diff --git "a/checkpoint-1572/trainer_state.json" "b/checkpoint-1572/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1572/trainer_state.json"
@@ -0,0 +1,9453 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.991113932085052,
+  "eval_steps": 500,
+  "global_step": 1572,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 1.3058,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 1.192,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.25e-06,
+      "loss": 1.2973,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.2629,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 1.2073,
+      "step": 5
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5e-06,
+      "loss": 1.1229,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 1.2677,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.1353,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.0524,
+      "step": 9
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 1.0016,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 1.1865,
+      "step": 11
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-06,
+      "loss": 1.0415,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 0.9764,
+      "step": 13
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 1.0045,
+      "step": 14
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6.25e-06,
+      "loss": 0.8317,
+      "step": 15
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.9207,
+      "step": 16
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.083333333333335e-06,
+      "loss": 0.8949,
+      "step": 17
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.8629,
+      "step": 18
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 0.9026,
+      "step": 19
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.8833,
+      "step": 20
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.8436,
+      "step": 21
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.868,
+      "step": 22
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.583333333333335e-06,
+      "loss": 0.8943,
+      "step": 23
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.835,
+      "step": 24
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.0416666666666668e-05,
+      "loss": 0.8719,
+      "step": 25
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.0833333333333334e-05,
+      "loss": 0.8289,
+      "step": 26
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.125e-05,
+      "loss": 0.8167,
+      "step": 27
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 0.8897,
+      "step": 28
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.2083333333333333e-05,
+      "loss": 0.895,
+      "step": 29
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.25e-05,
+      "loss": 0.8243,
+      "step": 30
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.2916666666666668e-05,
+      "loss": 0.8314,
+      "step": 31
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8049,
+      "step": 32
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.375e-05,
+      "loss": 0.8087,
+      "step": 33
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.416666666666667e-05,
+      "loss": 0.7565,
+      "step": 34
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.4583333333333333e-05,
+      "loss": 0.8179,
+      "step": 35
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.8248,
+      "step": 36
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.5416666666666668e-05,
+      "loss": 0.8214,
+      "step": 37
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.5833333333333333e-05,
+      "loss": 0.8612,
+      "step": 38
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 0.8087,
+      "step": 39
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.8338,
+      "step": 40
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.7083333333333333e-05,
+      "loss": 0.8538,
+      "step": 41
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 0.736,
+      "step": 42
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.7916666666666667e-05,
+      "loss": 0.7752,
+      "step": 43
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.8186,
+      "step": 44
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.9004,
+      "step": 45
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.916666666666667e-05,
+      "loss": 0.7464,
+      "step": 46
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9583333333333333e-05,
+      "loss": 0.8744,
+      "step": 47
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 0.7909,
+      "step": 48
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.998687664041995e-05,
+      "loss": 0.7354,
+      "step": 49
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9973753280839896e-05,
+      "loss": 0.7691,
+      "step": 50
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9960629921259843e-05,
+      "loss": 0.7398,
+      "step": 51
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9947506561679793e-05,
+      "loss": 0.814,
+      "step": 52
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9934383202099737e-05,
+      "loss": 0.8055,
+      "step": 53
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9921259842519688e-05,
+      "loss": 0.785,
+      "step": 54
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9908136482939635e-05,
+      "loss": 1.0027,
+      "step": 55
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9895013123359582e-05,
+      "loss": 0.8046,
+      "step": 56
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.988188976377953e-05,
+      "loss": 0.6848,
+      "step": 57
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9868766404199476e-05,
+      "loss": 0.786,
+      "step": 58
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9855643044619423e-05,
+      "loss": 0.8834,
+      "step": 59
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.984251968503937e-05,
+      "loss": 0.752,
+      "step": 60
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.982939632545932e-05,
+      "loss": 0.6875,
+      "step": 61
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9816272965879265e-05,
+      "loss": 0.7118,
+      "step": 62
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9803149606299215e-05,
+      "loss": 0.6849,
+      "step": 63
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9790026246719162e-05,
+      "loss": 0.8545,
+      "step": 64
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.977690288713911e-05,
+      "loss": 0.6823,
+      "step": 65
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9763779527559057e-05,
+      "loss": 0.7772,
+      "step": 66
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9750656167979004e-05,
+      "loss": 0.789,
+      "step": 67
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.973753280839895e-05,
+      "loss": 0.7375,
+      "step": 68
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.97244094488189e-05,
+      "loss": 0.8808,
+      "step": 69
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9711286089238845e-05,
+      "loss": 0.7628,
+      "step": 70
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9698162729658795e-05,
+      "loss": 0.732,
+      "step": 71
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9685039370078743e-05,
+      "loss": 0.7512,
+      "step": 72
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.967191601049869e-05,
+      "loss": 0.7593,
+      "step": 73
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9658792650918637e-05,
+      "loss": 0.7135,
+      "step": 74
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9645669291338584e-05,
+      "loss": 0.8312,
+      "step": 75
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.963254593175853e-05,
+      "loss": 0.7569,
+      "step": 76
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9619422572178478e-05,
+      "loss": 0.7697,
+      "step": 77
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.960629921259843e-05,
+      "loss": 0.7605,
+      "step": 78
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9593175853018372e-05,
+      "loss": 0.6993,
+      "step": 79
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9580052493438323e-05,
+      "loss": 0.7379,
+      "step": 80
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.956692913385827e-05,
+      "loss": 0.7636,
+      "step": 81
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9553805774278217e-05,
+      "loss": 0.7483,
+      "step": 82
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9540682414698164e-05,
+      "loss": 0.6927,
+      "step": 83
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.952755905511811e-05,
+      "loss": 0.7517,
+      "step": 84
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.951443569553806e-05,
+      "loss": 0.7909,
+      "step": 85
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9501312335958006e-05,
+      "loss": 0.8087,
+      "step": 86
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9488188976377956e-05,
+      "loss": 0.6628,
+      "step": 87
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.94750656167979e-05,
+      "loss": 0.7129,
+      "step": 88
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.946194225721785e-05,
+      "loss": 0.6922,
+      "step": 89
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9448818897637797e-05,
+      "loss": 0.7997,
+      "step": 90
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9435695538057745e-05,
+      "loss": 0.6905,
+      "step": 91
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9422572178477692e-05,
+      "loss": 0.8932,
+      "step": 92
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.940944881889764e-05,
+      "loss": 0.7211,
+      "step": 93
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9396325459317586e-05,
+      "loss": 0.7291,
+      "step": 94
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9383202099737536e-05,
+      "loss": 0.7578,
+      "step": 95
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.937007874015748e-05,
+      "loss": 0.8398,
+      "step": 96
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.935695538057743e-05,
+      "loss": 0.7768,
+      "step": 97
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9343832020997378e-05,
+      "loss": 0.8122,
+      "step": 98
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9330708661417325e-05,
+      "loss": 0.7434,
+      "step": 99
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9317585301837272e-05,
+      "loss": 0.7193,
+      "step": 100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.930446194225722e-05,
+      "loss": 0.7587,
+      "step": 101
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9291338582677166e-05,
+      "loss": 0.7144,
+      "step": 102
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9278215223097113e-05,
+      "loss": 0.629,
+      "step": 103
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9265091863517064e-05,
+      "loss": 0.7723,
+      "step": 104
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9251968503937008e-05,
+      "loss": 0.7523,
+      "step": 105
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9238845144356958e-05,
+      "loss": 0.7475,
+      "step": 106
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9225721784776905e-05,
+      "loss": 0.789,
+      "step": 107
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9212598425196852e-05,
+      "loss": 0.7289,
+      "step": 108
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.91994750656168e-05,
+      "loss": 0.7498,
+      "step": 109
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9186351706036747e-05,
+      "loss": 0.9547,
+      "step": 110
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9173228346456694e-05,
+      "loss": 0.8021,
+      "step": 111
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.916010498687664e-05,
+      "loss": 0.7983,
+      "step": 112
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.914698162729659e-05,
+      "loss": 0.6735,
+      "step": 113
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9133858267716535e-05,
+      "loss": 0.6658,
+      "step": 114
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9120734908136486e-05,
+      "loss": 0.7631,
+      "step": 115
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9107611548556433e-05,
+      "loss": 0.7936,
+      "step": 116
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.909448818897638e-05,
+      "loss": 0.7189,
+      "step": 117
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9081364829396327e-05,
+      "loss": 0.7468,
+      "step": 118
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9068241469816274e-05,
+      "loss": 0.6605,
+      "step": 119
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.905511811023622e-05,
+      "loss": 0.7652,
+      "step": 120
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9041994750656168e-05,
+      "loss": 0.7574,
+      "step": 121
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.902887139107612e-05,
+      "loss": 0.7762,
+      "step": 122
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9015748031496062e-05,
+      "loss": 0.8641,
+      "step": 123
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9002624671916013e-05,
+      "loss": 0.8135,
+      "step": 124
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.898950131233596e-05,
+      "loss": 0.7567,
+      "step": 125
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.8976377952755907e-05,
+      "loss": 0.7335,
+      "step": 126
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.8963254593175854e-05,
+      "loss": 0.7627,
+      "step": 127
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.89501312335958e-05,
+      "loss": 0.6773,
+      "step": 128
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.893700787401575e-05,
+      "loss": 0.729,
+      "step": 129
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.89238845144357e-05,
+      "loss": 0.7523,
+      "step": 130
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.8910761154855643e-05,
+      "loss": 0.8078,
+      "step": 131
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.8897637795275593e-05,
+      "loss": 0.6969,
+      "step": 132
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.888451443569554e-05,
+      "loss": 0.6729,
+      "step": 133
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.8871391076115488e-05,
+      "loss": 0.6691,
+      "step": 134
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.8858267716535435e-05,
+      "loss": 0.7116,
+      "step": 135
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.8845144356955382e-05,
+      "loss": 0.7435,
+      "step": 136
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.883202099737533e-05,
+      "loss": 0.7491,
+      "step": 137
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.8818897637795276e-05,
+      "loss": 0.7642,
+      "step": 138
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.8805774278215227e-05,
+      "loss": 0.7199,
+      "step": 139
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.879265091863517e-05,
+      "loss": 0.8203,
+      "step": 140
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.877952755905512e-05,
+      "loss": 0.7748,
+      "step": 141
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.8766404199475068e-05,
+      "loss": 0.7709,
+      "step": 142
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.8753280839895015e-05,
+      "loss": 0.8527,
+      "step": 143
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.8740157480314962e-05,
+      "loss": 0.6678,
+      "step": 144
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.872703412073491e-05,
+      "loss": 0.8097,
+      "step": 145
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.8713910761154856e-05,
+      "loss": 0.7239,
+      "step": 146
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.8700787401574803e-05,
+      "loss": 0.6475,
+      "step": 147
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.8687664041994754e-05,
+      "loss": 0.6625,
+      "step": 148
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.8674540682414698e-05,
+      "loss": 0.7769,
+      "step": 149
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.8661417322834648e-05,
+      "loss": 0.7377,
+      "step": 150
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.8648293963254595e-05,
+      "loss": 0.7023,
+      "step": 151
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.8635170603674542e-05,
+      "loss": 0.7637,
+      "step": 152
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.862204724409449e-05,
+      "loss": 0.7102,
+      "step": 153
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.8608923884514437e-05,
+      "loss": 0.7238,
+      "step": 154
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.8595800524934384e-05,
+      "loss": 0.7406,
+      "step": 155
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.858267716535433e-05,
+      "loss": 0.834,
+      "step": 156
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.856955380577428e-05,
+      "loss": 0.7231,
+      "step": 157
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.855643044619423e-05,
+      "loss": 0.6133,
+      "step": 158
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.8543307086614176e-05,
+      "loss": 0.7238,
+      "step": 159
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.8530183727034123e-05,
+      "loss": 0.6676,
+      "step": 160
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.851706036745407e-05,
+      "loss": 0.7014,
+      "step": 161
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.8503937007874017e-05,
+      "loss": 0.7926,
+      "step": 162
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.8490813648293964e-05,
+      "loss": 0.678,
+      "step": 163
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.847769028871391e-05,
+      "loss": 0.7271,
+      "step": 164
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.846456692913386e-05,
+      "loss": 0.7583,
+      "step": 165
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.8451443569553805e-05,
+      "loss": 0.7569,
+      "step": 166
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.8438320209973756e-05,
+      "loss": 0.6899,
+      "step": 167
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.8425196850393703e-05,
+      "loss": 0.6988,
+      "step": 168
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.841207349081365e-05,
+      "loss": 0.7114,
+      "step": 169
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.8398950131233597e-05,
+      "loss": 0.6809,
+      "step": 170
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.8385826771653544e-05,
+      "loss": 0.8045,
+      "step": 171
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.837270341207349e-05,
+      "loss": 0.7361,
+      "step": 172
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.835958005249344e-05,
+      "loss": 0.7402,
+      "step": 173
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.834645669291339e-05,
+      "loss": 0.7559,
+      "step": 174
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.7329,
+      "step": 175
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.8320209973753283e-05,
+      "loss": 0.735,
+      "step": 176
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.830708661417323e-05,
+      "loss": 0.6951,
+      "step": 177
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.8293963254593178e-05,
+      "loss": 0.7412,
+      "step": 178
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.8280839895013125e-05,
+      "loss": 0.8119,
+      "step": 179
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.8267716535433072e-05,
+      "loss": 0.7192,
+      "step": 180
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.825459317585302e-05,
+      "loss": 0.7069,
+      "step": 181
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.8241469816272966e-05,
+      "loss": 0.7952,
+      "step": 182
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.8228346456692917e-05,
+      "loss": 0.7064,
+      "step": 183
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.821522309711286e-05,
+      "loss": 0.7563,
+      "step": 184
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.820209973753281e-05,
+      "loss": 0.6829,
+      "step": 185
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.8188976377952758e-05,
+      "loss": 0.6525,
+      "step": 186
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.8175853018372705e-05,
+      "loss": 0.7344,
+      "step": 187
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.8162729658792652e-05,
+      "loss": 0.7181,
+      "step": 188
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.8149606299212603e-05,
+      "loss": 0.8006,
+      "step": 189
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.8136482939632546e-05,
+      "loss": 0.6388,
+      "step": 190
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.8123359580052497e-05,
+      "loss": 0.8009,
+      "step": 191
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.811023622047244e-05,
+      "loss": 0.7908,
+      "step": 192
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.809711286089239e-05,
+      "loss": 0.6814,
+      "step": 193
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.8083989501312338e-05,
+      "loss": 0.7376,
+      "step": 194
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8070866141732285e-05,
+      "loss": 0.7501,
+      "step": 195
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8057742782152232e-05,
+      "loss": 0.7356,
+      "step": 196
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.804461942257218e-05,
+      "loss": 0.7889,
+      "step": 197
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.8031496062992127e-05,
+      "loss": 0.7564,
+      "step": 198
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8018372703412074e-05,
+      "loss": 0.759,
+      "step": 199
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8005249343832024e-05,
+      "loss": 0.7271,
+      "step": 200
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.7992125984251968e-05,
+      "loss": 0.7391,
+      "step": 201
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.797900262467192e-05,
+      "loss": 0.7416,
+      "step": 202
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.7965879265091866e-05,
+      "loss": 0.816,
+      "step": 203
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.7952755905511813e-05,
+      "loss": 0.8431,
+      "step": 204
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.793963254593176e-05,
+      "loss": 0.7536,
+      "step": 205
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.7926509186351707e-05,
+      "loss": 0.8151,
+      "step": 206
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.7913385826771654e-05,
