diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7045 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9961612284069097,
+  "global_step": 1170,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 1.2046,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 1.2021,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 1.2134,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 1.188,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 1.1831,
+      "step": 5
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 1.1465,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 1.1309,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 1.0444,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 1.0173,
+      "step": 9
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 1.0151,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 1.0056,
+      "step": 11
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 0.9863,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 0.9473,
+      "step": 13
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 0.9294,
+      "step": 14
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 0.957,
+      "step": 15
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 0.9277,
+      "step": 16
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 0.9121,
+      "step": 17
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 0.9023,
+      "step": 18
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 0.8911,
+      "step": 19
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 0.8879,
+      "step": 20
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 0.8733,
+      "step": 21
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 0.8689,
+      "step": 22
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 0.853,
+      "step": 23
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.0212765957446808e-05,
+      "loss": 0.8608,
+      "step": 24
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.0638297872340426e-05,
+      "loss": 0.8423,
+      "step": 25
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.1063829787234044e-05,
+      "loss": 0.8281,
+      "step": 26
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.1489361702127662e-05,
+      "loss": 0.8394,
+      "step": 27
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.1914893617021277e-05,
+      "loss": 0.8311,
+      "step": 28
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.2340425531914895e-05,
+      "loss": 0.8267,
+      "step": 29
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.2765957446808513e-05,
+      "loss": 0.8081,
+      "step": 30
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.3191489361702127e-05,
+      "loss": 0.8013,
+      "step": 31
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.3617021276595745e-05,
+      "loss": 0.8062,
+      "step": 32
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.4042553191489363e-05,
+      "loss": 0.8025,
+      "step": 33
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.4468085106382981e-05,
+      "loss": 0.7817,
+      "step": 34
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.4893617021276596e-05,
+      "loss": 0.7937,
+      "step": 35
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.5319148936170214e-05,
+      "loss": 0.7864,
+      "step": 36
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.5744680851063832e-05,
+      "loss": 0.772,
+      "step": 37
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.6170212765957446e-05,
+      "loss": 0.7883,
+      "step": 38
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.6595744680851064e-05,
+      "loss": 0.77,
+      "step": 39
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.7021276595744682e-05,
+      "loss": 0.7642,
+      "step": 40
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.74468085106383e-05,
+      "loss": 0.7651,
+      "step": 41
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.7872340425531915e-05,
+      "loss": 0.7683,
+      "step": 42
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.8297872340425533e-05,
+      "loss": 0.7546,
+      "step": 43
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.872340425531915e-05,
+      "loss": 0.7681,
+      "step": 44
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.914893617021277e-05,
+      "loss": 0.7537,
+      "step": 45
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9574468085106384e-05,
+      "loss": 0.7583,
+      "step": 46
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 0.7471,
+      "step": 47
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9999960869990446e-05,
+      "loss": 0.7542,
+      "step": 48
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9999843480268005e-05,
+      "loss": 0.7417,
+      "step": 49
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9999647831751378e-05,
+      "loss": 0.7517,
+      "step": 50
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9999373925971705e-05,
+      "loss": 0.7522,
+      "step": 51
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9999021765072576e-05,
+      "loss": 0.7419,
+      "step": 52
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9998591351810002e-05,
+      "loss": 0.7349,
+      "step": 53
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9998082689552397e-05,
+      "loss": 0.7444,
+      "step": 54
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9997495782280554e-05,
+      "loss": 0.7373,
+      "step": 55
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.999683063458761e-05,
+      "loss": 0.7354,
+      "step": 56
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9996087251679012e-05,
+      "loss": 0.7368,
+      "step": 57
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9995265639372475e-05,
+      "loss": 0.731,
+      "step": 58
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9994365804097945e-05,
+      "loss": 0.7361,
+      "step": 59
