0.1949,
+      "step": 1913
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2,
+      "step": 1914
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 5e-05,
+      "loss": 0.2371,
+      "step": 1915
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5e-05,
+      "loss": 0.1798,
+      "step": 1916
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2236,
+      "step": 1917
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5e-05,
+      "loss": 0.2113,
+      "step": 1918
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5e-05,
+      "loss": 0.197,
+      "step": 1919
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 5e-05,
+      "loss": 0.1538,
+      "step": 1920
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2095,
+      "step": 1921
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2344,
+      "step": 1922
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2485,
+      "step": 1923
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5e-05,
+      "loss": 0.1633,
+      "step": 1924
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2698,
+      "step": 1925
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 5e-05,
+      "loss": 0.2666,
+      "step": 1926
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5e-05,
+      "loss": 0.269,
+      "step": 1927
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5e-05,
+      "loss": 0.26,
+      "step": 1928
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2415,
+      "step": 1929
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5e-05,
+      "loss": 0.1685,
+      "step": 1930
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2288,
+      "step": 1931
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 5e-05,
+      "loss": 0.2185,
+      "step": 1932
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2439,
+      "step": 1933
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2498,
+      "step": 1934
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5e-05,
+      "loss": 0.3018,
+      "step": 1935
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5e-05,
+      "loss": 0.2393,
+      "step": 1936
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 5e-05,
+      "loss": 0.1892,
+      "step": 1937
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2003,
+      "step": 1938
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5e-05,
+      "loss": 0.25,
+      "step": 1939
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2688,
+      "step": 1940
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5e-05,
+      "loss": 0.1429,
+      "step": 1941
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5e-05,
+      "loss": 0.1934,
+      "step": 1942
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 5e-05,
+      "loss": 0.2009,
+      "step": 1943
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5e-05,
+      "loss": 0.1649,
+      "step": 1944
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5e-05,
+      "loss": 0.365,
+      "step": 1945
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5e-05,
+      "loss": 0.1812,
+      "step": 1946
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5e-05,
+      "loss": 0.1819,
+      "step": 1947
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2627,
+      "step": 1948
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 5e-05,
+      "loss": 0.2561,
+      "step": 1949
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2615,
+      "step": 1950
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2659,
+      "step": 1951
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2596,
+      "step": 1952
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5e-05,
+      "loss": 0.1823,
+      "step": 1953
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 5e-05,
+      "loss": 0.2349,
+      "step": 1954
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5e-05,
+      "loss": 0.1897,
+      "step": 1955
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2029,
+      "step": 1956
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5e-05,
+      "loss": 0.1729,
+      "step": 1957
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5e-05,
+      "loss": 0.3042,
+      "step": 1958
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2114,
+      "step": 1959
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 5e-05,
+      "loss": 0.2234,
+      "step": 1960
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5e-05,
+      "loss": 0.3062,
+      "step": 1961
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2041,
+      "step": 1962
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5e-05,
+      "loss": 0.1473,
+      "step": 1963
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5e-05,
+      "loss": 0.241,
+      "step": 1964
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 5e-05,
+      "loss": 0.2979,
+      "step": 1965
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2321,
+      "step": 1966
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2151,
+      "step": 1967
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5e-05,
+      "loss": 0.1527,
+      "step": 1968
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2451,
+      "step": 1969
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5e-05,
+      "loss": 0.1779,
+      "step": 1970
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 5e-05,
+      "loss": 0.2,
+      "step": 1971
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2092,
+      "step": 1972
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2825,
+      "step": 1973
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2708,
+      "step": 1974
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2271,
+      "step": 1975
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5e-05,
+      "loss": 0.2612,
+      "step": 1976
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 5e-05,
+      "loss": 0.262,
+      "step": 1977
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2766,
+      "step": 1978
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5e-05,
+      "loss": 0.3552,
+      "step": 1979
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5e-05,
+      "loss": 0.207,
+      "step": 1980
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2114,
+      "step": 1981
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 5e-05,
+      "loss": 0.2317,