+      "loss": 0.7686,
+      "step": 207
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.79002624671916e-05,
+      "loss": 0.8059,
+      "step": 208
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.7887139107611552e-05,
+      "loss": 0.7326,
+      "step": 209
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.7874015748031495e-05,
+      "loss": 0.7793,
+      "step": 210
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.7860892388451446e-05,
+      "loss": 0.7138,
+      "step": 211
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.7847769028871393e-05,
+      "loss": 0.7229,
+      "step": 212
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.783464566929134e-05,
+      "loss": 0.6943,
+      "step": 213
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.7821522309711287e-05,
+      "loss": 0.7195,
+      "step": 214
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.7808398950131234e-05,
+      "loss": 0.7308,
+      "step": 215
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.779527559055118e-05,
+      "loss": 0.7402,
+      "step": 216
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.778215223097113e-05,
+      "loss": 0.744,
+      "step": 217
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.776902887139108e-05,
+      "loss": 0.7673,
+      "step": 218
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.7755905511811026e-05,
+      "loss": 0.7006,
+      "step": 219
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.7742782152230973e-05,
+      "loss": 0.6871,
+      "step": 220
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.772965879265092e-05,
+      "loss": 0.7029,
+      "step": 221
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.7716535433070868e-05,
+      "loss": 0.7153,
+      "step": 222
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.7703412073490815e-05,
+      "loss": 0.7385,
+      "step": 223
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.7690288713910762e-05,
+      "loss": 0.712,
+      "step": 224
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.767716535433071e-05,
+      "loss": 0.7185,
+      "step": 225
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.766404199475066e-05,
+      "loss": 0.7657,
+      "step": 226
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.7650918635170603e-05,
+      "loss": 0.7549,
+      "step": 227
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.7637795275590554e-05,
+      "loss": 0.7118,
+      "step": 228
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.76246719160105e-05,
+      "loss": 0.6845,
+      "step": 229
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.7611548556430448e-05,
+      "loss": 0.689,
+      "step": 230
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7598425196850395e-05,
+      "loss": 0.7365,
+      "step": 231
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7585301837270342e-05,
+      "loss": 0.7297,
+      "step": 232
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.757217847769029e-05,
+      "loss": 0.7449,
+      "step": 233
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.7559055118110236e-05,
+      "loss": 0.7652,
+      "step": 234
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7545931758530187e-05,
+      "loss": 0.719,
+      "step": 235
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.753280839895013e-05,
+      "loss": 0.6519,
+      "step": 236
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.751968503937008e-05,
+      "loss": 0.7801,
+      "step": 237
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.7506561679790028e-05,
+      "loss": 0.7334,
+      "step": 238
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7493438320209975e-05,
+      "loss": 0.689,
+      "step": 239
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7480314960629923e-05,
+      "loss": 0.8085,
+      "step": 240
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.746719160104987e-05,
+      "loss": 0.8845,
+      "step": 241
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.7454068241469817e-05,
+      "loss": 0.6735,
+      "step": 242
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.7440944881889764e-05,
+      "loss": 0.7002,
+      "step": 243
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.7427821522309714e-05,
+      "loss": 0.6246,
+      "step": 244
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.7414698162729658e-05,
+      "loss": 0.6858,
+      "step": 245
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.740157480314961e-05,
+      "loss": 0.6897,
+      "step": 246
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7388451443569556e-05,
+      "loss": 0.6294,
+      "step": 247
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7375328083989503e-05,
+      "loss": 0.6608,
+      "step": 248
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.736220472440945e-05,
+      "loss": 0.6644,
+      "step": 249
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.7349081364829397e-05,
+      "loss": 0.6894,
+      "step": 250
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7335958005249344e-05,
+      "loss": 0.6826,
+      "step": 251
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7322834645669295e-05,
+      "loss": 0.719,
+      "step": 252
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.7309711286089242e-05,
+      "loss": 0.7302,
+      "step": 253
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.729658792650919e-05,
+      "loss": 0.7835,
+      "step": 254
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7283464566929136e-05,
+      "loss": 0.8102,
+      "step": 255
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7270341207349083e-05,
+      "loss": 0.7183,
+      "step": 256
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.725721784776903e-05,
+      "loss": 0.7524,
+      "step": 257
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7244094488188977e-05,
+      "loss": 0.7385,
+      "step": 258
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7230971128608925e-05,
+      "loss": 0.7174,
+      "step": 259
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.721784776902887e-05,
+      "loss": 0.6565,
+      "step": 260
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7204724409448822e-05,
+      "loss": 0.7028,
+      "step": 261
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.7191601049868766e-05,
+      "loss": 0.6655,
+      "step": 262
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.7178477690288716e-05,
+      "loss": 0.6893,
+      "step": 263
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.7165354330708663e-05,
+      "loss": 0.6681,
+      "step": 264
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.715223097112861e-05,
+      "loss": 0.6808,
+      "step": 265
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7139107611548558e-05,
+      "loss": 0.6594,
+      "step": 266
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7125984251968505e-05,
+      "loss": 0.7971,
+      "step": 267
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.7112860892388452e-05,
+      "loss": 0.7395,
+      "step": 268
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.70997375328084e-05,
+      "loss": 0.6938,
+      "step": 269
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.708661417322835e-05,
+      "loss": 0.7269,
+      "step": 270
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7073490813648293e-05,
+      "loss": 0.6377,
+      "step": 271
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7060367454068244e-05,
+      "loss": 0.6876,
+      "step": 272
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.704724409448819e-05,
+      "loss": 0.7468,
+      "step": 273
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7034120734908138e-05,
+      "loss": 0.7081,
+      "step": 274
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7020997375328085e-05,
+      "loss": 0.7155,
+      "step": 275
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7007874015748032e-05,
+      "loss": 0.7403,
+      "step": 276
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.699475065616798e-05,
+      "loss": 0.7509,
+      "step": 277
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6981627296587927e-05,
+      "loss": 0.7691,
+      "step": 278
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.6968503937007877e-05,
+      "loss": 0.6769,
+      "step": 279
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.695538057742782e-05,
+      "loss": 0.7203,
+      "step": 280
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.694225721784777e-05,
+      "loss": 0.6667,
+      "step": 281
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.692913385826772e-05,
+      "loss": 0.6891,
+      "step": 282
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6916010498687665e-05,
+      "loss": 0.6798,
+      "step": 283
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6902887139107613e-05,
+      "loss": 0.6851,
+      "step": 284
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.6889763779527563e-05,
+      "loss": 0.6759,
+      "step": 285
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.6876640419947507e-05,
+      "loss": 0.704,
+      "step": 286
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.6863517060367457e-05,
+      "loss": 0.6572,
+      "step": 287
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.68503937007874e-05,
+      "loss": 0.7341,
+      "step": 288
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.683727034120735e-05,
+      "loss": 0.7115,
+      "step": 289
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.68241469816273e-05,
+      "loss": 0.7271,
+      "step": 290
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6811023622047246e-05,
+      "loss": 0.703,
+      "step": 291
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6797900262467193e-05,
+      "loss": 0.7439,
+      "step": 292
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.678477690288714e-05,
+      "loss": 0.6744,
+      "step": 293
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.6771653543307087e-05,
+      "loss": 0.7919,
+      "step": 294
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.6758530183727034e-05,
+      "loss": 0.773,
+      "step": 295
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.6745406824146985e-05,
+      "loss": 0.7258,
+      "step": 296
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.673228346456693e-05,
+      "loss": 0.7358,
+      "step": 297
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.671916010498688e-05,
+      "loss": 0.7336,
+      "step": 298
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.6706036745406826e-05,
+      "loss": 0.6757,
+      "step": 299
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.6692913385826773e-05,
+      "loss": 0.7411,
+      "step": 300
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.667979002624672e-05,
+      "loss": 0.8021,
+      "step": 301
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.7287,
+      "step": 302
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.6653543307086615e-05,
+      "loss": 0.7613,
+      "step": 303
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.6640419947506562e-05,
+      "loss": 0.8577,
+      "step": 304
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.6627296587926512e-05,
+      "loss": 0.7202,
+      "step": 305
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.6614173228346456e-05,
+      "loss": 0.7679,
+      "step": 306
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.6601049868766406e-05,
+      "loss": 0.6743,
+      "step": 307
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.6587926509186354e-05,
+      "loss": 0.7753,
+      "step": 308
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.65748031496063e-05,
+      "loss": 0.7147,
+      "step": 309
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.6561679790026248e-05,
+      "loss": 0.6653,
+      "step": 310
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.6548556430446195e-05,
+      "loss": 0.6641,
+      "step": 311
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.6535433070866142e-05,
+      "loss": 0.7154,
+      "step": 312
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.6522309711286093e-05,
+      "loss": 0.7679,
+      "step": 313
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.650918635170604e-05,
+      "loss": 0.6796,
+      "step": 314
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.6496062992125987e-05,
+      "loss": 0.7906,
+      "step": 315
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.6482939632545934e-05,
+      "loss": 0.6748,
+      "step": 316
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.646981627296588e-05,
+      "loss": 0.6933,
+      "step": 317
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.6456692913385828e-05,
+      "loss": 0.7913,
+      "step": 318
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.6443569553805775e-05,
+      "loss": 0.7508,
+      "step": 319
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.6430446194225722e-05,
+      "loss": 0.6814,
+      "step": 320
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.641732283464567e-05,
+      "loss": 0.7016,
+      "step": 321
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.640419947506562e-05,
+      "loss": 0.7073,
+      "step": 322
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6391076115485564e-05,
+      "loss": 0.7956,
+      "step": 323
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6377952755905514e-05,
+      "loss": 0.6889,
+      "step": 324
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.636482939632546e-05,
+      "loss": 0.6842,
+      "step": 325
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.635170603674541e-05,
+      "loss": 0.7083,
+      "step": 326
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6338582677165356e-05,
+      "loss": 0.695,
+      "step": 327
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6325459317585303e-05,
+      "loss": 0.6817,
+      "step": 328
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.631233595800525e-05,
+      "loss": 0.6597,
+      "step": 329
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6299212598425197e-05,
+      "loss": 0.8122,
+      "step": 330
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6286089238845147e-05,
+      "loss": 0.7084,
+      "step": 331
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.627296587926509e-05,
+      "loss": 0.6669,
+      "step": 332
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.625984251968504e-05,
+      "loss": 0.7016,
+      "step": 333
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.624671916010499e-05,
+      "loss": 0.6762,
+      "step": 334
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6233595800524936e-05,
+      "loss": 0.707,
+      "step": 335
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6220472440944883e-05,
+      "loss": 0.7484,
+      "step": 336
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.620734908136483e-05,
+      "loss": 0.6942,
+      "step": 337
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6194225721784777e-05,
+      "loss": 0.657,
+      "step": 338
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6181102362204724e-05,
+      "loss": 0.7008,
+      "step": 339
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6167979002624675e-05,
+      "loss": 0.7285,
+      "step": 340
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.615485564304462e-05,
+      "loss": 0.6876,
+      "step": 341
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.614173228346457e-05,
+      "loss": 0.7145,
+      "step": 342
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6128608923884516e-05,
+      "loss": 0.6675,
+      "step": 343
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6115485564304463e-05,
+      "loss": 0.6565,
+      "step": 344
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.610236220472441e-05,
+      "loss": 0.7392,
+      "step": 345
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.608923884514436e-05,
+      "loss": 0.6847,
+      "step": 346
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6076115485564305e-05,
+      "loss": 0.6308,
+      "step": 347
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6062992125984255e-05,
+      "loss": 0.6739,
+      "step": 348
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6049868766404202e-05,
+      "loss": 0.645,
+      "step": 349
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.603674540682415e-05,
+      "loss": 0.6569,
+      "step": 350
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6023622047244096e-05,
+      "loss": 0.8016,
+      "step": 351
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6010498687664044e-05,
+      "loss": 0.6679,
+      "step": 352
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.599737532808399e-05,
+      "loss": 0.7299,
+      "step": 353
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.5984251968503938e-05,
+      "loss": 0.6514,
+      "step": 354
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.5971128608923885e-05,
+      "loss": 0.6613,
+      "step": 355
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.5958005249343832e-05,
+      "loss": 0.6922,
+      "step": 356
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.5944881889763783e-05,
+      "loss": 0.7856,
+      "step": 357
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.5931758530183726e-05,
+      "loss": 0.7365,
+      "step": 358
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.5918635170603677e-05,
+      "loss": 0.6569,
+      "step": 359
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.5905511811023624e-05,
+      "loss": 0.7049,
+      "step": 360
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.589238845144357e-05,
+      "loss": 0.7298,
+      "step": 361
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5879265091863518e-05,
+      "loss": 0.6846,
+      "step": 362
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5866141732283465e-05,
+      "loss": 0.6876,
+      "step": 363
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5853018372703412e-05,
+      "loss": 0.8115,
+      "step": 364
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.583989501312336e-05,
+      "loss": 0.7524,
+      "step": 365
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.582677165354331e-05,
+      "loss": 0.6977,
+      "step": 366
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.5813648293963254e-05,
+      "loss": 0.7187,
+      "step": 367
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.5800524934383204e-05,
+      "loss": 0.7026,
+      "step": 368
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.578740157480315e-05,
+      "loss": 0.7194,
+      "step": 369
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.57742782152231e-05,
+      "loss": 0.6877,
+      "step": 370
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.5761154855643046e-05,
+      "loss": 0.8635,
+      "step": 371
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.5748031496062993e-05,
+      "loss": 0.6863,
+      "step": 372
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.573490813648294e-05,
+      "loss": 0.6409,
+      "step": 373
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5721784776902887e-05,
+      "loss": 0.7026,
+      "step": 374
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5708661417322837e-05,
+      "loss": 0.7152,
+      "step": 375
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5695538057742785e-05,
+      "loss": 0.6851,
+      "step": 376
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.568241469816273e-05,
+      "loss": 0.7097,
+      "step": 377
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.566929133858268e-05,
+      "loss": 0.6708,
+      "step": 378
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5656167979002626e-05,
+      "loss": 0.7262,
+      "step": 379
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5643044619422573e-05,
+      "loss": 0.692,
+      "step": 380
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5629921259842524e-05,
+      "loss": 0.7964,
+      "step": 381
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5616797900262467e-05,
+      "loss": 0.6485,
+      "step": 382
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5603674540682418e-05,
+      "loss": 0.6548,
+      "step": 383
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.559055118110236e-05,
+      "loss": 0.7234,
+      "step": 384
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5577427821522312e-05,
+      "loss": 0.6184,
+      "step": 385
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.556430446194226e-05,
+      "loss": 0.6784,
+      "step": 386
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5551181102362206e-05,
+      "loss": 0.6862,
+      "step": 387
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5538057742782153e-05,
+      "loss": 0.6898,
+      "step": 388
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.55249343832021e-05,
+      "loss": 0.6877,
+      "step": 389
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5511811023622048e-05,
+      "loss": 0.63,
+      "step": 390
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5498687664041995e-05,
+      "loss": 0.7392,
+      "step": 391
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5485564304461945e-05,
+      "loss": 0.7621,
+      "step": 392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.547244094488189e-05,
+      "loss": 0.7172,
+      "step": 393
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.545931758530184e-05,
+      "loss": 0.6579,
+      "step": 394
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5446194225721787e-05,
+      "loss": 0.6238,
+      "step": 395
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5433070866141734e-05,
+      "loss": 0.6394,
+      "step": 396
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.541994750656168e-05,
+      "loss": 0.6816,
+      "step": 397
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5406824146981628e-05,
+      "loss": 0.6576,
+      "step": 398
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5393700787401575e-05,
+      "loss": 0.657,
+      "step": 399
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5380577427821522e-05,
+      "loss": 0.7055,
+      "step": 400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5367454068241473e-05,
+      "loss": 0.5969,
+      "step": 401
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5354330708661416e-05,
+      "loss": 0.7107,
+      "step": 402
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5341207349081367e-05,
+      "loss": 0.6595,
+      "step": 403
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5328083989501314e-05,
+      "loss": 0.7404,