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9993387752897522e-05,
+      "loss": 0.7307,
+      "step": 60
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.999233149342545e-05,
+      "loss": 0.7351,
+      "step": 61
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.999119703394801e-05,
+      "loss": 0.7358,
+      "step": 62
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9989984383343486e-05,
+      "loss": 0.7317,
+      "step": 63
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.998869355110208e-05,
+      "loss": 0.741,
+      "step": 64
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9987324547325856e-05,
+      "loss": 0.7312,
+      "step": 65
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9985877382728634e-05,
+      "loss": 0.7292,
+      "step": 66
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9984352068635927e-05,
+      "loss": 0.728,
+      "step": 67
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.998274861698485e-05,
+      "loss": 0.7285,
+      "step": 68
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9981067040324014e-05,
+      "loss": 0.7236,
+      "step": 69
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9979307351813443e-05,
+      "loss": 0.7244,
+      "step": 70
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.997746956522446e-05,
+      "loss": 0.7246,
+      "step": 71
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9975553694939593e-05,
+      "loss": 0.7332,
+      "step": 72
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9973559755952442e-05,
+      "loss": 0.7249,
+      "step": 73
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9971487763867578e-05,
+      "loss": 0.7205,
+      "step": 74
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9969337734900415e-05,
+      "loss": 0.7227,
+      "step": 75
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.996710968587708e-05,
+      "loss": 0.7217,
+      "step": 76
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9964803634234297e-05,
+      "loss": 0.7212,
+      "step": 77
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9962419598019226e-05,
+      "loss": 0.7209,
+      "step": 78
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.995995759588934e-05,
+      "loss": 0.7258,
+      "step": 79
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.995741764711227e-05,
+      "loss": 0.7244,
+      "step": 80
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.995479977156566e-05,
+      "loss": 0.7156,
+      "step": 81
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9952103989737013e-05,
+      "loss": 0.7244,
+      "step": 82
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9949330322723518e-05,
+      "loss": 0.7109,
+      "step": 83
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9946478792231904e-05,
+      "loss": 0.7087,
+      "step": 84
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.994354942057825e-05,
+      "loss": 0.7097,
+      "step": 85
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9940542230687822e-05,
+      "loss": 0.7136,
+      "step": 86
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.99374572460949e-05,
+      "loss": 0.7141,
+      "step": 87
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9934294490942573e-05,
+      "loss": 0.7195,
+      "step": 88
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9931053989982575e-05,
+      "loss": 0.7158,
+      "step": 89
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.992773576857507e-05,
+      "loss": 0.7214,
+      "step": 90
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9924339852688463e-05,
+      "loss": 0.7153,
+      "step": 91
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9920866268899203e-05,
+      "loss": 0.7229,
+      "step": 92
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9917315044391558e-05,
+      "loss": 0.7107,
+      "step": 93
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9913686206957417e-05,
+      "loss": 0.7065,
+      "step": 94
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9909979784996075e-05,
+      "loss": 0.7083,
+      "step": 95
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9906195807513994e-05,
+      "loss": 0.718,
+      "step": 96
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9902334304124592e-05,
+      "loss": 0.7092,
+      "step": 97
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.989839530504799e-05,
+      "loss": 0.72,
+      "step": 98
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.989437884111082e-05,
+      "loss": 0.7173,
+      "step": 99
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9890284943745927e-05,
+      "loss": 0.7087,
+      "step": 100
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.988611364499216e-05,
+      "loss": 0.7104,
+      "step": 101
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.988186497749411e-05,
+      "loss": 0.7188,
+      "step": 102
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9877538974501856e-05,
+      "loss": 0.7178,
+      "step": 103
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.987313566987071e-05,
+      "loss": 0.7119,
+      "step": 104
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9868655098060942e-05,
+      "loss": 0.7061,
+      "step": 105
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9864097294137517e-05,
+      "loss": 0.7073,
+      "step": 106
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9859462293769808e-05,
+      "loss": 0.7083,
+      "step": 107
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.985475013323135e-05,
+      "loss": 0.7209,
+      "step": 108
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.984996084939951e-05,
+      "loss": 0.709,
+      "step": 109
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9845094479755234e-05,
+      "loss": 0.72,
+      "step": 110
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9840151062382743e-05,