+      "step": 1982
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5e-05,
+      "loss": 0.1934,
+      "step": 1983
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5e-05,
+      "loss": 0.1771,
+      "step": 1984
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5e-05,
+      "loss": 0.2075,
+      "step": 1985
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5e-05,
+      "loss": 0.3228,
+      "step": 1986
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5e-05,
+      "loss": 0.1948,
+      "step": 1987
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 5e-05,
+      "loss": 0.1973,
+      "step": 1988
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5e-05,
+      "loss": 0.1858,
+      "step": 1989
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2236,
+      "step": 1990
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2087,
+      "step": 1991
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5e-05,
+      "loss": 0.1704,
+      "step": 1992
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5e-05,
+      "loss": 0.2866,
+      "step": 1993
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 5e-05,
+      "loss": 0.209,
+      "step": 1994
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2434,
+      "step": 1995
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2351,
+      "step": 1996
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5e-05,
+      "loss": 0.2219,
+      "step": 1997
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5e-05,
+      "loss": 0.1953,
+      "step": 1998
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 5e-05,
+      "loss": 0.1956,
+      "step": 1999
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2981,
+      "step": 2000
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5e-05,
+      "loss": 0.24,
+      "step": 2001
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5e-05,
+      "loss": 0.252,
+      "step": 2002
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2588,
+      "step": 2003
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5e-05,
+      "loss": 0.1719,
+      "step": 2004
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 5e-05,
+      "loss": 0.2207,
+      "step": 2005
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5e-05,
+      "loss": 0.1918,
+      "step": 2006
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2291,
+      "step": 2007
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2576,
+      "step": 2008
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2441,
+      "step": 2009
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 5e-05,
+      "loss": 0.2158,
+      "step": 2010
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2866,
+      "step": 2011
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5e-05,
+      "loss": 0.1753,
+      "step": 2012
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2529,
+      "step": 2013
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5e-05,
+      "loss": 0.1714,
+      "step": 2014
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2485,
+      "step": 2015
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5e-05,
+      "loss": 0.2275,
+      "step": 2016
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5e-05,
+      "loss": 0.189,
+      "step": 2017
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5e-05,
+      "loss": 0.1897,
+      "step": 2018
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2092,
+      "step": 2019
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5e-05,
+      "loss": 0.2125,
+      "step": 2020
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5e-05,
+      "loss": 0.197,
+      "step": 2021
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5e-05,
+      "loss": 0.1799,
+      "step": 2022
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2585,
+      "step": 2023
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2856,
+      "step": 2024
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2693,
+      "step": 2025
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5e-05,
+      "loss": 0.1794,
+      "step": 2026
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5e-05,
+      "loss": 0.2288,
+      "step": 2027
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2676,
+      "step": 2028
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5e-05,
+      "loss": 0.1793,
+      "step": 2029
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2256,
+      "step": 2030
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5e-05,
+      "loss": 0.1758,
+      "step": 2031
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2156,
+      "step": 2032
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5e-05,
+      "loss": 0.2134,
+      "step": 2033
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2251,
+      "step": 2034
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5e-05,
+      "loss": 0.2332,
+      "step": 2035
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5e-05,
+      "loss": 0.1841,
+      "step": 2036
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5e-05,
+      "loss": 0.1514,
+      "step": 2037
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5e-05,
+      "loss": 0.178,
+      "step": 2038
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 5e-05,
+      "loss": 0.229,
+      "step": 2039
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2854,
+      "step": 2040
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2559,
+      "step": 2041
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5e-05,
+      "loss": 0.1663,
+      "step": 2042
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5e-05,
+      "loss": 0.1741,
+      "step": 2043
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 5e-05,
+      "loss": 0.2378,
+      "step": 2044
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5e-05,
+      "loss": 0.1619,
+      "step": 2045
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5e-05,
+      "loss": 0.203,
+      "step": 2046
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5e-05,
+      "loss": 0.267,
+      "step": 2047
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5e-05,
+      "loss": 0.2759,
+      "step": 2048
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5e-05,
+      "loss": 0.208,
+      "step": 2049
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 5e-05,