+      "step": 404
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.531496062992126e-05,
+      "loss": 0.6249,
+      "step": 405
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5301837270341208e-05,
+      "loss": 0.6453,
+      "step": 406
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.528871391076116e-05,
+      "loss": 0.7113,
+      "step": 407
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5275590551181102e-05,
+      "loss": 0.7842,
+      "step": 408
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5262467191601053e-05,
+      "loss": 0.6437,
+      "step": 409
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5249343832021e-05,
+      "loss": 0.6802,
+      "step": 410
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5236220472440946e-05,
+      "loss": 0.6295,
+      "step": 411
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5223097112860894e-05,
+      "loss": 0.6469,
+      "step": 412
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5209973753280841e-05,
+      "loss": 0.7367,
+      "step": 413
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5196850393700789e-05,
+      "loss": 0.6417,
+      "step": 414
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5183727034120737e-05,
+      "loss": 0.7288,
+      "step": 415
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5170603674540683e-05,
+      "loss": 0.796,
+      "step": 416
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5157480314960632e-05,
+      "loss": 0.6668,
+      "step": 417
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5144356955380579e-05,
+      "loss": 0.7408,
+      "step": 418
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5131233595800526e-05,
+      "loss": 0.6293,
+      "step": 419
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5118110236220473e-05,
+      "loss": 0.6605,
+      "step": 420
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5104986876640422e-05,
+      "loss": 0.6289,
+      "step": 421
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5091863517060367e-05,
+      "loss": 0.585,
+      "step": 422
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5078740157480316e-05,
+      "loss": 0.6207,
+      "step": 423
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.5065616797900265e-05,
+      "loss": 0.671,
+      "step": 424
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.505249343832021e-05,
+      "loss": 0.6277,
+      "step": 425
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.5039370078740159e-05,
+      "loss": 0.6535,
+      "step": 426
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.5026246719160106e-05,
+      "loss": 0.5868,
+      "step": 427
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.5013123359580053e-05,
+      "loss": 0.7606,
+      "step": 428
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.6605,
+      "step": 429
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.498687664041995e-05,
+      "loss": 0.6228,
+      "step": 430
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4973753280839896e-05,
+      "loss": 0.7267,
+      "step": 431
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4960629921259843e-05,
+      "loss": 0.6723,
+      "step": 432
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4947506561679792e-05,
+      "loss": 0.6098,
+      "step": 433
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4934383202099738e-05,
+      "loss": 0.6253,
+      "step": 434
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4921259842519686e-05,
+      "loss": 0.6369,
+      "step": 435
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4908136482939635e-05,
+      "loss": 0.7005,
+      "step": 436
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.489501312335958e-05,
+      "loss": 0.679,
+      "step": 437
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.488188976377953e-05,
+      "loss": 0.7713,
+      "step": 438
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4868766404199477e-05,
+      "loss": 0.6451,
+      "step": 439
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4855643044619424e-05,
+      "loss": 0.6409,
+      "step": 440
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4842519685039371e-05,
+      "loss": 0.7342,
+      "step": 441
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.482939632545932e-05,
+      "loss": 0.6575,
+      "step": 442
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4816272965879265e-05,
+      "loss": 0.6358,
+      "step": 443
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4803149606299214e-05,
+      "loss": 0.5813,
+      "step": 444
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4790026246719161e-05,
+      "loss": 0.6427,
+      "step": 445
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4776902887139108e-05,
+      "loss": 0.7327,
+      "step": 446
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4763779527559057e-05,
+      "loss": 0.6833,
+      "step": 447
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4750656167979002e-05,
+      "loss": 0.6715,
+      "step": 448
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4737532808398951e-05,
+      "loss": 0.6244,
+      "step": 449
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.47244094488189e-05,
+      "loss": 0.5874,
+      "step": 450
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4711286089238845e-05,
+      "loss": 0.8781,
+      "step": 451
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4698162729658794e-05,
+      "loss": 0.5761,
+      "step": 452
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4685039370078741e-05,
+      "loss": 0.6387,
+      "step": 453
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4671916010498688e-05,
+      "loss": 0.6827,
+      "step": 454
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4658792650918636e-05,
+      "loss": 0.697,
+      "step": 455
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4645669291338584e-05,
+      "loss": 0.7584,
+      "step": 456
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.463254593175853e-05,
+      "loss": 0.6165,
+      "step": 457
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4619422572178479e-05,
+      "loss": 0.6614,
+      "step": 458
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.4606299212598427e-05,
+      "loss": 0.6898,
+      "step": 459
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.4593175853018373e-05,
+      "loss": 0.7243,
+      "step": 460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.4580052493438322e-05,
+      "loss": 0.6284,
+      "step": 461
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.456692913385827e-05,
+      "loss": 0.6523,
+      "step": 462
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.4553805774278216e-05,
+      "loss": 0.6886,
+      "step": 463
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.4540682414698165e-05,
+      "loss": 0.6592,
+      "step": 464
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.4527559055118112e-05,
+      "loss": 0.5917,
+      "step": 465
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.4514435695538059e-05,
+      "loss": 0.6923,
+      "step": 466
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.4501312335958006e-05,
+      "loss": 0.7126,
+      "step": 467
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.4488188976377955e-05,
+      "loss": 0.6608,
+      "step": 468
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.44750656167979e-05,
+      "loss": 0.6697,
+      "step": 469
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.4461942257217849e-05,
+      "loss": 0.6865,
+      "step": 470
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.4448818897637798e-05,
+      "loss": 0.6339,
+      "step": 471
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.4435695538057743e-05,
+      "loss": 0.6488,
+      "step": 472
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.4422572178477692e-05,
+      "loss": 0.6327,
+      "step": 473
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.440944881889764e-05,
+      "loss": 0.6231,
+      "step": 474
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.4396325459317586e-05,
+      "loss": 0.6524,
+      "step": 475
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.4383202099737535e-05,
+      "loss": 0.6779,
+      "step": 476
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.437007874015748e-05,
+      "loss": 0.5976,
+      "step": 477
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.435695538057743e-05,
+      "loss": 0.6892,
+      "step": 478
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.4343832020997377e-05,
+      "loss": 0.6442,
+      "step": 479
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.4330708661417324e-05,
+      "loss": 0.662,
+      "step": 480
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.431758530183727e-05,
+      "loss": 0.7119,
+      "step": 481
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.430446194225722e-05,
+      "loss": 0.7041,
+      "step": 482
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.4291338582677165e-05,
+      "loss": 0.6616,
+      "step": 483
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.4278215223097114e-05,
+      "loss": 0.6021,
+      "step": 484
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.4265091863517063e-05,
+      "loss": 0.6857,
+      "step": 485
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.4251968503937008e-05,
+      "loss": 0.6354,
+      "step": 486
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.4238845144356957e-05,
+      "loss": 0.7394,
+      "step": 487
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.4225721784776904e-05,
+      "loss": 0.6941,
+      "step": 488
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.4212598425196851e-05,
+      "loss": 0.6883,
+      "step": 489
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.4199475065616798e-05,
+      "loss": 0.6294,
+      "step": 490
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.4186351706036747e-05,
+      "loss": 0.6858,
+      "step": 491
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.4173228346456694e-05,
+      "loss": 0.7239,
+      "step": 492
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.4160104986876641e-05,
+      "loss": 0.6474,
+      "step": 493
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.414698162729659e-05,
+      "loss": 0.7456,
+      "step": 494
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.4133858267716535e-05,
+      "loss": 0.7308,
+      "step": 495
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.4120734908136484e-05,
+      "loss": 0.6171,
+      "step": 496
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.4107611548556433e-05,
+      "loss": 0.6873,
+      "step": 497
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.4094488188976379e-05,
+      "loss": 0.6792,
+      "step": 498
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.4081364829396327e-05,
+      "loss": 0.6495,
+      "step": 499
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.4068241469816274e-05,
+      "loss": 0.6466,
+      "step": 500
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.4055118110236222e-05,
+      "loss": 0.6188,
+      "step": 501
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.4041994750656169e-05,
+      "loss": 0.7188,
+      "step": 502
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.4028871391076117e-05,
+      "loss": 0.6322,
+      "step": 503
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.4015748031496063e-05,
+      "loss": 0.719,
+      "step": 504
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.4002624671916012e-05,
+      "loss": 0.6446,
+      "step": 505
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.398950131233596e-05,
+      "loss": 0.6482,
+      "step": 506
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.3976377952755906e-05,
+      "loss": 0.727,
+      "step": 507
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.3963254593175855e-05,
+      "loss": 0.5877,
+      "step": 508
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.39501312335958e-05,
+      "loss": 0.626,
+      "step": 509
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.3937007874015749e-05,
+      "loss": 0.7112,
+      "step": 510
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.3923884514435698e-05,
+      "loss": 0.6759,
+      "step": 511
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.3910761154855643e-05,
+      "loss": 0.6676,
+      "step": 512
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.3897637795275592e-05,
+      "loss": 0.6789,
+      "step": 513
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.388451443569554e-05,
+      "loss": 0.7629,
+      "step": 514
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.3871391076115486e-05,
+      "loss": 0.6798,
+      "step": 515
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.3858267716535433e-05,
+      "loss": 0.7309,
+      "step": 516
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.3845144356955382e-05,
+      "loss": 0.6596,
+      "step": 517
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3832020997375328e-05,
+      "loss": 0.6956,
+      "step": 518
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3818897637795276e-05,
+      "loss": 0.712,
+      "step": 519
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3805774278215225e-05,
+      "loss": 0.7014,
+      "step": 520
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.379265091863517e-05,
+      "loss": 0.6591,
+      "step": 521
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.377952755905512e-05,
+      "loss": 0.6482,
+      "step": 522
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3766404199475068e-05,
+      "loss": 0.6946,
+      "step": 523
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3753280839895014e-05,
+      "loss": 0.5933,
+      "step": 524
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3740157480314963e-05,
+      "loss": 0.6783,
+      "step": 525
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.372703412073491e-05,
+      "loss": 0.6338,
+      "step": 526
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3713910761154857e-05,
+      "loss": 0.6873,
+      "step": 527
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3700787401574804e-05,
+      "loss": 0.5833,
+      "step": 528
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3687664041994753e-05,
+      "loss": 0.6784,
+      "step": 529
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3674540682414698e-05,
+      "loss": 0.6976,
+      "step": 530
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3661417322834647e-05,
+      "loss": 0.7271,
+      "step": 531
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3648293963254596e-05,
+      "loss": 0.6426,
+      "step": 532
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3635170603674541e-05,
+      "loss": 0.6955,
+      "step": 533
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.362204724409449e-05,
+      "loss": 0.7484,
+      "step": 534
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3608923884514437e-05,
+      "loss": 0.6906,
+      "step": 535
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3595800524934384e-05,
+      "loss": 0.6727,
+      "step": 536
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3582677165354331e-05,
+      "loss": 0.6647,
+      "step": 537
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.356955380577428e-05,
+      "loss": 0.6372,
+      "step": 538
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3556430446194227e-05,
+      "loss": 0.5744,
+      "step": 539
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3543307086614174e-05,
+      "loss": 0.625,
+      "step": 540
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3530183727034121e-05,
+      "loss": 0.6893,
+      "step": 541
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3517060367454069e-05,
+      "loss": 0.6194,
+      "step": 542
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3503937007874017e-05,
+      "loss": 0.6335,
+      "step": 543
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3490813648293963e-05,
+      "loss": 0.6877,
+      "step": 544
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3477690288713912e-05,
+      "loss": 0.6932,
+      "step": 545
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.346456692913386e-05,
+      "loss": 0.7087,
+      "step": 546
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3451443569553806e-05,
+      "loss": 0.6365,
+      "step": 547
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3438320209973755e-05,
+      "loss": 0.6548,
+      "step": 548
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3425196850393702e-05,
+      "loss": 0.6501,
+      "step": 549
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.3412073490813649e-05,
+      "loss": 0.8315,
+      "step": 550
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.3398950131233596e-05,
+      "loss": 0.7062,
+      "step": 551
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.3385826771653545e-05,
+      "loss": 0.6554,
+      "step": 552
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.337270341207349e-05,
+      "loss": 0.6587,
+      "step": 553
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.3359580052493439e-05,
+      "loss": 0.6427,
+      "step": 554
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.3346456692913388e-05,
+      "loss": 0.6918,
+      "step": 555
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6768,
+      "step": 556
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.3320209973753282e-05,
+      "loss": 0.6826,
+      "step": 557
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.3307086614173231e-05,
+      "loss": 0.6255,
+      "step": 558
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.3293963254593176e-05,
+      "loss": 0.6723,
+      "step": 559
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.3280839895013125e-05,
+      "loss": 0.6109,
+      "step": 560
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.3267716535433072e-05,
+      "loss": 0.6529,
+      "step": 561
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.325459317585302e-05,
+      "loss": 0.6983,
+      "step": 562
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.3241469816272966e-05,
+      "loss": 0.6914,
+      "step": 563
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.3228346456692915e-05,
+      "loss": 0.6383,
+      "step": 564
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.321522309711286e-05,
+      "loss": 0.6393,
+      "step": 565
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.320209973753281e-05,
+      "loss": 0.7771,
+      "step": 566
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.3188976377952758e-05,
+      "loss": 0.8403,
+      "step": 567
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.3175853018372704e-05,
+      "loss": 0.728,
+      "step": 568
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.3162729658792653e-05,
+      "loss": 0.7123,
+      "step": 569
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.3149606299212601e-05,
+      "loss": 0.6901,
+      "step": 570
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.3136482939632547e-05,
+      "loss": 0.6541,
+      "step": 571
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.3123359580052496e-05,
+      "loss": 0.6734,
+      "step": 572
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.3110236220472441e-05,
+      "loss": 0.6712,
+      "step": 573
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.309711286089239e-05,
+      "loss": 0.6116,
+      "step": 574
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.3083989501312337e-05,
+      "loss": 0.6835,
+      "step": 575
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.3070866141732284e-05,
+      "loss": 0.6363,
+      "step": 576
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.3057742782152231e-05,
+      "loss": 0.6991,
+      "step": 577
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.304461942257218e-05,
+      "loss": 0.6465,
+      "step": 578
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.3031496062992125e-05,
+      "loss": 0.6969,
+      "step": 579
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.3018372703412074e-05,
+      "loss": 0.7316,
+      "step": 580
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.3005249343832023e-05,
+      "loss": 0.6521,
+      "step": 581
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.2992125984251968e-05,
+      "loss": 0.6664,
+      "step": 582
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.2979002624671917e-05,
+      "loss": 0.6338,
+      "step": 583
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.2965879265091864e-05,
+      "loss": 0.637,
+      "step": 584
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.2952755905511812e-05,
+      "loss": 0.6806,
+      "step": 585
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.293963254593176e-05,
+      "loss": 0.7268,
+      "step": 586
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.2926509186351707e-05,
+      "loss": 0.6092,
+      "step": 587
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.2913385826771655e-05,
+      "loss": 0.7613,
+      "step": 588
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.2900262467191602e-05,
+      "loss": 0.6138,
+      "step": 589
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.288713910761155e-05,
+      "loss": 0.6425,
+      "step": 590
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.2874015748031496e-05,
+      "loss": 0.6976,
+      "step": 591
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.2860892388451445e-05,
+      "loss": 0.7524,
+      "step": 592
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.2847769028871394e-05,
+      "loss": 0.6862,
+      "step": 593
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.2834645669291339e-05,
+      "loss": 0.6916,
+      "step": 594
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.2821522309711288e-05,
+      "loss": 0.6429,
+      "step": 595
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.2808398950131235e-05,
+      "loss": 0.7474,
+      "step": 596
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.2795275590551182e-05,
+      "loss": 0.6388,
+      "step": 597
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.2782152230971129e-05,
+      "loss": 0.7393,
+      "step": 598
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.2769028871391078e-05,
+      "loss": 0.632,
+      "step": 599
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.2755905511811025e-05,
+      "loss": 0.6251,
+      "step": 600
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.2742782152230972e-05,
+      "loss": 0.6984,
+      "step": 601