+      "loss": 0.7124,
+      "step": 111
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9835130635969228e-05,
+      "loss": 0.7158,
+      "step": 112
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.983003323980456e-05,
+      "loss": 0.7051,
+      "step": 113
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.982485891378097e-05,
+      "loss": 0.7058,
+      "step": 114
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.981960769839274e-05,
+      "loss": 0.7048,
+      "step": 115
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9814279634735896e-05,
+      "loss": 0.7073,
+      "step": 116
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.980887476450787e-05,
+      "loss": 0.708,
+      "step": 117
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9803393130007186e-05,
+      "loss": 0.7048,
+      "step": 118
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9797834774133133e-05,
+      "loss": 0.7036,
+      "step": 119
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.979219974038541e-05,
+      "loss": 0.7095,
+      "step": 120
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9786488072863802e-05,
+      "loss": 0.7053,
+      "step": 121
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.978069981626783e-05,
+      "loss": 0.7039,
+      "step": 122
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.97748350158964e-05,
+      "loss": 0.7056,
+      "step": 123
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9768893717647456e-05,
+      "loss": 0.7058,
+      "step": 124
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9762875968017602e-05,
+      "loss": 0.6985,
+      "step": 125
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9756781814101765e-05,
+      "loss": 0.7041,
+      "step": 126
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.97506113035928e-05,
+      "loss": 0.696,
+      "step": 127
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.974436448478114e-05,
+      "loss": 0.7039,
+      "step": 128
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9738041406554393e-05,
+      "loss": 0.7041,
+      "step": 129
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.973164211839699e-05,
+      "loss": 0.7041,
+      "step": 130
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9725166670389765e-05,
+      "loss": 0.7002,
+      "step": 131
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9718615113209588e-05,
+      "loss": 0.7031,
+      "step": 132
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.971198749812896e-05,
+      "loss": 0.709,
+      "step": 133
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9705283877015612e-05,
+      "loss": 0.7051,
+      "step": 134
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.969850430233209e-05,
+      "loss": 0.7102,
+      "step": 135
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9691648827135358e-05,
+      "loss": 0.7026,
+      "step": 136
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9684717505076386e-05,
+      "loss": 0.7078,
+      "step": 137
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9677710390399705e-05,
+      "loss": 0.6936,
+      "step": 138
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9670627537943013e-05,
+      "loss": 0.7065,
+      "step": 139
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9663469003136723e-05,
+      "loss": 0.7051,
+      "step": 140
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9656234842003546e-05,
+      "loss": 0.7029,
+      "step": 141
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9648925111158037e-05,
+      "loss": 0.7075,
+      "step": 142
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9641539867806164e-05,
+      "loss": 0.7029,
+      "step": 143
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9634079169744865e-05,
+      "loss": 0.71,
+      "step": 144
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.962654307536156e-05,
+      "loss": 0.7039,
+      "step": 145
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9618931643633754e-05,
+      "loss": 0.696,
+      "step": 146
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.961124493412852e-05,
+      "loss": 0.6934,
+      "step": 147
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9603483007002055e-05,
+      "loss": 0.699,
+      "step": 148
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9595645922999228e-05,
+      "loss": 0.7039,
+      "step": 149
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9587733743453065e-05,
+      "loss": 0.6982,
+      "step": 150
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.95797465302843e-05,
+      "loss": 0.7041,
+      "step": 151
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9571684346000877e-05,
+      "loss": 0.6885,
+      "step": 152
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.956354725369747e-05,
+      "loss": 0.6995,
+      "step": 153
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9555335317054974e-05,
+      "loss": 0.7014,
+      "step": 154
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9547048600340025e-05,
+      "loss": 0.7048,
+      "step": 155
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.953868716840448e-05,
+      "loss": 0.7014,
+      "step": 156
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9530251086684923e-05,
+      "loss": 0.6931,
+      "step": 157
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.952174042120215e-05,
+      "loss": 0.6941,
+      "step": 158
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9513155238560638e-05,
+      "loss": 0.7009,
+      "step": 159
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9504495605948047e-05,
+      "loss": 0.7034,
+      "step": 160
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9495761591134676e-05,
+      "loss": 0.7056,
+      "step": 161
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9486953262472945e-05,