+      "loss": 0.1846,
+      "step": 2050
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5e-05,
+      "loss": 0.4829,
+      "step": 2051
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5e-05,
+      "loss": 0.199,
+      "step": 2052
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2297,
+      "step": 2053
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2512,
+      "step": 2054
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 5e-05,
+      "loss": 0.2168,
+      "step": 2055
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2693,
+      "step": 2056
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2202,
+      "step": 2057
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2739,
+      "step": 2058
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2,
+      "step": 2059
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5e-05,
+      "loss": 0.3423,
+      "step": 2060
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 5e-05,
+      "loss": 0.2156,
+      "step": 2061
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2041,
+      "step": 2062
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5e-05,
+      "loss": 0.183,
+      "step": 2063
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5e-05,
+      "loss": 0.1858,
+      "step": 2064
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5e-05,
+      "loss": 0.2594,
+      "step": 2065
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5e-05,
+      "loss": 0.272,
+      "step": 2066
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 5e-05,
+      "loss": 0.272,
+      "step": 2067
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2504,
+      "step": 2068
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5e-05,
+      "loss": 0.2561,
+      "step": 2069
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5e-05,
+      "loss": 0.3066,
+      "step": 2070
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5e-05,
+      "loss": 0.1776,
+      "step": 2071
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 5e-05,
+      "loss": 0.1868,
+      "step": 2072
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5e-05,
+      "loss": 0.1702,
+      "step": 2073
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5e-05,
+      "loss": 0.1611,
+      "step": 2074
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2579,
+      "step": 2075
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5e-05,
+      "loss": 0.261,
+      "step": 2076
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5e-05,
+      "loss": 0.2183,
+      "step": 2077
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 5e-05,
+      "loss": 0.209,
+      "step": 2078
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2241,
+      "step": 2079
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5e-05,
+      "loss": 0.1837,
+      "step": 2080
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2502,
+      "step": 2081
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5e-05,
+      "loss": 0.1758,
+      "step": 2082
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5e-05,
+      "loss": 0.1347,
+      "step": 2083
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 5e-05,
+      "loss": 0.2212,
+      "step": 2084
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5e-05,
+      "loss": 0.1765,
+      "step": 2085
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2222,
+      "step": 2086
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2312,
+      "step": 2087
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5e-05,
+      "loss": 0.207,
+      "step": 2088
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 5e-05,
+      "loss": 0.2471,
+      "step": 2089
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5e-05,
+      "loss": 0.1997,
+      "step": 2090
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5e-05,
+      "loss": 0.1838,
+      "step": 2091
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2947,
+      "step": 2092
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2034,
+      "step": 2093
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5e-05,
+      "loss": 0.1827,
+      "step": 2094
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 5e-05,
+      "loss": 0.2151,
+      "step": 2095
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5e-05,
+      "loss": 0.1926,
+      "step": 2096
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5e-05,
+      "loss": 0.1948,
+      "step": 2097
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5e-05,
+      "loss": 0.5336,
+      "step": 2098
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2296,
+      "step": 2099
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 5e-05,
+      "loss": 0.2052,
+      "step": 2100
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2416,
+      "step": 2101
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5e-05,
+      "loss": 0.3159,
+      "step": 2102
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2163,
+      "step": 2103
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5e-05,
+      "loss": 0.1887,
+      "step": 2104
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2708,
+      "step": 2105
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5e-05,
+      "loss": 0.2029,
+      "step": 2106
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2419,
+      "step": 2107
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2937,
+      "step": 2108
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2927,
+      "step": 2109
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2246,
+      "step": 2110
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5e-05,
+      "loss": 0.2175,
+      "step": 2111
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5e-05,
+      "loss": 0.1527,
+      "step": 2112
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2554,
+      "step": 2113
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5e-05,
+      "loss": 0.226,
+      "step": 2114
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2388,
+      "step": 2115
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2607,
+      "step": 2116
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5e-05,
+      "loss": 0.2422,
+      "step": 2117
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2739,
+      "step": 2118