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.2729658792650921e-05,
+      "loss": 0.6141,
+      "step": 602
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.2716535433070866e-05,
+      "loss": 0.6312,
+      "step": 603
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.2703412073490815e-05,
+      "loss": 0.5493,
+      "step": 604
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.269028871391076e-05,
+      "loss": 0.6029,
+      "step": 605
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.267716535433071e-05,
+      "loss": 0.6128,
+      "step": 606
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.2664041994750658e-05,
+      "loss": 0.6787,
+      "step": 607
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.2650918635170604e-05,
+      "loss": 0.6808,
+      "step": 608
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.2637795275590552e-05,
+      "loss": 0.7465,
+      "step": 609
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.26246719160105e-05,
+      "loss": 0.5791,
+      "step": 610
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.2611548556430447e-05,
+      "loss": 0.7059,
+      "step": 611
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.2598425196850394e-05,
+      "loss": 0.6348,
+      "step": 612
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.2585301837270343e-05,
+      "loss": 0.6382,
+      "step": 613
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.2572178477690288e-05,
+      "loss": 0.6318,
+      "step": 614
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.2559055118110237e-05,
+      "loss": 0.6412,
+      "step": 615
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.2545931758530186e-05,
+      "loss": 0.6998,
+      "step": 616
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.2532808398950131e-05,
+      "loss": 0.6655,
+      "step": 617
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.251968503937008e-05,
+      "loss": 0.6509,
+      "step": 618
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.2506561679790029e-05,
+      "loss": 0.6825,
+      "step": 619
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.2493438320209974e-05,
+      "loss": 0.7011,
+      "step": 620
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.2480314960629923e-05,
+      "loss": 0.6242,
+      "step": 621
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.246719160104987e-05,
+      "loss": 0.6564,
+      "step": 622
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.2454068241469817e-05,
+      "loss": 0.744,
+      "step": 623
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.2440944881889764e-05,
+      "loss": 0.6374,
+      "step": 624
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.2427821522309713e-05,
+      "loss": 0.6434,
+      "step": 625
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.2414698162729659e-05,
+      "loss": 0.7321,
+      "step": 626
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.2401574803149607e-05,
+      "loss": 0.6143,
+      "step": 627
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.2388451443569556e-05,
+      "loss": 0.6818,
+      "step": 628
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.2375328083989502e-05,
+      "loss": 0.6714,
+      "step": 629
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.236220472440945e-05,
+      "loss": 0.6964,
+      "step": 630
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.2349081364829398e-05,
+      "loss": 0.7756,
+      "step": 631
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.2335958005249345e-05,
+      "loss": 0.7802,
+      "step": 632
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.2322834645669293e-05,
+      "loss": 0.6589,
+      "step": 633
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.230971128608924e-05,
+      "loss": 0.7549,
+      "step": 634
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.2296587926509188e-05,
+      "loss": 0.7144,
+      "step": 635
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.2283464566929135e-05,
+      "loss": 0.6988,
+      "step": 636
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.2270341207349082e-05,
+      "loss": 0.6001,
+      "step": 637
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.2257217847769029e-05,
+      "loss": 0.6947,
+      "step": 638
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.2244094488188978e-05,
+      "loss": 0.645,
+      "step": 639
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.2230971128608923e-05,
+      "loss": 0.66,
+      "step": 640
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.2217847769028872e-05,
+      "loss": 0.6684,
+      "step": 641
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.2204724409448821e-05,
+      "loss": 0.7285,
+      "step": 642
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.2191601049868766e-05,
+      "loss": 0.6325,
+      "step": 643
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2178477690288715e-05,
+      "loss": 0.6233,
+      "step": 644
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2165354330708662e-05,
+      "loss": 0.6203,
+      "step": 645
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.215223097112861e-05,
+      "loss": 0.6795,
+      "step": 646
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2139107611548558e-05,
+      "loss": 0.6835,
+      "step": 647
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2125984251968505e-05,
+      "loss": 0.6351,
+      "step": 648
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2112860892388452e-05,
+      "loss": 0.6153,
+      "step": 649
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.20997375328084e-05,
+      "loss": 0.6549,
+      "step": 650
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2086614173228348e-05,
+      "loss": 0.5985,
+      "step": 651
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2073490813648294e-05,
+      "loss": 0.6196,
+      "step": 652
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2060367454068243e-05,
+      "loss": 0.6036,
+      "step": 653
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2047244094488191e-05,
+      "loss": 0.733,
+      "step": 654
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2034120734908137e-05,
+      "loss": 0.655,
+      "step": 655
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2020997375328086e-05,
+      "loss": 0.6435,
+      "step": 656
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2007874015748033e-05,
+      "loss": 0.6405,
+      "step": 657
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.199475065616798e-05,
+      "loss": 0.703,
+      "step": 658
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.1981627296587927e-05,
+      "loss": 0.6925,
+      "step": 659
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.1968503937007876e-05,
+      "loss": 0.6758,
+      "step": 660
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.1955380577427821e-05,
+      "loss": 0.7093,
+      "step": 661
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.194225721784777e-05,
+      "loss": 0.6541,
+      "step": 662
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.1929133858267719e-05,
+      "loss": 0.5868,
+      "step": 663
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.1916010498687664e-05,
+      "loss": 0.6866,
+      "step": 664
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.1902887139107613e-05,
+      "loss": 0.6209,
+      "step": 665
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.1889763779527562e-05,
+      "loss": 0.6694,
+      "step": 666
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.1876640419947507e-05,
+      "loss": 0.8347,
+      "step": 667
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1863517060367456e-05,
+      "loss": 0.8037,
+      "step": 668
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1850393700787401e-05,
+      "loss": 0.6485,
+      "step": 669
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.183727034120735e-05,
+      "loss": 0.6699,
+      "step": 670
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1824146981627297e-05,
+      "loss": 0.7128,
+      "step": 671
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1811023622047245e-05,
+      "loss": 0.6803,
+      "step": 672
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1797900262467192e-05,
+      "loss": 0.6019,
+      "step": 673
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.178477690288714e-05,
+      "loss": 0.603,
+      "step": 674
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1771653543307086e-05,
+      "loss": 0.7078,
+      "step": 675
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1758530183727035e-05,
+      "loss": 0.6425,
+      "step": 676
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1745406824146984e-05,
+      "loss": 0.6275,
+      "step": 677
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1732283464566929e-05,
+      "loss": 0.6658,
+      "step": 678
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1719160104986878e-05,
+      "loss": 0.6794,
+      "step": 679
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1706036745406827e-05,
+      "loss": 0.6406,
+      "step": 680
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1692913385826772e-05,
+      "loss": 0.6341,
+      "step": 681
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.167979002624672e-05,
+      "loss": 0.6533,
+      "step": 682
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 0.7331,
+      "step": 683
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1653543307086615e-05,
+      "loss": 0.658,
+      "step": 684
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1640419947506562e-05,
+      "loss": 0.6517,
+      "step": 685
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1627296587926511e-05,
+      "loss": 0.6398,
+      "step": 686
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1614173228346456e-05,
+      "loss": 0.674,
+      "step": 687
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1601049868766405e-05,
+      "loss": 0.6094,
+      "step": 688
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1587926509186354e-05,
+      "loss": 0.5922,
+      "step": 689
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.15748031496063e-05,
+      "loss": 0.6527,
+      "step": 690
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1561679790026248e-05,
+      "loss": 0.6102,
+      "step": 691
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1548556430446195e-05,
+      "loss": 0.6422,
+      "step": 692
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1535433070866142e-05,
+      "loss": 0.6254,
+      "step": 693
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1522309711286091e-05,
+      "loss": 0.5598,
+      "step": 694
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1509186351706038e-05,
+      "loss": 0.6273,
+      "step": 695
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.1496062992125985e-05,
+      "loss": 0.6342,
+      "step": 696
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.1482939632545933e-05,
+      "loss": 0.7083,
+      "step": 697
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.1469816272965881e-05,
+      "loss": 0.683,
+      "step": 698
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.1456692913385827e-05,
+      "loss": 0.5848,
+      "step": 699
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.1443569553805776e-05,
+      "loss": 0.7239,
+      "step": 700
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.1430446194225721e-05,
+      "loss": 0.5985,
+      "step": 701
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.141732283464567e-05,
+      "loss": 0.699,
+      "step": 702
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.1404199475065619e-05,
+      "loss": 0.6788,
+      "step": 703
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.1391076115485564e-05,
+      "loss": 0.6577,
+      "step": 704
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.1377952755905513e-05,
+      "loss": 0.6856,
+      "step": 705
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.136482939632546e-05,
+      "loss": 0.6631,
+      "step": 706
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.1351706036745407e-05,
+      "loss": 0.7377,
+      "step": 707
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.1338582677165354e-05,
+      "loss": 0.7019,
+      "step": 708
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.1325459317585303e-05,
+      "loss": 0.7614,
+      "step": 709
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.131233595800525e-05,
+      "loss": 0.6117,
+      "step": 710
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.1299212598425197e-05,
+      "loss": 0.6598,
+      "step": 711
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.1286089238845146e-05,
+      "loss": 0.6825,
+      "step": 712
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.1272965879265092e-05,
+      "loss": 0.6691,
+      "step": 713
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.125984251968504e-05,
+      "loss": 0.6533,
+      "step": 714
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.124671916010499e-05,
+      "loss": 0.5929,
+      "step": 715
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.1233595800524935e-05,
+      "loss": 0.6919,
+      "step": 716
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.1220472440944883e-05,
+      "loss": 0.7194,
+      "step": 717
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.120734908136483e-05,
+      "loss": 0.6258,
+      "step": 718
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.1194225721784778e-05,
+      "loss": 0.7814,
+      "step": 719
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.1181102362204725e-05,
+      "loss": 0.6551,
+      "step": 720
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.1167979002624674e-05,
+      "loss": 0.6137,
+      "step": 721
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.1154855643044619e-05,
+      "loss": 0.6571,
+      "step": 722
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.1141732283464568e-05,
+      "loss": 0.6935,
+      "step": 723
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.1128608923884517e-05,
+      "loss": 0.6791,
+      "step": 724
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.1115485564304462e-05,
+      "loss": 0.6395,
+      "step": 725
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.1102362204724411e-05,
+      "loss": 0.7327,
+      "step": 726
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.108923884514436e-05,
+      "loss": 0.5885,
+      "step": 727
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.1076115485564305e-05,
+      "loss": 0.728,
+      "step": 728
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.1062992125984254e-05,
+      "loss": 0.652,
+      "step": 729
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.1049868766404201e-05,
+      "loss": 0.7011,
+      "step": 730
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.1036745406824148e-05,
+      "loss": 0.737,
+      "step": 731
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.1023622047244095e-05,
+      "loss": 0.6353,
+      "step": 732
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.1010498687664042e-05,
+      "loss": 0.5945,
+      "step": 733
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.099737532808399e-05,
+      "loss": 0.621,
+      "step": 734
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.0984251968503938e-05,
+      "loss": 0.8395,
+      "step": 735
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.0971128608923884e-05,
+      "loss": 0.6474,
+      "step": 736
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.0958005249343833e-05,
+      "loss": 0.6155,
+      "step": 737
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.0944881889763781e-05,
+      "loss": 0.7122,
+      "step": 738
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.0931758530183727e-05,
+      "loss": 0.6375,
+      "step": 739
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.0918635170603676e-05,
+      "loss": 0.8032,
+      "step": 740
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.0905511811023624e-05,
+      "loss": 0.6403,
+      "step": 741
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.089238845144357e-05,
+      "loss": 0.6304,
+      "step": 742
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.0879265091863519e-05,
+      "loss": 0.6455,
+      "step": 743
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.0866141732283466e-05,
+      "loss": 0.6916,
+      "step": 744
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.0853018372703413e-05,
+      "loss": 0.646,
+      "step": 745
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.083989501312336e-05,
+      "loss": 0.717,
+      "step": 746
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.0826771653543309e-05,
+      "loss": 0.7486,
+      "step": 747
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.0813648293963254e-05,
+      "loss": 0.7672,
+      "step": 748
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.0800524934383203e-05,
+      "loss": 0.6813,
+      "step": 749
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.0787401574803152e-05,
+      "loss": 0.6072,
+      "step": 750
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.0774278215223097e-05,
+      "loss": 0.6769,
+      "step": 751
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.0761154855643046e-05,
+      "loss": 0.7087,
+      "step": 752
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.0748031496062993e-05,
+      "loss": 0.6325,
+      "step": 753
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.073490813648294e-05,
+      "loss": 0.6437,
+      "step": 754
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.0721784776902887e-05,
+      "loss": 0.6623,
+      "step": 755
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.0708661417322836e-05,
+      "loss": 0.6227,
+      "step": 756
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.0695538057742783e-05,
+      "loss": 0.723,
+      "step": 757
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.068241469816273e-05,
+      "loss": 0.729,
+      "step": 758
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.066929133858268e-05,
+      "loss": 0.6285,
+      "step": 759
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.0656167979002625e-05,
+      "loss": 0.8486,
+      "step": 760
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.0643044619422573e-05,
+      "loss": 0.629,
+      "step": 761
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.0629921259842522e-05,
+      "loss": 0.6881,
+      "step": 762
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.0616797900262468e-05,
+      "loss": 0.6187,
+      "step": 763
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.0603674540682417e-05,
+      "loss": 0.7408,
+      "step": 764
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.0590551181102362e-05,
+      "loss": 0.5962,
+      "step": 765
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.057742782152231e-05,
+      "loss": 0.8271,
+      "step": 766
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.0564304461942258e-05,
+      "loss": 0.7126,
+      "step": 767
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.0551181102362205e-05,
+      "loss": 0.6346,
+      "step": 768
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.0538057742782152e-05,
+      "loss": 0.6933,
+      "step": 769
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.0524934383202101e-05,
+      "loss": 0.6603,
+      "step": 770
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.0511811023622048e-05,
+      "loss": 0.7081,
+      "step": 771
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.0498687664041995e-05,
+      "loss": 0.6914,
+      "step": 772
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.0485564304461944e-05,
+      "loss": 0.632,
+      "step": 773
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.047244094488189e-05,
+      "loss": 0.6257,
+      "step": 774
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.0459317585301838e-05,
+      "loss": 0.6747,
+      "step": 775
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.0446194225721787e-05,
+      "loss": 0.7726,
+      "step": 776
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.0433070866141732e-05,
+      "loss": 0.6323,
+      "step": 777
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.0419947506561681e-05,
+      "loss": 0.6529,
+      "step": 778
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.0406824146981628e-05,
+      "loss": 0.6803,
+      "step": 779
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.0393700787401575e-05,
+      "loss": 0.5972,
+      "step": 780
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.0380577427821523e-05,
+      "loss": 0.664,
+      "step": 781
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.0367454068241471e-05,
+      "loss": 0.6321,
+      "step": 782
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.0354330708661417e-05,
+      "loss": 0.673,
+      "step": 783
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.0341207349081366e-05,
+      "loss": 0.6909,
+      "step": 784
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.0328083989501314e-05,
+      "loss": 0.6972,
+      "step": 785
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.031496062992126e-05,
+      "loss": 0.7406,
+      "step": 786
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.0301837270341209e-05,
+      "loss": 0.6397,
+      "step": 787
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.0288713910761157e-05,
+      "loss": 0.6186,
+      "step": 788
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.0275590551181103e-05,
+      "loss": 0.6089,
+      "step": 789
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.0262467191601052e-05,
+      "loss": 0.7009,
+      "step": 790
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.0249343832020999e-05,
+      "loss": 0.6637,
+      "step": 791
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.0236220472440946e-05,
+      "loss": 0.6239,
+      "step": 792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.0223097112860893e-05,
+      "loss": 0.6173,
+      "step": 793
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.0209973753280842e-05,
+      "loss": 0.6577,
+      "step": 794
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.0196850393700787e-05,
+      "loss": 0.6167,
+      "step": 795
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.0183727034120736e-05,
+      "loss": 0.6545,
+      "step": 796
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.0170603674540682e-05,
+      "loss": 0.7052,
+      "step": 797
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.015748031496063e-05,
+      "loss": 0.6266,