+      "loss": 0.6946,
+      "step": 162
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.947807068889685e-05,
+      "loss": 0.7021,
+      "step": 163
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9469113939921425e-05,
+      "loss": 0.7017,
+      "step": 164
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.946008308564221e-05,
+      "loss": 0.6919,
+      "step": 165
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.945097819673468e-05,
+      "loss": 0.6948,
+      "step": 166
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9441799344453725e-05,
+      "loss": 0.7,
+      "step": 167
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.943254660063305e-05,
+      "loss": 0.7017,
+      "step": 168
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.942322003768465e-05,
+      "loss": 0.6892,
+      "step": 169
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9413819728598225e-05,
+      "loss": 0.6882,
+      "step": 170
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.940434574694061e-05,
+      "loss": 0.6982,
+      "step": 171
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9394798166855205e-05,
+      "loss": 0.6992,
+      "step": 172
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9385177063061385e-05,
+      "loss": 0.6982,
+      "step": 173
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9375482510853938e-05,
+      "loss": 0.6975,
+      "step": 174
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.936571458610244e-05,
+      "loss": 0.6873,
+      "step": 175
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9355873365250688e-05,
+      "loss": 0.7039,
+      "step": 176
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9345958925316094e-05,
+      "loss": 0.6934,
+      "step": 177
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.933597134388909e-05,
+      "loss": 0.6997,
+      "step": 178
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.932591069913251e-05,
+      "loss": 0.6953,
+      "step": 179
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.931577706978097e-05,
+      "loss": 0.6904,
+      "step": 180
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9305570535140273e-05,
+      "loss": 0.6956,
+      "step": 181
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9295291175086782e-05,
+      "loss": 0.698,
+      "step": 182
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.928493907006679e-05,
+      "loss": 0.6863,
+      "step": 183
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.927451430109589e-05,
+      "loss": 0.7012,
+      "step": 184
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9264016949758343e-05,
+      "loss": 0.6831,
+      "step": 185
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9253447098206436e-05,
+      "loss": 0.6885,
+      "step": 186
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9242804829159857e-05,
+      "loss": 0.6946,
+      "step": 187
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9232090225905016e-05,
+      "loss": 0.6926,
+      "step": 188
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.922130337229442e-05,
+      "loss": 0.686,
+      "step": 189
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9210444352746012e-05,
+      "loss": 0.6917,
+      "step": 190
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.919951325224249e-05,
+      "loss": 0.6965,
+      "step": 191
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9188510156330676e-05,
+      "loss": 0.6938,
+      "step": 192
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9177435151120815e-05,
+      "loss": 0.6934,
+      "step": 193
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9166288323285917e-05,
+      "loss": 0.6899,
+      "step": 194
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9155069760061086e-05,
+      "loss": 0.6934,
+      "step": 195
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.914377954924281e-05,
+      "loss": 0.7012,
+      "step": 196
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.913241777918831e-05,
+      "loss": 0.7002,
+      "step": 197
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.912098453881481e-05,
+      "loss": 0.6902,
+      "step": 198
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9109479917598876e-05,
+      "loss": 0.689,
+      "step": 199
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.90979040055757e-05,
+      "loss": 0.6855,
+      "step": 200
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9086256893338382e-05,
+      "loss": 0.6963,
+      "step": 201
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9074538672037254e-05,
+      "loss": 0.6995,
+      "step": 202
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.906274943337913e-05,
+      "loss": 0.6973,
+      "step": 203
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.905088926962662e-05,
+      "loss": 0.6948,
+      "step": 204
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9038958273597385e-05,
+      "loss": 0.6902,
+      "step": 205
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9026956538663426e-05,
+      "loss": 0.6995,
+      "step": 206
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9014884158750342e-05,
+      "loss": 0.6887,
+      "step": 207
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.90027412283366e-05,
+      "loss": 0.6943,
+      "step": 208
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8990527842452797e-05,
+      "loss": 0.6899,
+      "step": 209
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8978244096680914e-05,
+      "loss": 0.6865,
+      "step": 210
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8965890087153575e-05,
+      "loss": 0.6965,
+      "step": 211
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8953465910553274e-05,
+      "loss": 0.6941,
+      "step": 212
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8940971664111643e-05,