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2151,
+      "step": 2119
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2266,
+      "step": 2120
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5e-05,
+      "loss": 0.4092,
+      "step": 2121
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5e-05,
+      "loss": 0.2465,
+      "step": 2122
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5e-05,
+      "loss": 0.3003,
+      "step": 2123
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2046,
+      "step": 2124
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5e-05,
+      "loss": 0.1992,
+      "step": 2125
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2336,
+      "step": 2126
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5e-05,
+      "loss": 0.1521,
+      "step": 2127
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5e-05,
+      "loss": 0.2166,
+      "step": 2128
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5e-05,
+      "loss": 0.1824,
+      "step": 2129
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5e-05,
+      "loss": 0.1678,
+      "step": 2130
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2302,
+      "step": 2131
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5e-05,
+      "loss": 0.1843,
+      "step": 2132
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5e-05,
+      "loss": 0.2285,
+      "step": 2133
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5e-05,
+      "loss": 0.3281,
+      "step": 2134
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2939,
+      "step": 2135
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2095,
+      "step": 2136
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5e-05,
+      "loss": 0.1667,
+      "step": 2137
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5e-05,
+      "loss": 0.2104,
+      "step": 2138
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5e-05,
+      "loss": 0.1582,
+      "step": 2139
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5e-05,
+      "loss": 0.225,
+      "step": 2140
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5e-05,
+      "loss": 0.228,
+      "step": 2141
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5e-05,
+      "loss": 0.1869,
+      "step": 2142
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5e-05,
+      "loss": 0.2395,
+      "step": 2143
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5e-05,
+      "loss": 0.1793,
+      "step": 2144
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5e-05,
+      "loss": 0.1825,
+      "step": 2145
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5e-05,
+      "loss": 0.208,
+      "step": 2146
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2322,
+      "step": 2147
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2056,
+      "step": 2148
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2388,
+      "step": 2149
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5e-05,
+      "loss": 0.175,
+      "step": 2150
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5e-05,
+      "loss": 0.2878,
+      "step": 2151
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5e-05,
+      "loss": 0.1675,
+      "step": 2152
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5e-05,
+      "loss": 0.1992,
+      "step": 2153
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5e-05,
+      "loss": 0.1635,
+      "step": 2154
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2456,
+      "step": 2155
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5e-05,
+      "loss": 0.1495,
+      "step": 2156
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5e-05,
+      "loss": 0.2561,
+      "step": 2157
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5e-05,
+      "loss": 0.1604,
+      "step": 2158
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5e-05,
+      "loss": 0.1882,
+      "step": 2159
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5e-05,
+      "loss": 0.219,
+      "step": 2160
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2607,
+      "step": 2161
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5e-05,
+      "loss": 0.2217,
+      "step": 2162
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2708,
+      "step": 2163
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2026,
+      "step": 2164
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2388,
+      "step": 2165
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2437,
+      "step": 2166
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2053,
+      "step": 2167
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5e-05,
+      "loss": 0.2554,
+      "step": 2168
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2209,
+      "step": 2169
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2129,
+      "step": 2170
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5e-05,
+      "loss": 0.1985,
+      "step": 2171
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5e-05,
+      "loss": 0.25,
+      "step": 2172
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5e-05,
+      "loss": 0.251,
+      "step": 2173
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5e-05,
+      "loss": 0.2888,
+      "step": 2174
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2285,
+      "step": 2175
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2751,
+      "step": 2176
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5e-05,
+      "loss": 0.2271,
+      "step": 2177
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5e-05,
+      "loss": 0.27,
+      "step": 2178
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5e-05,
+      "loss": 0.3208,
+      "step": 2179
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2166,
+      "step": 2180
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5e-05,
+      "loss": 0.3369,
+      "step": 2181
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2773,
+      "step": 2182
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5e-05,
+      "loss": 0.229,
+      "step": 2183
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5e-05,
+      "loss": 0.2383,
+      "step": 2184
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5e-05,
+      "loss": 0.282,
+      "step": 2185
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5e-05,
+      "loss": 0.1813,
+      "step": 2186