+      "step": 798
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.0144356955380579e-05,
+      "loss": 0.6369,
+      "step": 799
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.0131233595800525e-05,
+      "loss": 0.6615,
+      "step": 800
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.0118110236220473e-05,
+      "loss": 0.5865,
+      "step": 801
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.010498687664042e-05,
+      "loss": 0.6904,
+      "step": 802
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.0091863517060368e-05,
+      "loss": 0.6218,
+      "step": 803
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.0078740157480316e-05,
+      "loss": 0.6065,
+      "step": 804
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.0065616797900264e-05,
+      "loss": 0.6221,
+      "step": 805
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.005249343832021e-05,
+      "loss": 0.6232,
+      "step": 806
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.0039370078740158e-05,
+      "loss": 0.6404,
+      "step": 807
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.0026246719160107e-05,
+      "loss": 0.6067,
+      "step": 808
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.0013123359580052e-05,
+      "loss": 0.6303,
+      "step": 809
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5762,
+      "step": 810
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 9.986876640419948e-06,
+      "loss": 0.5743,
+      "step": 811
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 9.973753280839897e-06,
+      "loss": 0.6424,
+      "step": 812
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 9.960629921259844e-06,
+      "loss": 0.6624,
+      "step": 813
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 9.947506561679791e-06,
+      "loss": 0.6268,
+      "step": 814
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 9.934383202099738e-06,
+      "loss": 0.6103,
+      "step": 815
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 9.921259842519685e-06,
+      "loss": 0.6463,
+      "step": 816
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 9.908136482939632e-06,
+      "loss": 0.703,
+      "step": 817
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 9.895013123359581e-06,
+      "loss": 0.5118,
+      "step": 818
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 9.881889763779528e-06,
+      "loss": 0.6595,
+      "step": 819
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 9.868766404199475e-06,
+      "loss": 0.5826,
+      "step": 820
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 9.855643044619422e-06,
+      "loss": 0.6616,
+      "step": 821
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 9.842519685039371e-06,
+      "loss": 0.6775,
+      "step": 822
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 9.829396325459318e-06,
+      "loss": 0.6202,
+      "step": 823
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 9.816272965879266e-06,
+      "loss": 0.607,
+      "step": 824
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 9.803149606299214e-06,
+      "loss": 0.6514,
+      "step": 825
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.790026246719161e-06,
+      "loss": 0.605,
+      "step": 826
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.776902887139109e-06,
+      "loss": 0.6088,
+      "step": 827
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.763779527559056e-06,
+      "loss": 0.5596,
+      "step": 828
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.750656167979003e-06,
+      "loss": 0.5667,
+      "step": 829
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 9.73753280839895e-06,
+      "loss": 0.6606,
+      "step": 830
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 9.724409448818899e-06,
+      "loss": 0.6374,
+      "step": 831
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 9.711286089238846e-06,
+      "loss": 0.628,
+      "step": 832
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 9.698162729658793e-06,
+      "loss": 0.6703,
+      "step": 833
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 9.68503937007874e-06,
+      "loss": 0.6053,
+      "step": 834
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 9.671916010498689e-06,
+      "loss": 0.6029,
+      "step": 835
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 9.658792650918636e-06,
+      "loss": 0.6459,
+      "step": 836
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 9.645669291338583e-06,
+      "loss": 0.6399,
+      "step": 837
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 9.632545931758532e-06,
+      "loss": 0.6433,
+      "step": 838
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 9.619422572178479e-06,
+      "loss": 0.5805,
+      "step": 839
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 9.606299212598426e-06,
+      "loss": 0.6291,
+      "step": 840
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 9.593175853018373e-06,
+      "loss": 0.705,
+      "step": 841
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 9.58005249343832e-06,
+      "loss": 0.6765,
+      "step": 842
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 9.566929133858268e-06,
+      "loss": 0.6457,
+      "step": 843
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 9.553805774278216e-06,
+      "loss": 0.6815,
+      "step": 844
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 9.540682414698163e-06,
+      "loss": 0.6055,
+      "step": 845
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 9.52755905511811e-06,
+      "loss": 0.6402,
+      "step": 846
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 9.51443569553806e-06,
+      "loss": 0.6346,
+      "step": 847
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 9.501312335958006e-06,
+      "loss": 0.6713,
+      "step": 848
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 9.488188976377954e-06,
+      "loss": 0.6157,
+      "step": 849
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 9.4750656167979e-06,
+      "loss": 0.6422,
+      "step": 850
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 9.46194225721785e-06,
+      "loss": 0.5974,
+      "step": 851
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 9.448818897637797e-06,
+      "loss": 0.6071,
+      "step": 852
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 9.435695538057744e-06,
+      "loss": 0.682,
+      "step": 853
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 9.422572178477691e-06,
+      "loss": 0.6037,
+      "step": 854
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 9.409448818897638e-06,
+      "loss": 0.6007,
+      "step": 855
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 9.396325459317585e-06,
+      "loss": 0.7192,
+      "step": 856
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 9.383202099737534e-06,
+      "loss": 0.6368,
+      "step": 857
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 9.370078740157481e-06,
+      "loss": 0.6612,
+      "step": 858
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 9.356955380577428e-06,
+      "loss": 0.6396,
+      "step": 859
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 9.343832020997377e-06,
+      "loss": 0.6173,
+      "step": 860
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 9.330708661417324e-06,
+      "loss": 0.6089,
+      "step": 861
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 9.317585301837271e-06,
+      "loss": 0.6533,
+      "step": 862
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 9.304461942257218e-06,
+      "loss": 0.6849,
+      "step": 863
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 9.291338582677165e-06,
+      "loss": 0.5914,
+      "step": 864
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 9.278215223097114e-06,
+      "loss": 0.703,
+      "step": 865
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 9.265091863517061e-06,
+      "loss": 0.6436,
+      "step": 866
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 9.251968503937008e-06,
+      "loss": 0.658,
+      "step": 867
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 9.238845144356956e-06,
+      "loss": 0.5291,
+      "step": 868
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 9.225721784776903e-06,
+      "loss": 0.5614,
+      "step": 869
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 9.212598425196852e-06,
+      "loss": 0.6571,
+      "step": 870
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 9.199475065616799e-06,
+      "loss": 0.6326,
+      "step": 871
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 9.186351706036746e-06,
+      "loss": 0.6077,
+      "step": 872
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 9.173228346456695e-06,
+      "loss": 0.6227,
+      "step": 873
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 9.160104986876642e-06,
+      "loss": 0.6865,
+      "step": 874
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 9.146981627296589e-06,
+      "loss": 0.6281,
+      "step": 875
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 9.133858267716536e-06,
+      "loss": 0.5794,
+      "step": 876
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 9.120734908136483e-06,
+      "loss": 0.5695,
+      "step": 877
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 9.10761154855643e-06,
+      "loss": 0.686,
+      "step": 878
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 9.094488188976379e-06,
+      "loss": 0.6254,
+      "step": 879
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 9.081364829396326e-06,
+      "loss": 0.6484,
+      "step": 880
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 9.068241469816273e-06,
+      "loss": 0.6519,
+      "step": 881
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 9.05511811023622e-06,
+      "loss": 0.6125,
+      "step": 882
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 9.041994750656169e-06,
+      "loss": 0.6773,
+      "step": 883
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 9.028871391076116e-06,
+      "loss": 0.6738,
+      "step": 884
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 9.015748031496063e-06,
+      "loss": 0.6506,
+      "step": 885
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 9.002624671916012e-06,
+      "loss": 0.6296,
+      "step": 886
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 8.98950131233596e-06,
+      "loss": 0.6392,
+      "step": 887
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 8.976377952755906e-06,
+      "loss": 0.5881,
+      "step": 888
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 8.963254593175854e-06,
+      "loss": 0.6211,
+      "step": 889
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 8.9501312335958e-06,
+      "loss": 0.5231,
+      "step": 890
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 8.937007874015748e-06,
+      "loss": 0.6116,
+      "step": 891
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 8.923884514435697e-06,
+      "loss": 0.6918,
+      "step": 892
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.910761154855644e-06,
+      "loss": 0.6539,
+      "step": 893
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.89763779527559e-06,
+      "loss": 0.6007,
+      "step": 894
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.88451443569554e-06,
+      "loss": 0.6268,
+      "step": 895
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.871391076115487e-06,
+      "loss": 0.6474,
+      "step": 896
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 8.858267716535434e-06,
+      "loss": 0.5751,
+      "step": 897
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 8.845144356955381e-06,
+      "loss": 0.6662,
+      "step": 898
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 8.83202099737533e-06,
+      "loss": 0.6464,
+      "step": 899
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 8.818897637795277e-06,
+      "loss": 0.7425,
+      "step": 900
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 8.805774278215224e-06,
+      "loss": 0.646,
+      "step": 901
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 8.792650918635171e-06,
+      "loss": 0.567,
+      "step": 902
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 8.779527559055118e-06,
+      "loss": 0.6417,
+      "step": 903
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 8.766404199475065e-06,
+      "loss": 0.5776,
+      "step": 904
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 8.753280839895014e-06,
+      "loss": 0.5452,
+      "step": 905
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 8.740157480314961e-06,
+      "loss": 0.624,
+      "step": 906
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 8.727034120734908e-06,
+      "loss": 0.778,
+      "step": 907
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 8.713910761154857e-06,
+      "loss": 0.6186,
+      "step": 908
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 8.700787401574804e-06,
+      "loss": 0.7648,
+      "step": 909
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 8.687664041994751e-06,
+      "loss": 0.6451,
+      "step": 910
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 8.674540682414699e-06,
+      "loss": 0.5695,
+      "step": 911
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 8.661417322834647e-06,
+      "loss": 0.6104,
+      "step": 912
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 8.648293963254594e-06,
+      "loss": 0.6005,
+      "step": 913
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 8.635170603674542e-06,
+      "loss": 0.6166,
+      "step": 914
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 8.622047244094489e-06,
+      "loss": 0.6402,
+      "step": 915
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 8.608923884514436e-06,
+      "loss": 0.5688,
+      "step": 916
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 8.595800524934383e-06,
+      "loss": 0.6487,
+      "step": 917
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 8.582677165354332e-06,
+      "loss": 0.6313,
+      "step": 918
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 8.569553805774279e-06,
+      "loss": 0.632,
+      "step": 919
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 8.556430446194226e-06,
+      "loss": 0.6531,
+      "step": 920
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 8.543307086614175e-06,
+      "loss": 0.6736,
+      "step": 921
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 8.530183727034122e-06,
+      "loss": 0.7554,
+      "step": 922
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 8.517060367454069e-06,
+      "loss": 0.6537,
+      "step": 923
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 8.503937007874016e-06,
+      "loss": 0.6095,
+      "step": 924
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 8.490813648293963e-06,
+      "loss": 0.577,
+      "step": 925
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 8.47769028871391e-06,
+      "loss": 0.643,
+      "step": 926
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 8.46456692913386e-06,
+      "loss": 0.68,
+      "step": 927
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 8.451443569553806e-06,
+      "loss": 0.5881,
+      "step": 928
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 8.438320209973753e-06,
+      "loss": 0.6181,
+      "step": 929
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 8.4251968503937e-06,
+      "loss": 0.6661,
+      "step": 930
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 8.41207349081365e-06,
+      "loss": 0.6526,
+      "step": 931
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 8.398950131233596e-06,
+      "loss": 0.6136,
+      "step": 932
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 8.385826771653544e-06,
+      "loss": 0.6261,
+      "step": 933
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 8.372703412073492e-06,
+      "loss": 0.7581,
+      "step": 934
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 8.35958005249344e-06,
+      "loss": 0.7365,
+      "step": 935
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 8.346456692913387e-06,
+      "loss": 0.543,
+      "step": 936
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.6141,
+      "step": 937
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 8.320209973753281e-06,
+      "loss": 0.6007,
+      "step": 938
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 8.307086614173228e-06,
+      "loss": 0.6645,
+      "step": 939
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 8.293963254593177e-06,
+      "loss": 0.616,
+      "step": 940
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 8.280839895013124e-06,
+      "loss": 0.5948,
+      "step": 941
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 8.267716535433071e-06,
+      "loss": 0.6621,
+      "step": 942
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 8.25459317585302e-06,
+      "loss": 0.5853,
+      "step": 943
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 8.241469816272967e-06,
+      "loss": 0.6448,
+      "step": 944
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 8.228346456692914e-06,
+      "loss": 0.6693,
+      "step": 945
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 8.215223097112861e-06,
+      "loss": 0.7148,
+      "step": 946
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 8.20209973753281e-06,
+      "loss": 0.7296,
+      "step": 947
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 8.188976377952757e-06,
+      "loss": 0.6092,
+      "step": 948
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 8.175853018372704e-06,
+      "loss": 0.6536,
+      "step": 949
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 8.162729658792651e-06,
+      "loss": 0.5782,
+      "step": 950
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 8.149606299212598e-06,
+      "loss": 0.5929,
+      "step": 951
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 8.136482939632546e-06,
+      "loss": 0.6732,
+      "step": 952
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 8.123359580052494e-06,
+      "loss": 0.6452,
+      "step": 953
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 8.110236220472441e-06,
+      "loss": 0.6733,
+      "step": 954
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 8.097112860892389e-06,
+      "loss": 0.5783,
+      "step": 955
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.083989501312337e-06,
+      "loss": 0.6939,
+      "step": 956
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.070866141732285e-06,
+      "loss": 0.649,
+      "step": 957
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.057742782152232e-06,
+      "loss": 0.6461,
+      "step": 958
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.04461942257218e-06,
+      "loss": 0.6792,
+      "step": 959
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 8.031496062992128e-06,
+      "loss": 0.6308,
+      "step": 960
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 8.018372703412075e-06,
+      "loss": 0.6007,
+      "step": 961
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 8.005249343832022e-06,
+      "loss": 0.5804,
+      "step": 962
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 7.992125984251969e-06,
+      "loss": 0.5736,
+      "step": 963
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 7.979002624671916e-06,
+      "loss": 0.705,
+      "step": 964
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 7.965879265091863e-06,
+      "loss": 0.6313,
+      "step": 965
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 7.952755905511812e-06,
+      "loss": 0.8025,
+      "step": 966
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 7.939632545931759e-06,
+      "loss": 0.5595,
+      "step": 967
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 7.926509186351706e-06,
+      "loss": 0.6043,
+      "step": 968
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 7.913385826771655e-06,
+      "loss": 0.7616,
+      "step": 969
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 7.900262467191602e-06,
+      "loss": 0.6214,
+      "step": 970
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 7.88713910761155e-06,
+      "loss": 0.5419,
+      "step": 971
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 7.874015748031496e-06,
+      "loss": 0.6296,
+      "step": 972
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 7.860892388451443e-06,
+      "loss": 0.6272,
+      "step": 973
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 7.847769028871392e-06,
+      "loss": 0.6284,
+      "step": 974
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 7.83464566929134e-06,
+      "loss": 0.6204,
+      "step": 975
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 7.821522309711287e-06,
+      "loss": 0.5949,
+      "step": 976
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 7.808398950131234e-06,
+      "loss": 0.6518,
+      "step": 977
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 7.79527559055118e-06,
+      "loss": 0.5697,
+      "step": 978
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 7.78215223097113e-06,
+      "loss": 0.6188,
+      "step": 979
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 7.769028871391077e-06,
+      "loss": 0.6032,
+      "step": 980
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 7.755905511811024e-06,
+      "loss": 0.6406,
+      "step": 981
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 7.742782152230973e-06,
+      "loss": 0.5665,
+      "step": 982
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 7.72965879265092e-06,
+      "loss": 0.5978,
+      "step": 983
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 7.716535433070867e-06,
+      "loss": 0.5523,
+      "step": 984
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 7.703412073490814e-06,
+      "loss": 0.5657,
+      "step": 985
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 7.690288713910761e-06,
+      "loss": 0.6334,
+      "step": 986
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 7.677165354330708e-06,
+      "loss": 0.645,
+      "step": 987
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 7.664041994750657e-06,
+      "loss": 0.6624,
+      "step": 988
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 7.650918635170604e-06,
+      "loss": 0.6452,
+      "step": 989
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 7.637795275590551e-06,
+      "loss": 0.6135,
+      "step": 990
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 7.6246719160105e-06,
+      "loss": 0.584,
+      "step": 991
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 7.611548556430447e-06,
+      "loss": 0.61,
+      "step": 992
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 7.598425196850394e-06,
+      "loss": 0.5792,
+      "step": 993