+      "loss": 0.6956,
+      "step": 213
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.892840744560868e-05,
+      "loss": 0.6853,
+      "step": 214
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8915773353371983e-05,
+      "loss": 0.6907,
+      "step": 215
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.890306948627598e-05,
+      "loss": 0.689,
+      "step": 216
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8890295943741163e-05,
+      "loss": 0.6819,
+      "step": 217
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8877452825733296e-05,
+      "loss": 0.6982,
+      "step": 218
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8864540232762646e-05,
+      "loss": 0.7053,
+      "step": 219
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.885155826588319e-05,
+      "loss": 0.687,
+      "step": 220
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8838507026691833e-05,
+      "loss": 0.6912,
+      "step": 221
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8825386617327586e-05,
+      "loss": 0.6848,
+      "step": 222
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8812197140470803e-05,
+      "loss": 0.6846,
+      "step": 223
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.879893869934236e-05,
+      "loss": 0.689,
+      "step": 224
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8785611397702835e-05,
+      "loss": 0.6865,
+      "step": 225
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8772215339851716e-05,
+      "loss": 0.6875,
+      "step": 226
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8758750630626584e-05,
+      "loss": 0.6868,
+      "step": 227
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.874521737540228e-05,
+      "loss": 0.6875,
+      "step": 228
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8731615680090075e-05,
+      "loss": 0.6938,
+      "step": 229
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8717945651136872e-05,
+      "loss": 0.6934,
+      "step": 230
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8704207395524337e-05,
+      "loss": 0.6819,
+      "step": 231
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.869040102076809e-05,
+      "loss": 0.6794,
+      "step": 232
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8676526634916843e-05,
+      "loss": 0.6975,
+      "step": 233
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8662584346551566e-05,
+      "loss": 0.6809,
+      "step": 234
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.864857426478464e-05,
+      "loss": 0.6892,
+      "step": 235
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.863449649925898e-05,
+      "loss": 0.6892,
+      "step": 236
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.862035116014722e-05,
+      "loss": 0.6902,
+      "step": 237
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8606138358150804e-05,
+      "loss": 0.6863,
+      "step": 238
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8591858204499143e-05,
+      "loss": 0.6917,
+      "step": 239
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8577510810948758e-05,
+      "loss": 0.6841,
+      "step": 240
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8563096289782366e-05,
+      "loss": 0.6865,
+      "step": 241
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8548614753808046e-05,
+      "loss": 0.6824,
+      "step": 242
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8534066316358325e-05,
+      "loss": 0.6902,
+      "step": 243
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8519451091289298e-05,
+      "loss": 0.6914,
+      "step": 244
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8504769192979748e-05,
+      "loss": 0.6812,
+      "step": 245
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8490020736330236e-05,
+      "loss": 0.6792,
+      "step": 246
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8475205836762215e-05,
+      "loss": 0.6941,
+      "step": 247
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8460324610217117e-05,
+      "loss": 0.6838,
+      "step": 248
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8445377173155447e-05,
+      "loss": 0.6907,
+      "step": 249
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.843036364255588e-05,
+      "loss": 0.6956,
+      "step": 250
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8415284135914337e-05,
+      "loss": 0.6924,
+      "step": 251
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8400138771243056e-05,
+      "loss": 0.6765,
+      "step": 252
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8384927667069696e-05,
+      "loss": 0.6846,
+      "step": 253
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8369650942436387e-05,
+      "loss": 0.6804,
+      "step": 254
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8354308716898807e-05,
+      "loss": 0.6833,
+      "step": 255
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.833890111052524e-05,
+      "loss": 0.6838,
+      "step": 256
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8323428243895644e-05,
+      "loss": 0.6787,
+      "step": 257
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8307890238100698e-05,
+      "loss": 0.6907,
+      "step": 258
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8292287214740872e-05,
+      "loss": 0.6887,
+      "step": 259
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8276619295925457e-05,
+      "loss": 0.6729,
+      "step": 260
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.826088660427161e-05,
+      "loss": 0.6887,
+      "step": 261
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8245089262903407e-05,
+      "loss": 0.6838,
+      "step": 262
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8229227395450876e-05,
+      "loss": 0.697,
+      "step": 263
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.821330112604902e-05,
+      "loss": 0.6855,