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2046,
+      "step": 2187
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5e-05,
+      "loss": 0.2935,
+      "step": 2188
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5e-05,
+      "loss": 0.3289,
+      "step": 2189
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5e-05,
+      "loss": 0.3652,
+      "step": 2190
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5e-05,
+      "loss": 0.1458,
+      "step": 2191
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5e-05,
+      "loss": 0.1775,
+      "step": 2192
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2925,
+      "step": 2193
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5e-05,
+      "loss": 0.2747,
+      "step": 2194
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5e-05,
+      "loss": 0.302,
+      "step": 2195
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5e-05,
+      "loss": 0.1902,
+      "step": 2196
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5e-05,
+      "loss": 0.1963,
+      "step": 2197
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2432,
+      "step": 2198
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5e-05,
+      "loss": 0.3765,
+      "step": 2199
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2063,
+      "step": 2200
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2327,
+      "step": 2201
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5e-05,
+      "loss": 0.2141,
+      "step": 2202
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2451,
+      "step": 2203
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5e-05,
+      "loss": 0.1616,
+      "step": 2204
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2026,
+      "step": 2205
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2458,
+      "step": 2206
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5e-05,
+      "loss": 0.2039,
+      "step": 2207
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5e-05,
+      "loss": 0.1921,
+      "step": 2208
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2512,
+      "step": 2209
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5e-05,
+      "loss": 0.1941,
+      "step": 2210
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5e-05,
+      "loss": 0.2539,
+      "step": 2211
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5e-05,
+      "loss": 0.3235,
+      "step": 2212
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5e-05,
+      "loss": 0.1664,
+      "step": 2213
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5e-05,
+      "loss": 0.217,
+      "step": 2214
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5e-05,
+      "loss": 0.3011,
+      "step": 2215
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5e-05,
+      "loss": 0.1875,
+      "step": 2216
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2734,
+      "step": 2217
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2668,
+      "step": 2218
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5e-05,
+      "loss": 0.2253,
+      "step": 2219
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2283,
+      "step": 2220
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2329,
+      "step": 2221
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2957,
+      "step": 2222
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5e-05,
+      "loss": 0.2039,
+      "step": 2223
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5e-05,
+      "loss": 0.1799,
+      "step": 2224
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5e-05,
+      "loss": 0.3074,
+      "step": 2225
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5e-05,
+      "loss": 0.259,
+      "step": 2226
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2717,
+      "step": 2227
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2618,
+      "step": 2228
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2693,
+      "step": 2229
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5e-05,
+      "loss": 0.2947,
+      "step": 2230
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2405,
+      "step": 2231
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5e-05,
+      "loss": 0.1953,
+      "step": 2232
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5e-05,
+      "loss": 0.2544,
+      "step": 2233
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5e-05,
+      "loss": 0.1906,
+      "step": 2234
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5e-05,
+      "loss": 0.374,
+      "step": 2235
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5e-05,
+      "loss": 0.1978,
+      "step": 2236
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5e-05,
+      "loss": 0.1866,
+      "step": 2237
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2744,
+      "step": 2238
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2295,
+      "step": 2239
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2104,
+      "step": 2240
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5e-05,
+      "loss": 0.2394,
+      "step": 2241
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 5e-05,
+      "loss": 0.219,
+      "step": 2242
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 5e-05,
+      "loss": 0.1956,
+      "step": 2243
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2273,
+      "step": 2244
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2122,
+      "step": 2245
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2156,
+      "step": 2246
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 5e-05,
+      "loss": 0.2461,
+      "step": 2247
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 5e-05,
+      "loss": 0.2009,
+      "step": 2248
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 0.22490152716636658,
+      "eval_runtime": 118.6312,
+      "eval_samples_per_second": 2.529,
+      "eval_steps_per_second": 0.632,
+      "step": 2248
+    },
+    {
+      "epoch": 4.0,
+      "step": 2248,
+      "total_flos": 3.506559652919771e+19,
+      "train_loss": 0.2778411539423932,
+      "train_runtime": 160981.3537,
+      "train_samples_per_second": 0.671,
+      "train_steps_per_second": 0.014
+    }
+  ],
+  "max_steps": 2248,
+  "num_train_epochs": 4,
+  "total_flos": 3.506559652919771e+19,
+  "trial_name": null,
+  "trial_params": null
+}