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 7.585301837270341e-06,
+      "loss": 0.5922,
+      "step": 994
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.572178477690289e-06,
+      "loss": 0.6651,
+      "step": 995
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.5590551181102365e-06,
+      "loss": 0.7296,
+      "step": 996
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.545931758530184e-06,
+      "loss": 0.5467,
+      "step": 997
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.532808398950132e-06,
+      "loss": 0.7355,
+      "step": 998
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.5196850393700795e-06,
+      "loss": 0.6573,
+      "step": 999
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.506561679790027e-06,
+      "loss": 0.6198,
+      "step": 1000
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.493438320209975e-06,
+      "loss": 0.5829,
+      "step": 1001
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.480314960629922e-06,
+      "loss": 0.6324,
+      "step": 1002
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.467191601049869e-06,
+      "loss": 0.5266,
+      "step": 1003
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.454068241469818e-06,
+      "loss": 0.6182,
+      "step": 1004
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.440944881889765e-06,
+      "loss": 0.6859,
+      "step": 1005
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.427821522309712e-06,
+      "loss": 0.5596,
+      "step": 1006
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.41469816272966e-06,
+      "loss": 0.5993,
+      "step": 1007
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.401574803149607e-06,
+      "loss": 0.6331,
+      "step": 1008
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.388451443569554e-06,
+      "loss": 0.6644,
+      "step": 1009
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.375328083989501e-06,
+      "loss": 0.6473,
+      "step": 1010
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.36220472440945e-06,
+      "loss": 0.6239,
+      "step": 1011
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.349081364829397e-06,
+      "loss": 0.6353,
+      "step": 1012
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.335958005249344e-06,
+      "loss": 0.6204,
+      "step": 1013
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.322834645669292e-06,
+      "loss": 0.5977,
+      "step": 1014
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.309711286089239e-06,
+      "loss": 0.6741,
+      "step": 1015
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.2965879265091864e-06,
+      "loss": 0.6163,
+      "step": 1016
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.283464566929135e-06,
+      "loss": 0.6252,
+      "step": 1017
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.270341207349082e-06,
+      "loss": 0.558,
+      "step": 1018
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 7.2572178477690295e-06,
+      "loss": 0.7629,
+      "step": 1019
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 7.2440944881889774e-06,
+      "loss": 0.6893,
+      "step": 1020
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 7.2309711286089245e-06,
+      "loss": 0.6833,
+      "step": 1021
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 7.217847769028872e-06,
+      "loss": 0.6381,
+      "step": 1022
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 7.20472440944882e-06,
+      "loss": 0.606,
+      "step": 1023
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 7.191601049868768e-06,
+      "loss": 0.6269,
+      "step": 1024
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 7.178477690288715e-06,
+      "loss": 0.6477,
+      "step": 1025
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 7.165354330708662e-06,
+      "loss": 0.6106,
+      "step": 1026
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 7.15223097112861e-06,
+      "loss": 0.5857,
+      "step": 1027
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 7.139107611548557e-06,
+      "loss": 0.707,
+      "step": 1028
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 7.125984251968504e-06,
+      "loss": 0.6282,
+      "step": 1029
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 7.112860892388452e-06,
+      "loss": 0.5898,
+      "step": 1030
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 7.099737532808399e-06,
+      "loss": 0.6308,
+      "step": 1031
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 7.086614173228347e-06,
+      "loss": 0.7467,
+      "step": 1032
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 7.073490813648295e-06,
+      "loss": 0.5981,
+      "step": 1033
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.060367454068242e-06,
+      "loss": 0.7398,
+      "step": 1034
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.047244094488189e-06,
+      "loss": 0.6665,
+      "step": 1035
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.034120734908137e-06,
+      "loss": 0.6329,
+      "step": 1036
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.020997375328084e-06,
+      "loss": 0.604,
+      "step": 1037
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.0078740157480315e-06,
+      "loss": 0.5794,
+      "step": 1038
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.99475065616798e-06,
+      "loss": 0.7249,
+      "step": 1039
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.981627296587927e-06,
+      "loss": 0.5788,
+      "step": 1040
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.9685039370078745e-06,
+      "loss": 0.6256,
+      "step": 1041
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.955380577427822e-06,
+      "loss": 0.6902,
+      "step": 1042
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.94225721784777e-06,
+      "loss": 0.6513,
+      "step": 1043
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.929133858267717e-06,
+      "loss": 0.6074,
+      "step": 1044
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.916010498687664e-06,
+      "loss": 0.627,
+      "step": 1045
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.902887139107613e-06,
+      "loss": 0.6453,
+      "step": 1046
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.88976377952756e-06,
+      "loss": 0.6491,
+      "step": 1047
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.876640419947507e-06,
+      "loss": 0.6324,
+      "step": 1048
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.863517060367455e-06,
+      "loss": 0.6118,
+      "step": 1049
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.850393700787402e-06,
+      "loss": 0.6043,
+      "step": 1050
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.837270341207349e-06,
+      "loss": 0.6692,
+      "step": 1051
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.824146981627298e-06,
+      "loss": 0.649,
+      "step": 1052
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.811023622047245e-06,
+      "loss": 0.626,
+      "step": 1053
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.797900262467192e-06,
+      "loss": 0.6614,
+      "step": 1054
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.78477690288714e-06,
+      "loss": 0.5828,
+      "step": 1055
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.771653543307087e-06,
+      "loss": 0.6297,
+      "step": 1056
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.758530183727034e-06,
+      "loss": 0.6677,
+      "step": 1057
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 6.745406824146981e-06,
+      "loss": 0.5872,
+      "step": 1058
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 6.73228346456693e-06,
+      "loss": 0.5833,
+      "step": 1059
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 6.719160104986877e-06,
+      "loss": 0.5769,
+      "step": 1060
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 6.7060367454068245e-06,
+      "loss": 0.5524,
+      "step": 1061
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 6.692913385826772e-06,
+      "loss": 0.6334,
+      "step": 1062
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 6.6797900262467195e-06,
+      "loss": 0.6272,
+      "step": 1063
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.5321,
+      "step": 1064
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 6.6535433070866155e-06,
+      "loss": 0.6737,
+      "step": 1065
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 6.6404199475065626e-06,
+      "loss": 0.6599,
+      "step": 1066
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 6.62729658792651e-06,
+      "loss": 0.5876,
+      "step": 1067
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 6.614173228346458e-06,
+      "loss": 0.6939,
+      "step": 1068
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 6.601049868766405e-06,
+      "loss": 0.5652,
+      "step": 1069
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 6.587926509186352e-06,
+      "loss": 0.6255,
+      "step": 1070
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 6.574803149606301e-06,
+      "loss": 0.5783,
+      "step": 1071
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 6.561679790026248e-06,
+      "loss": 0.6066,
+      "step": 1072
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 6.548556430446195e-06,
+      "loss": 0.6528,
+      "step": 1073
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 6.535433070866142e-06,
+      "loss": 0.5878,
+      "step": 1074
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 6.52230971128609e-06,
+      "loss": 0.612,
+      "step": 1075
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 6.509186351706037e-06,
+      "loss": 0.6516,
+      "step": 1076
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 6.496062992125984e-06,
+      "loss": 0.6555,
+      "step": 1077
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 6.482939632545932e-06,
+      "loss": 0.6241,
+      "step": 1078
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 6.46981627296588e-06,
+      "loss": 0.5879,
+      "step": 1079
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 6.456692913385827e-06,
+      "loss": 0.5961,
+      "step": 1080
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 6.443569553805775e-06,
+      "loss": 0.63,
+      "step": 1081
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 6.430446194225722e-06,
+      "loss": 0.5751,
+      "step": 1082
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 6.4173228346456695e-06,
+      "loss": 0.6776,
+      "step": 1083
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 6.4041994750656174e-06,
+      "loss": 0.6335,
+      "step": 1084
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 6.3910761154855646e-06,
+      "loss": 0.6898,
+      "step": 1085
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 6.3779527559055125e-06,
+      "loss": 0.6258,
+      "step": 1086
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 6.3648293963254605e-06,
+      "loss": 0.7041,
+      "step": 1087
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 6.351706036745408e-06,
+      "loss": 0.6841,
+      "step": 1088
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 6.338582677165355e-06,
+      "loss": 0.6457,
+      "step": 1089
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 6.325459317585302e-06,
+      "loss": 0.6021,
+      "step": 1090
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 6.31233595800525e-06,
+      "loss": 0.7876,
+      "step": 1091
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 6.299212598425197e-06,
+      "loss": 0.6511,
+      "step": 1092
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 6.286089238845144e-06,
+      "loss": 0.5693,
+      "step": 1093
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 6.272965879265093e-06,
+      "loss": 0.6689,
+      "step": 1094
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 6.25984251968504e-06,
+      "loss": 0.5794,
+      "step": 1095
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 6.246719160104987e-06,
+      "loss": 0.6095,
+      "step": 1096
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 6.233595800524935e-06,
+      "loss": 0.6198,
+      "step": 1097
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 6.220472440944882e-06,
+      "loss": 0.728,
+      "step": 1098
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 6.207349081364829e-06,
+      "loss": 0.6237,
+      "step": 1099
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 6.194225721784778e-06,
+      "loss": 0.5208,
+      "step": 1100
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 6.181102362204725e-06,
+      "loss": 0.6644,
+      "step": 1101
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 6.167979002624672e-06,
+      "loss": 0.6237,
+      "step": 1102
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 6.15485564304462e-06,
+      "loss": 0.6408,
+      "step": 1103
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 6.141732283464567e-06,
+      "loss": 0.5923,
+      "step": 1104
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 6.1286089238845145e-06,
+      "loss": 0.5986,
+      "step": 1105
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 6.115485564304462e-06,
+      "loss": 0.6119,
+      "step": 1106
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 6.1023622047244104e-06,
+      "loss": 0.5958,
+      "step": 1107
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 6.0892388451443576e-06,
+      "loss": 0.5795,
+      "step": 1108
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 6.076115485564305e-06,
+      "loss": 0.5703,
+      "step": 1109
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 6.062992125984253e-06,
+      "loss": 0.5626,
+      "step": 1110
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 6.0498687664042e-06,
+      "loss": 0.5996,
+      "step": 1111
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 6.036745406824147e-06,
+      "loss": 0.6301,
+      "step": 1112
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 6.023622047244096e-06,
+      "loss": 0.6289,
+      "step": 1113
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 6.010498687664043e-06,
+      "loss": 0.5775,
+      "step": 1114
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5.99737532808399e-06,
+      "loss": 0.7325,
+      "step": 1115
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 5.984251968503938e-06,
+      "loss": 0.6303,
+      "step": 1116
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5.971128608923885e-06,
+      "loss": 0.6087,
+      "step": 1117
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5.958005249343832e-06,
+      "loss": 0.6375,
+      "step": 1118
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5.944881889763781e-06,
+      "loss": 0.6599,
+      "step": 1119
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 5.931758530183728e-06,
+      "loss": 0.6024,
+      "step": 1120
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5.918635170603675e-06,
+      "loss": 0.6323,
+      "step": 1121
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5.905511811023622e-06,
+      "loss": 0.6376,
+      "step": 1122
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5.89238845144357e-06,
+      "loss": 0.6958,
+      "step": 1123
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5.879265091863517e-06,
+      "loss": 0.6853,
+      "step": 1124
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.8661417322834645e-06,
+      "loss": 0.6357,
+      "step": 1125
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.853018372703413e-06,
+      "loss": 0.6529,
+      "step": 1126
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.83989501312336e-06,
+      "loss": 0.6125,
+      "step": 1127
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.8267716535433075e-06,
+      "loss": 0.648,
+      "step": 1128
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5.8136482939632555e-06,
+      "loss": 0.655,
+      "step": 1129
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5.800524934383203e-06,
+      "loss": 0.5752,
+      "step": 1130
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5.78740157480315e-06,
+      "loss": 0.7096,
+      "step": 1131
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 5.774278215223098e-06,
+      "loss": 0.6383,
+      "step": 1132
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5.761154855643046e-06,
+      "loss": 0.731,
+      "step": 1133
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5.748031496062993e-06,
+      "loss": 0.6733,
+      "step": 1134
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5.734908136482941e-06,
+      "loss": 0.6333,
+      "step": 1135
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 5.721784776902888e-06,
+      "loss": 0.7259,
+      "step": 1136
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5.708661417322835e-06,
+      "loss": 0.6367,
+      "step": 1137
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5.695538057742782e-06,
+      "loss": 0.7418,
+      "step": 1138
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5.68241469816273e-06,
+      "loss": 0.6129,
+      "step": 1139
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5.669291338582677e-06,
+      "loss": 0.646,
+      "step": 1140
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.656167979002625e-06,
+      "loss": 0.6234,
+      "step": 1141
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.643044619422573e-06,
+      "loss": 0.6498,
+      "step": 1142
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.62992125984252e-06,
+      "loss": 0.5893,
+      "step": 1143
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.616797900262467e-06,
+      "loss": 0.6116,
+      "step": 1144
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.603674540682415e-06,
+      "loss": 0.6926,
+      "step": 1145
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.590551181102362e-06,
+      "loss": 0.5774,
+      "step": 1146
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.5774278215223095e-06,
+      "loss": 0.5413,
+      "step": 1147
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.564304461942258e-06,
+      "loss": 0.6365,
+      "step": 1148
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.5511811023622054e-06,
+      "loss": 0.6671,
+      "step": 1149
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.5380577427821525e-06,
+      "loss": 0.5876,
+      "step": 1150
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.5249343832021005e-06,
+      "loss": 0.6142,
+      "step": 1151
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 5.511811023622048e-06,
+      "loss": 0.6116,
+      "step": 1152
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5.498687664041995e-06,
+      "loss": 0.652,
+      "step": 1153
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5.485564304461942e-06,
+      "loss": 0.5818,
+      "step": 1154
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5.472440944881891e-06,
+      "loss": 0.6518,
+      "step": 1155
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 5.459317585301838e-06,
+      "loss": 0.6722,
+      "step": 1156
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5.446194225721785e-06,
+      "loss": 0.6402,
+      "step": 1157
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5.433070866141733e-06,
+      "loss": 0.5531,
+      "step": 1158
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 5.41994750656168e-06,
+      "loss": 0.6255,
+      "step": 1159
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5.406824146981627e-06,
+      "loss": 0.6685,
+      "step": 1160
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5.393700787401576e-06,
+      "loss": 0.621,
+      "step": 1161
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5.380577427821523e-06,
+      "loss": 0.6379,
+      "step": 1162
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5.36745406824147e-06,
+      "loss": 0.606,
+      "step": 1163
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.354330708661418e-06,
+      "loss": 0.6268,
+      "step": 1164
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.341207349081365e-06,
+      "loss": 0.7007,
+      "step": 1165
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.328083989501312e-06,
+      "loss": 0.6645,
+      "step": 1166
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.314960629921261e-06,
+      "loss": 0.6981,
+      "step": 1167
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.301837270341208e-06,
+      "loss": 0.5873,
+      "step": 1168
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.288713910761155e-06,
+      "loss": 0.6321,
+      "step": 1169
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.2755905511811025e-06,
+      "loss": 0.6274,
+      "step": 1170
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.2624671916010505e-06,
+      "loss": 0.6425,
+      "step": 1171
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5.2493438320209976e-06,
+      "loss": 0.6479,
+      "step": 1172
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5.236220472440945e-06,
+      "loss": 0.6338,
+      "step": 1173
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5.2230971128608935e-06,
+      "loss": 0.6026,
+      "step": 1174
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 5.209973753280841e-06,
+      "loss": 0.6891,
+      "step": 1175
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.196850393700788e-06,
+      "loss": 0.6816,
+      "step": 1176
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.183727034120736e-06,
+      "loss": 0.5745,
+      "step": 1177
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.170603674540683e-06,
+      "loss": 0.621,
+      "step": 1178
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.15748031496063e-06,
+      "loss": 0.5757,
+      "step": 1179
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5.144356955380579e-06,
+      "loss": 0.6507,
+      "step": 1180
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5.131233595800526e-06,
+      "loss": 0.6384,
+      "step": 1181
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5.118110236220473e-06,
+      "loss": 0.5381,
+      "step": 1182
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 5.104986876640421e-06,
+      "loss": 0.6693,
+      "step": 1183
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5.091863517060368e-06,
+      "loss": 0.6092,
+      "step": 1184
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5.078740157480315e-06,
+      "loss": 0.5909,
+      "step": 1185
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5.065616797900262e-06,
+      "loss": 0.5797,
+      "step": 1186
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 5.05249343832021e-06,
+      "loss": 0.5953,
+      "step": 1187
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5.039370078740158e-06,