+      "step": 264
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.819731057933685e-05,
+      "loss": 0.6855,
+      "step": 265
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8181255880456422e-05,
+      "loss": 0.6865,
+      "step": 266
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8165137155051832e-05,
+      "loss": 0.6821,
+      "step": 267
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.8148954529268262e-05,
+      "loss": 0.6804,
+      "step": 268
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.8132708129750972e-05,
+      "loss": 0.6887,
+      "step": 269
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.811639808364431e-05,
+      "loss": 0.6887,
+      "step": 270
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.8100024518590737e-05,
+      "loss": 0.6885,
+      "step": 271
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.8083587562729796e-05,
+      "loss": 0.6851,
+      "step": 272
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.806708734469714e-05,
+      "loss": 0.6851,
+      "step": 273
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.8050523993623504e-05,
+      "loss": 0.6775,
+      "step": 274
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.803389763913371e-05,
+      "loss": 0.6882,
+      "step": 275
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.8017208411345636e-05,
+      "loss": 0.6851,
+      "step": 276
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.800045644086921e-05,
+      "loss": 0.6826,
+      "step": 277
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7983641858805384e-05,
+      "loss": 0.689,
+      "step": 278
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.796676479674511e-05,
+      "loss": 0.6807,
+      "step": 279
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7949825386768312e-05,
+      "loss": 0.6836,
+      "step": 280
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7932823761442845e-05,
+      "loss": 0.6807,
+      "step": 281
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7915760053823454e-05,
+      "loss": 0.689,
+      "step": 282
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7898634397450757e-05,
+      "loss": 0.6907,
+      "step": 283
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7881446926350167e-05,
+      "loss": 0.6841,
+      "step": 284
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7864197775030863e-05,
+      "loss": 0.6851,
+      "step": 285
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7846887078484743e-05,
+      "loss": 0.6809,
+      "step": 286
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7829514972185353e-05,
+      "loss": 0.6792,
+      "step": 287
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7812081592086818e-05,
+      "loss": 0.6851,
+      "step": 288
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7794587074622815e-05,
+      "loss": 0.6946,
+      "step": 289
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.777703155670547e-05,
+      "loss": 0.6855,
+      "step": 290
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7759415175724296e-05,
+      "loss": 0.6843,
+      "step": 291
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7741738069545126e-05,
+      "loss": 0.688,
+      "step": 292
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7724000376509023e-05,
+      "loss": 0.6799,
+      "step": 293
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7706202235431217e-05,
+      "loss": 0.6736,
+      "step": 294
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.768834378559998e-05,
+      "loss": 0.6917,
+      "step": 295
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7670425166775588e-05,
+      "loss": 0.6846,
+      "step": 296
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7652446519189173e-05,
+      "loss": 0.6772,
+      "step": 297
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.763440798354168e-05,
+      "loss": 0.6821,
+      "step": 298
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7616309701002713e-05,
+      "loss": 0.6848,
+      "step": 299
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7598151813209464e-05,
+      "loss": 0.6763,
+      "step": 300
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.757993446226561e-05,
+      "loss": 0.6833,
+      "step": 301
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.756165779074016e-05,
+      "loss": 0.6873,
+      "step": 302
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7543321941666388e-05,
+      "loss": 0.6809,
+      "step": 303
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7524927058540683e-05,
+      "loss": 0.6914,
+      "step": 304
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7506473285321438e-05,
+      "loss": 0.6809,
+      "step": 305
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7487960766427913e-05,
+      "loss": 0.6814,
+      "step": 306
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7469389646739117e-05,
+      "loss": 0.6843,
+      "step": 307
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7450760071592668e-05,
+      "loss": 0.6819,
+      "step": 308
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.743207218678366e-05,
+      "loss": 0.6816,
+      "step": 309
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.741332613856351e-05,
+      "loss": 0.6824,
+      "step": 310
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7394522073638833e-05,
+      "loss": 0.6772,
+      "step": 311
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7375660139170276e-05,
+      "loss": 0.6848,
+      "step": 312
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7356740482771367e-05,
+      "loss": 0.6753,
+      "step": 313
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7337763252507383e-05,
+      "loss": 0.6741,
+      "step": 314
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7318728596894157e-05,
+      "loss": 0.6843,