+      "loss": 0.6111,
+      "step": 1188
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5.026246719160105e-06,
+      "loss": 0.6314,
+      "step": 1189
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5.013123359580053e-06,
+      "loss": 0.5515,
+      "step": 1190
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 5e-06,
+      "loss": 0.5821,
+      "step": 1191
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 4.986876640419948e-06,
+      "loss": 0.5963,
+      "step": 1192
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 4.9737532808398955e-06,
+      "loss": 0.5521,
+      "step": 1193
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 4.960629921259843e-06,
+      "loss": 0.6068,
+      "step": 1194
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 4.9475065616797906e-06,
+      "loss": 0.5339,
+      "step": 1195
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 4.934383202099738e-06,
+      "loss": 0.5205,
+      "step": 1196
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 4.921259842519686e-06,
+      "loss": 0.5731,
+      "step": 1197
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 4.908136482939633e-06,
+      "loss": 0.5512,
+      "step": 1198
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 4.895013123359581e-06,
+      "loss": 0.6023,
+      "step": 1199
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 4.881889763779528e-06,
+      "loss": 0.635,
+      "step": 1200
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 4.868766404199475e-06,
+      "loss": 0.6287,
+      "step": 1201
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 4.855643044619423e-06,
+      "loss": 0.6664,
+      "step": 1202
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 4.84251968503937e-06,
+      "loss": 0.6193,
+      "step": 1203
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 4.829396325459318e-06,
+      "loss": 0.7184,
+      "step": 1204
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 4.816272965879266e-06,
+      "loss": 0.6296,
+      "step": 1205
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 4.803149606299213e-06,
+      "loss": 0.7193,
+      "step": 1206
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 4.79002624671916e-06,
+      "loss": 0.571,
+      "step": 1207
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 4.776902887139108e-06,
+      "loss": 0.6051,
+      "step": 1208
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 4.763779527559055e-06,
+      "loss": 0.5848,
+      "step": 1209
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 4.750656167979003e-06,
+      "loss": 0.6236,
+      "step": 1210
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 4.73753280839895e-06,
+      "loss": 0.6097,
+      "step": 1211
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 4.724409448818898e-06,
+      "loss": 0.6099,
+      "step": 1212
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 4.7112860892388454e-06,
+      "loss": 0.578,
+      "step": 1213
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 4.6981627296587926e-06,
+      "loss": 0.6556,
+      "step": 1214
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 4.6850393700787405e-06,
+      "loss": 0.6387,
+      "step": 1215
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 4.6719160104986885e-06,
+      "loss": 0.6708,
+      "step": 1216
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 4.658792650918636e-06,
+      "loss": 0.5391,
+      "step": 1217
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 4.645669291338583e-06,
+      "loss": 0.6061,
+      "step": 1218
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 4.632545931758531e-06,
+      "loss": 0.5698,
+      "step": 1219
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 4.619422572178478e-06,
+      "loss": 0.5522,
+      "step": 1220
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 4.606299212598426e-06,
+      "loss": 0.6396,
+      "step": 1221
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 4.593175853018373e-06,
+      "loss": 0.7344,
+      "step": 1222
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 4.580052493438321e-06,
+      "loss": 0.5797,
+      "step": 1223
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 4.566929133858268e-06,
+      "loss": 0.6183,
+      "step": 1224
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 4.553805774278215e-06,
+      "loss": 0.5866,
+      "step": 1225
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 4.540682414698163e-06,
+      "loss": 0.6462,
+      "step": 1226
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 4.52755905511811e-06,
+      "loss": 0.5587,
+      "step": 1227
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 4.514435695538058e-06,
+      "loss": 0.6201,
+      "step": 1228
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 4.501312335958006e-06,
+      "loss": 0.6224,
+      "step": 1229
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 4.488188976377953e-06,
+      "loss": 0.5492,
+      "step": 1230
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 4.4750656167979e-06,
+      "loss": 0.6344,
+      "step": 1231
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 4.461942257217848e-06,
+      "loss": 0.6005,
+      "step": 1232
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 4.448818897637795e-06,
+      "loss": 0.6069,
+      "step": 1233
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 4.435695538057743e-06,
+      "loss": 0.6872,
+      "step": 1234
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 4.4225721784776905e-06,
+      "loss": 0.6643,
+      "step": 1235
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 4.4094488188976384e-06,
+      "loss": 0.5951,
+      "step": 1236
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 4.3963254593175856e-06,
+      "loss": 0.6105,
+      "step": 1237
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 4.383202099737533e-06,
+      "loss": 0.5205,
+      "step": 1238
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 4.370078740157481e-06,
+      "loss": 0.6237,
+      "step": 1239
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 4.356955380577429e-06,
+      "loss": 0.5476,
+      "step": 1240
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 4.343832020997376e-06,
+      "loss": 0.7175,
+      "step": 1241
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 4.330708661417324e-06,
+      "loss": 0.6419,
+      "step": 1242
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 4.317585301837271e-06,
+      "loss": 0.6084,
+      "step": 1243
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 4.304461942257218e-06,
+      "loss": 0.5893,
+      "step": 1244
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 4.291338582677166e-06,
+      "loss": 0.5489,
+      "step": 1245
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 4.278215223097113e-06,
+      "loss": 0.6424,
+      "step": 1246
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 4.265091863517061e-06,
+      "loss": 0.6146,
+      "step": 1247
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 4.251968503937008e-06,
+      "loss": 0.6359,
+      "step": 1248
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 4.238845144356955e-06,
+      "loss": 0.6106,
+      "step": 1249
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 4.225721784776903e-06,
+      "loss": 0.6153,
+      "step": 1250
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 4.21259842519685e-06,
+      "loss": 0.5912,
+      "step": 1251
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 4.199475065616798e-06,
+      "loss": 0.6647,
+      "step": 1252
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 4.186351706036746e-06,
+      "loss": 0.6085,
+      "step": 1253
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 4.173228346456693e-06,
+      "loss": 0.528,
+      "step": 1254
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 4.1601049868766404e-06,
+      "loss": 0.5937,
+      "step": 1255
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 4.146981627296588e-06,
+      "loss": 0.605,
+      "step": 1256
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 4.1338582677165355e-06,
+      "loss": 0.5645,
+      "step": 1257
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 4.1207349081364835e-06,
+      "loss": 0.6515,
+      "step": 1258
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 4.107611548556431e-06,
+      "loss": 0.582,
+      "step": 1259
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 4.0944881889763785e-06,
+      "loss": 0.6275,
+      "step": 1260
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 4.081364829396326e-06,
+      "loss": 0.5371,
+      "step": 1261
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 4.068241469816273e-06,
+      "loss": 0.6456,
+      "step": 1262
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 4.055118110236221e-06,
+      "loss": 0.6005,
+      "step": 1263
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 4.041994750656169e-06,
+      "loss": 0.5733,
+      "step": 1264
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 4.028871391076116e-06,
+      "loss": 0.6131,
+      "step": 1265
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 4.015748031496064e-06,
+      "loss": 0.639,
+      "step": 1266
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 4.002624671916011e-06,
+      "loss": 0.583,
+      "step": 1267
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 3.989501312335958e-06,
+      "loss": 0.6704,
+      "step": 1268
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 3.976377952755906e-06,
+      "loss": 0.6779,
+      "step": 1269
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 3.963254593175853e-06,
+      "loss": 0.5783,
+      "step": 1270
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 3.950131233595801e-06,
+      "loss": 0.545,
+      "step": 1271
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 3.937007874015748e-06,
+      "loss": 0.5147,
+      "step": 1272
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 3.923884514435696e-06,
+      "loss": 0.667,
+      "step": 1273
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 3.910761154855643e-06,
+      "loss": 0.6262,
+      "step": 1274
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 3.89763779527559e-06,
+      "loss": 0.6188,
+      "step": 1275
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 3.884514435695538e-06,
+      "loss": 0.6738,
+      "step": 1276
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 3.871391076115486e-06,
+      "loss": 0.5153,
+      "step": 1277
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 3.858267716535433e-06,
+      "loss": 0.5038,
+      "step": 1278
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 3.8451443569553805e-06,
+      "loss": 0.5774,
+      "step": 1279
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 3.8320209973753285e-06,
+      "loss": 0.6238,
+      "step": 1280
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 3.818897637795276e-06,
+      "loss": 0.5444,
+      "step": 1281
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 3.8057742782152236e-06,
+      "loss": 0.6276,
+      "step": 1282
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 3.7926509186351707e-06,
+      "loss": 0.6282,
+      "step": 1283
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 3.7795275590551182e-06,
+      "loss": 0.6081,
+      "step": 1284
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 3.766404199475066e-06,
+      "loss": 0.6392,
+      "step": 1285
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 3.7532808398950133e-06,
+      "loss": 0.579,
+      "step": 1286
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 3.740157480314961e-06,
+      "loss": 0.698,
+      "step": 1287
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 3.727034120734909e-06,
+      "loss": 0.5969,
+      "step": 1288
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 3.713910761154856e-06,
+      "loss": 0.6102,
+      "step": 1289
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 3.7007874015748035e-06,
+      "loss": 0.5752,
+      "step": 1290
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 3.6876640419947506e-06,
+      "loss": 0.6185,
+      "step": 1291
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 3.6745406824146986e-06,
+      "loss": 0.6071,
+      "step": 1292
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 3.661417322834646e-06,
+      "loss": 0.576,
+      "step": 1293
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 3.6482939632545932e-06,
+      "loss": 0.6004,
+      "step": 1294
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 3.635170603674541e-06,
+      "loss": 0.5846,
+      "step": 1295
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 3.6220472440944887e-06,
+      "loss": 0.6752,
+      "step": 1296
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 3.608923884514436e-06,
+      "loss": 0.5583,
+      "step": 1297
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 3.595800524934384e-06,
+      "loss": 0.6503,
+      "step": 1298
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 3.582677165354331e-06,
+      "loss": 0.6422,
+      "step": 1299
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 3.5695538057742785e-06,
+      "loss": 0.6243,
+      "step": 1300
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 3.556430446194226e-06,
+      "loss": 0.6539,
+      "step": 1301
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 3.5433070866141735e-06,
+      "loss": 0.62,
+      "step": 1302
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 3.530183727034121e-06,
+      "loss": 0.5681,
+      "step": 1303
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 3.5170603674540686e-06,
+      "loss": 0.5639,
+      "step": 1304
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 3.5039370078740157e-06,
+      "loss": 0.6487,
+      "step": 1305
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 3.4908136482939637e-06,
+      "loss": 0.5617,
+      "step": 1306
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 3.477690288713911e-06,
+      "loss": 0.5928,
+      "step": 1307
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 3.4645669291338583e-06,
+      "loss": 0.6927,
+      "step": 1308
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 3.4514435695538063e-06,
+      "loss": 0.5844,
+      "step": 1309
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.4383202099737534e-06,
+      "loss": 0.5415,
+      "step": 1310
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.425196850393701e-06,
+      "loss": 0.5315,
+      "step": 1311
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.412073490813649e-06,
+      "loss": 0.5704,
+      "step": 1312
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.398950131233596e-06,
+      "loss": 0.515,
+      "step": 1313
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 3.3858267716535436e-06,
+      "loss": 0.6536,
+      "step": 1314
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 3.3727034120734907e-06,
+      "loss": 0.5739,
+      "step": 1315
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 3.3595800524934387e-06,
+      "loss": 0.5631,
+      "step": 1316
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 3.346456692913386e-06,
+      "loss": 0.5802,
+      "step": 1317
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.6491,
+      "step": 1318
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 3.3202099737532813e-06,
+      "loss": 0.5736,
+      "step": 1319
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 3.307086614173229e-06,
+      "loss": 0.6119,
+      "step": 1320
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 3.293963254593176e-06,
+      "loss": 0.6322,
+      "step": 1321
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 3.280839895013124e-06,
+      "loss": 0.5798,
+      "step": 1322
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 3.267716535433071e-06,
+      "loss": 0.565,
+      "step": 1323
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 3.2545931758530186e-06,
+      "loss": 0.5726,
+      "step": 1324
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 3.241469816272966e-06,
+      "loss": 0.6044,
+      "step": 1325
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 3.2283464566929136e-06,
+      "loss": 0.5637,
+      "step": 1326
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 3.215223097112861e-06,
+      "loss": 0.5935,
+      "step": 1327
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 3.2020997375328087e-06,
+      "loss": 0.5824,
+      "step": 1328
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 3.1889763779527563e-06,
+      "loss": 0.594,
+      "step": 1329
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 3.175853018372704e-06,
+      "loss": 0.5998,
+      "step": 1330
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 3.162729658792651e-06,
+      "loss": 0.5911,
+      "step": 1331
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 3.1496062992125985e-06,
+      "loss": 0.6149,
+      "step": 1332
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 3.1364829396325464e-06,
+      "loss": 0.6125,
+      "step": 1333
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 3.1233595800524935e-06,
+      "loss": 0.7193,
+      "step": 1334
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 3.110236220472441e-06,
+      "loss": 0.57,
+      "step": 1335
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 3.097112860892389e-06,
+      "loss": 0.625,
+      "step": 1336
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 3.083989501312336e-06,
+      "loss": 0.6525,
+      "step": 1337
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 3.0708661417322837e-06,
+      "loss": 0.5988,
+      "step": 1338
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 3.057742782152231e-06,
+      "loss": 0.6858,
+      "step": 1339
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 3.0446194225721788e-06,
+      "loss": 0.6876,
+      "step": 1340
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 3.0314960629921263e-06,
+      "loss": 0.5405,
+      "step": 1341
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 3.0183727034120734e-06,
+      "loss": 0.5811,
+      "step": 1342
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 3.0052493438320214e-06,
+      "loss": 0.6261,
+      "step": 1343
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 2.992125984251969e-06,
+      "loss": 0.6444,
+      "step": 1344
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 2.979002624671916e-06,
+      "loss": 0.6008,
+      "step": 1345
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 2.965879265091864e-06,
+      "loss": 0.6031,
+      "step": 1346
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 2.952755905511811e-06,
+      "loss": 0.636,
+      "step": 1347
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 2.9396325459317587e-06,
+      "loss": 0.5578,
+      "step": 1348
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 2.9265091863517066e-06,
+      "loss": 0.6241,
+      "step": 1349
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 2.9133858267716538e-06,
+      "loss": 0.6523,
+      "step": 1350
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 2.9002624671916013e-06,
+      "loss": 0.6231,
+      "step": 1351
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 2.887139107611549e-06,
+      "loss": 0.6905,
+      "step": 1352
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 2.8740157480314964e-06,
+      "loss": 0.5796,
+      "step": 1353
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 2.860892388451444e-06,
+      "loss": 0.6134,
+      "step": 1354
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 2.847769028871391e-06,
+      "loss": 0.6189,
+      "step": 1355
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 2.8346456692913386e-06,
+      "loss": 0.6193,
+      "step": 1356
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 2.8215223097112865e-06,
+      "loss": 0.5601,
+      "step": 1357
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 2.8083989501312337e-06,
+      "loss": 0.5667,
+      "step": 1358
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 2.795275590551181e-06,
+      "loss": 0.6043,
+      "step": 1359
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 2.782152230971129e-06,
+      "loss": 0.6244,
+      "step": 1360
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 2.7690288713910763e-06,
+      "loss": 0.6286,
+      "step": 1361
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 2.755905511811024e-06,
+      "loss": 0.6662,
+      "step": 1362
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 2.742782152230971e-06,
+      "loss": 0.5494,
+      "step": 1363
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 2.729658792650919e-06,
+      "loss": 0.688,
+      "step": 1364
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 2.7165354330708664e-06,
+      "loss": 0.5471,
+      "step": 1365
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 2.7034120734908135e-06,
+      "loss": 0.5926,
+      "step": 1366
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 2.6902887139107615e-06,
+      "loss": 0.6642,
+      "step": 1367
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 2.677165354330709e-06,
+      "loss": 0.5687,
+      "step": 1368
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 2.664041994750656e-06,
+      "loss": 0.6278,
+      "step": 1369
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 2.650918635170604e-06,
+      "loss": 0.6126,
+      "step": 1370
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 2.6377952755905512e-06,
+      "loss": 0.5179,
+      "step": 1371
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 2.6246719160104988e-06,
+      "loss": 0.579,
+      "step": 1372
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 2.6115485564304468e-06,
+      "loss": 0.6222,
+      "step": 1373
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 2.598425196850394e-06,
+      "loss": 0.645,
+      "step": 1374
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 2.5853018372703414e-06,
+      "loss": 0.6468,
+      "step": 1375
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 2.5721784776902894e-06,
+      "loss": 0.6632,
+      "step": 1376
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 2.5590551181102365e-06,
+      "loss": 0.6044,
+      "step": 1377
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 2.545931758530184e-06,
+      "loss": 0.6571,
+      "step": 1378
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 2.532808398950131e-06,
+      "loss": 0.5599,
+      "step": 1379
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 2.519685039370079e-06,
+      "loss": 0.6641,
+      "step": 1380
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 2.5065616797900266e-06,
+      "loss": 0.5726,
+      "step": 1381
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 2.493438320209974e-06,
+      "loss": 0.5685,
+      "step": 1382
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 2.4803149606299213e-06,
+      "loss": 0.5564,
+      "step": 1383
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 2.467191601049869e-06,
+      "loss": 0.6384,
+      "step": 1384
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 2.4540682414698164e-06,
+      "loss": 0.6037,
+      "step": 1385
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 2.440944881889764e-06,
+      "loss": 0.5677,
+      "step": 1386
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 2.4278215223097115e-06,
+      "loss": 0.6128,
+      "step": 1387
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 2.414698162729659e-06,
+      "loss": 0.626,
+      "step": 1388
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 2.4015748031496065e-06,
+      "loss": 0.5141,
+      "step": 1389
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 2.388451443569554e-06,
+      "loss": 0.6525,
+      "step": 1390
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 2.3753280839895016e-06,
+      "loss": 0.5761,
+      "step": 1391
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 2.362204724409449e-06,
+      "loss": 0.6983,
+      "step": 1392
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 2.3490813648293963e-06,
+      "loss": 0.5376,
+      "step": 1393
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 2.3359580052493442e-06,
+      "loss": 0.6287,
+      "step": 1394
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 2.3228346456692914e-06,
+      "loss": 0.6647,
+      "step": 1395
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 2.309711286089239e-06,
+      "loss": 0.7124,
+      "step": 1396
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 2.2965879265091864e-06,
+      "loss": 0.6219,
+      "step": 1397
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 2.283464566929134e-06,
+      "loss": 0.641,
+      "step": 1398
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 2.2703412073490815e-06,
+      "loss": 0.5652,
+      "step": 1399
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 2.257217847769029e-06,
+      "loss": 0.5909,
+      "step": 1400
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 2.2440944881889766e-06,
+      "loss": 0.6209,
+      "step": 1401
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 2.230971128608924e-06,
+      "loss": 0.6429,
+      "step": 1402
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 2.2178477690288717e-06,
+      "loss": 0.6605,
+      "step": 1403
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 2.2047244094488192e-06,
+      "loss": 0.7196,
+      "step": 1404
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 2.1916010498687663e-06,
+      "loss": 0.637,
+      "step": 1405
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 2.1784776902887143e-06,
+      "loss": 0.6716,
+      "step": 1406
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 2.165354330708662e-06,
+      "loss": 0.5949,
+      "step": 1407
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 2.152230971128609e-06,
+      "loss": 0.571,
+      "step": 1408
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 2.1391076115485565e-06,
+      "loss": 0.6095,
+      "step": 1409
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 2.125984251968504e-06,
+      "loss": 0.6222,
+      "step": 1410
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 2.1128608923884516e-06,
+      "loss": 0.5718,
+      "step": 1411
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 2.099737532808399e-06,
+      "loss": 0.6395,
+      "step": 1412
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 2.0866141732283467e-06,
+      "loss": 0.5932,
+      "step": 1413
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 2.073490813648294e-06,
+      "loss": 0.5273,
+      "step": 1414
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 2.0603674540682417e-06,
+      "loss": 0.5639,
+      "step": 1415
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 2.0472440944881893e-06,
+      "loss": 0.6417,
+      "step": 1416
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 2.0341207349081364e-06,
+      "loss": 0.5556,
+      "step": 1417
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 2.0209973753280844e-06,
+      "loss": 0.5647,
+      "step": 1418
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 2.007874015748032e-06,
+      "loss": 0.5799,
+      "step": 1419
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1.994750656167979e-06,
+      "loss": 0.5283,
+      "step": 1420
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1.9816272965879266e-06,
+      "loss": 0.5321,
+      "step": 1421
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1.968503937007874e-06,
+      "loss": 0.6185,
+      "step": 1422
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1.9553805774278216e-06,
+      "loss": 0.5786,
+      "step": 1423
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.942257217847769e-06,
+      "loss": 0.5717,
+      "step": 1424
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.9291338582677167e-06,
+      "loss": 0.6168,
+      "step": 1425
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.9160104986876642e-06,
+      "loss": 0.5913,
+      "step": 1426
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.9028871391076118e-06,
+      "loss": 0.6777,
+      "step": 1427
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1.8897637795275591e-06,
+      "loss": 0.6172,
+      "step": 1428
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1.8766404199475067e-06,
+      "loss": 0.5928,
+      "step": 1429
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1.8635170603674544e-06,
+      "loss": 0.6762,
+      "step": 1430
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1.8503937007874017e-06,
+      "loss": 0.6074,
+      "step": 1431
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.8372703412073493e-06,
+      "loss": 0.5603,
+      "step": 1432
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.8241469816272966e-06,
+      "loss": 0.5717,
+      "step": 1433
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.8110236220472444e-06,
+      "loss": 0.643,
+      "step": 1434
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.797900262467192e-06,
+      "loss": 0.6208,
+      "step": 1435
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1.7847769028871392e-06,
+      "loss": 0.621,
+      "step": 1436
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1.7716535433070868e-06,
+      "loss": 0.5176,
+      "step": 1437
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1.7585301837270343e-06,
+      "loss": 0.6123,
+      "step": 1438
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1.7454068241469818e-06,
+      "loss": 0.6051,
+      "step": 1439
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.7322834645669292e-06,
+      "loss": 0.6126,
+      "step": 1440
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.7191601049868767e-06,
+      "loss": 0.6502,
+      "step": 1441
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.7060367454068245e-06,
+      "loss": 0.5842,
+      "step": 1442
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.6929133858267718e-06,
+      "loss": 0.6322,
+      "step": 1443
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.6797900262467193e-06,
+      "loss": 0.598,
+      "step": 1444
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.5934,
+      "step": 1445
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.6535433070866144e-06,
+      "loss": 0.6414,
+      "step": 1446
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.640419947506562e-06,
+      "loss": 0.6002,
+      "step": 1447
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.6272965879265093e-06,
+      "loss": 0.6249,
+      "step": 1448
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.6141732283464568e-06,
+      "loss": 0.6056,
+      "step": 1449
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.6010498687664044e-06,
+      "loss": 0.5995,
+      "step": 1450
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.587926509186352e-06,
+      "loss": 0.6305,
+      "step": 1451
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.5748031496062992e-06,
+      "loss": 0.6167,
+      "step": 1452
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.5616797900262468e-06,
+      "loss": 0.6691,
+      "step": 1453
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.5485564304461945e-06,
+      "loss": 0.6373,
+      "step": 1454
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.5354330708661418e-06,
+      "loss": 0.6748,
+      "step": 1455
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.5223097112860894e-06,
+      "loss": 0.5977,
+      "step": 1456
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.5091863517060367e-06,
+      "loss": 0.5841,
+      "step": 1457
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.4960629921259845e-06,
+      "loss": 0.6255,
+      "step": 1458
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.482939632545932e-06,
+      "loss": 0.6535,
+      "step": 1459
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.4698162729658793e-06,
+      "loss": 0.6219,
+      "step": 1460
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.4566929133858269e-06,
+      "loss": 0.6383,
+      "step": 1461
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.4435695538057744e-06,
+      "loss": 0.6382,
+      "step": 1462
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.430446194225722e-06,
+      "loss": 0.5947,
+      "step": 1463
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.4173228346456693e-06,
+      "loss": 0.585,
+      "step": 1464
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.4041994750656168e-06,
+      "loss": 0.6183,
+      "step": 1465
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.3910761154855646e-06,
+      "loss": 0.5613,
+      "step": 1466
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.377952755905512e-06,
+      "loss": 0.6758,
+      "step": 1467
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.3648293963254594e-06,
+      "loss": 0.6109,
+      "step": 1468
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.3517060367454068e-06,
+      "loss": 0.6069,
+      "step": 1469
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.3385826771653545e-06,
+      "loss": 0.6242,
+      "step": 1470
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.325459317585302e-06,
+      "loss": 0.63,
+      "step": 1471
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.3123359580052494e-06,
+      "loss": 0.6479,
+      "step": 1472
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.299212598425197e-06,
+      "loss": 0.6132,
+      "step": 1473
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.2860892388451447e-06,
+      "loss": 0.6606,
+      "step": 1474
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.272965879265092e-06,
+      "loss": 0.6714,
+      "step": 1475
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.2598425196850396e-06,
+      "loss": 0.5975,
+      "step": 1476
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.246719160104987e-06,
+      "loss": 0.5945,
+      "step": 1477
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.2335958005249344e-06,
+      "loss": 0.6447,
+      "step": 1478
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.220472440944882e-06,
+      "loss": 0.5509,
+      "step": 1479
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.2073490813648295e-06,
+      "loss": 0.5844,
+      "step": 1480
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.194225721784777e-06,
+      "loss": 0.6216,
+      "step": 1481
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.1811023622047246e-06,
+      "loss": 0.6026,
+      "step": 1482
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.1679790026246721e-06,
+      "loss": 0.5624,
+      "step": 1483
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.1548556430446194e-06,
+      "loss": 0.7082,
+      "step": 1484
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.141732283464567e-06,
+      "loss": 0.6144,
+      "step": 1485
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.1286089238845145e-06,
+      "loss": 0.5692,
+      "step": 1486
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.115485564304462e-06,
+      "loss": 0.6262,
+      "step": 1487
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.1023622047244096e-06,
+      "loss": 0.653,
+      "step": 1488
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.0892388451443571e-06,
+      "loss": 0.5855,
+      "step": 1489
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.0761154855643045e-06,
+      "loss": 0.5973,
+      "step": 1490
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.062992125984252e-06,
+      "loss": 0.6105,
+      "step": 1491
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.0498687664041996e-06,
+      "loss": 0.5299,
+      "step": 1492
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.036745406824147e-06,
+      "loss": 0.6346,
+      "step": 1493
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.0236220472440946e-06,
+      "loss": 0.6264,
+      "step": 1494
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1.0104986876640422e-06,
+      "loss": 0.6193,
+      "step": 1495
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 9.973753280839895e-07,
+      "loss": 0.595,
+      "step": 1496
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 9.84251968503937e-07,
+      "loss": 0.6105,
+      "step": 1497
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 9.711286089238846e-07,
+      "loss": 0.6622,
+      "step": 1498
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.580052493438321e-07,
+      "loss": 0.6297,
+      "step": 1499
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.448818897637796e-07,
+      "loss": 0.6445,
+      "step": 1500
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.317585301837272e-07,
+      "loss": 0.6366,
+      "step": 1501
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.186351706036746e-07,
+      "loss": 0.6128,
+      "step": 1502
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 9.055118110236222e-07,
+      "loss": 0.5532,
+      "step": 1503
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 8.923884514435696e-07,
+      "loss": 0.6056,
+      "step": 1504
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 8.792650918635172e-07,
+      "loss": 0.5792,
+      "step": 1505
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 8.661417322834646e-07,
+      "loss": 0.5819,
+      "step": 1506
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 8.530183727034122e-07,
+      "loss": 0.7251,
+      "step": 1507
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 8.398950131233597e-07,
+      "loss": 0.692,
+      "step": 1508
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 8.267716535433072e-07,
+      "loss": 0.6952,
+      "step": 1509
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 8.136482939632546e-07,
+      "loss": 0.6062,
+      "step": 1510
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 8.005249343832022e-07,
+      "loss": 0.5239,
+      "step": 1511
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 7.874015748031496e-07,
+      "loss": 0.5989,
+      "step": 1512
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 7.742782152230973e-07,
+      "loss": 0.661,
+      "step": 1513
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 7.611548556430447e-07,
+      "loss": 0.5685,
+      "step": 1514
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 7.480314960629922e-07,
+      "loss": 0.6027,
+      "step": 1515
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 7.349081364829397e-07,
+      "loss": 0.6204,
+      "step": 1516
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 7.217847769028872e-07,
+      "loss": 0.5488,
+      "step": 1517
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 7.086614173228346e-07,
+      "loss": 0.6015,
+      "step": 1518
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 6.955380577427823e-07,
+      "loss": 0.6198,
+      "step": 1519
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 6.824146981627297e-07,
+      "loss": 0.6718,
+      "step": 1520
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 6.692913385826773e-07,
+      "loss": 0.5899,
+      "step": 1521
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 6.561679790026247e-07,
+      "loss": 0.5739,
+      "step": 1522
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 6.430446194225723e-07,
+      "loss": 0.6321,
+      "step": 1523
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 6.299212598425198e-07,
+      "loss": 0.6345,
+      "step": 1524
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 6.167979002624672e-07,
+      "loss": 0.5331,
+      "step": 1525
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 6.036745406824148e-07,
+      "loss": 0.6951,
+      "step": 1526
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5.905511811023623e-07,
+      "loss": 0.6193,
+      "step": 1527
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5.774278215223097e-07,
+      "loss": 0.5597,
+      "step": 1528
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5.643044619422573e-07,
+      "loss": 0.66,
+      "step": 1529
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5.511811023622048e-07,
+      "loss": 0.6475,
+      "step": 1530
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5.380577427821522e-07,
+      "loss": 0.6018,
+      "step": 1531
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5.249343832020998e-07,
+      "loss": 0.5828,
+      "step": 1532
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5.118110236220473e-07,
+      "loss": 0.5536,
+      "step": 1533
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 4.986876640419948e-07,
+      "loss": 0.6034,
+      "step": 1534
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 4.855643044619423e-07,
+      "loss": 0.5538,
+      "step": 1535
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 4.724409448818898e-07,
+      "loss": 0.5911,
+      "step": 1536
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 4.593175853018373e-07,
+      "loss": 0.6138,
+      "step": 1537
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 4.461942257217848e-07,
+      "loss": 0.5463,
+      "step": 1538
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 4.330708661417323e-07,
+      "loss": 0.562,
+      "step": 1539
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 4.1994750656167983e-07,
+      "loss": 0.6298,
+      "step": 1540
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 4.068241469816273e-07,
+      "loss": 0.5527,
+      "step": 1541
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 3.937007874015748e-07,
+      "loss": 0.5853,
+      "step": 1542
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 3.8057742782152235e-07,
+      "loss": 0.6271,
+      "step": 1543
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 3.6745406824146983e-07,
+      "loss": 0.5871,
+      "step": 1544
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 3.543307086614173e-07,
+      "loss": 0.5981,
+      "step": 1545
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 3.4120734908136486e-07,
+      "loss": 0.6062,
+      "step": 1546
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 3.2808398950131235e-07,
+      "loss": 0.7301,
+      "step": 1547
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 3.149606299212599e-07,
+      "loss": 0.5831,
+      "step": 1548
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 3.018372703412074e-07,
+      "loss": 0.5431,
+      "step": 1549
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 2.8871391076115486e-07,
+      "loss": 0.5531,
+      "step": 1550
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 2.755905511811024e-07,
+      "loss": 0.5909,
+      "step": 1551
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 2.624671916010499e-07,
+      "loss": 0.5495,
+      "step": 1552
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 2.493438320209974e-07,
+      "loss": 0.6404,
+      "step": 1553
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 2.362204724409449e-07,
+      "loss": 0.5947,
+      "step": 1554
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 2.230971128608924e-07,
+      "loss": 0.6294,
+      "step": 1555
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 2.0997375328083992e-07,
+      "loss": 0.6102,
+      "step": 1556
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1.968503937007874e-07,
+      "loss": 0.5825,
+      "step": 1557
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.8372703412073492e-07,
+      "loss": 0.5933,
+      "step": 1558
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.7060367454068243e-07,
+      "loss": 0.5506,
+      "step": 1559
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.5748031496062994e-07,
+      "loss": 0.6542,
+      "step": 1560
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.4435695538057743e-07,
+      "loss": 0.5879,
+      "step": 1561
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.3123359580052494e-07,
+      "loss": 0.5935,
+      "step": 1562
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.1811023622047244e-07,
+      "loss": 0.5989,
+      "step": 1563
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.0498687664041996e-07,
+      "loss": 0.5682,
+      "step": 1564
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 9.186351706036746e-08,
+      "loss": 0.6218,
+      "step": 1565
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 7.874015748031497e-08,
+      "loss": 0.6013,
+      "step": 1566
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 6.561679790026247e-08,
+      "loss": 0.6295,
+      "step": 1567
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5.249343832020998e-08,
+      "loss": 0.5939,
+      "step": 1568
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 3.9370078740157486e-08,
+      "loss": 0.5385,
+      "step": 1569
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 2.624671916010499e-08,
+      "loss": 0.5929,
+      "step": 1570
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.3123359580052495e-08,
+      "loss": 0.5997,
+      "step": 1571
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.0,
+      "loss": 0.6515,
+      "step": 1572
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1572,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "total_flos": 4.2441624869719245e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}