+      "step": 315
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7299636664896945e-05,
+      "loss": 0.676,
+      "step": 316
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.728048760592924e-05,
+      "loss": 0.6875,
+      "step": 317
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7261281569851615e-05,
+      "loss": 0.6851,
+      "step": 318
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7242018706970545e-05,
+      "loss": 0.675,
+      "step": 319
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7222699168037236e-05,
+      "loss": 0.6812,
+      "step": 320
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.720332310424643e-05,
+      "loss": 0.6824,
+      "step": 321
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7183890667235245e-05,
+      "loss": 0.6721,
+      "step": 322
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.716440200908196e-05,
+      "loss": 0.6746,
+      "step": 323
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.7144857282304865e-05,
+      "loss": 0.6804,
+      "step": 324
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.712525663986102e-05,
+      "loss": 0.6797,
+      "step": 325
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.7105600235145092e-05,
+      "loss": 0.6748,
+      "step": 326
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.7085888221988144e-05,
+      "loss": 0.6843,
+      "step": 327
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.7066120754656426e-05,
+      "loss": 0.6819,
+      "step": 328
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.7046297987850177e-05,
+      "loss": 0.6729,
+      "step": 329
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.7026420076702406e-05,
+      "loss": 0.687,
+      "step": 330
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.7006487176777685e-05,
+      "loss": 0.6777,
+      "step": 331
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.698649944407093e-05,
+      "loss": 0.676,
+      "step": 332
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.696645703500617e-05,
+      "loss": 0.6733,
+      "step": 333
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.694636010643534e-05,
+      "loss": 0.6836,
+      "step": 334
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6926208815637037e-05,
+      "loss": 0.6807,
+      "step": 335
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6906003320315306e-05,
+      "loss": 0.686,
+      "step": 336
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6885743778598388e-05,
+      "loss": 0.6826,
+      "step": 337
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.68654303490375e-05,
+      "loss": 0.6804,
+      "step": 338
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.684506319060558e-05,
+      "loss": 0.6768,
+      "step": 339
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6824642462696043e-05,
+      "loss": 0.6648,
+      "step": 340
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6804168325121552e-05,
+      "loss": 0.6733,
+      "step": 341
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6783640938112742e-05,
+      "loss": 0.6738,
+      "step": 342
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6763060462316986e-05,
+      "loss": 0.6746,
+      "step": 343
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6742427058797122e-05,
+      "loss": 0.6838,
+      "step": 344
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.672174088903021e-05,
+      "loss": 0.6821,
+      "step": 345
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6701002114906254e-05,
+      "loss": 0.6721,
+      "step": 346
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.668021089872694e-05,
+      "loss": 0.6714,
+      "step": 347
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.665936740320436e-05,
+      "loss": 0.6907,
+      "step": 348
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6638471791459757e-05,
+      "loss": 0.6777,
+      "step": 349
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6617524227022225e-05,
+      "loss": 0.6746,
+      "step": 350
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6596524873827444e-05,
+      "loss": 0.6804,
+      "step": 351
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6575473896216393e-05,
+      "loss": 0.6846,
+      "step": 352
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.655437145893406e-05,
+      "loss": 0.6799,
+      "step": 353
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6533217727128165e-05,
+      "loss": 0.6785,
+      "step": 354
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.651201286634785e-05,
+      "loss": 0.6841,
+      "step": 355
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6490757042542393e-05,
+      "loss": 0.6763,
+      "step": 356
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6469450422059914e-05,
+      "loss": 0.6755,
+      "step": 357
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.644809317164607e-05,
+      "loss": 0.6777,
+      "step": 358
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.642668545844274e-05,
+      "loss": 0.6863,
+      "step": 359
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6405227449986728e-05,
+      "loss": 0.6768,
+      "step": 360
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.638371931420845e-05,
+      "loss": 0.6831,
+      "step": 361
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6362161219430618e-05,
+      "loss": 0.668,
+      "step": 362
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6340553334366922e-05,
+      "loss": 0.6758,
+      "step": 363
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6318895828120714e-05,
+      "loss": 0.6821,
+      "step": 364
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6297188870183673e-05,
+      "loss": 0.6653,
+      "step": 365
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.62754326304345e-05,
+      "loss": 0.6809,
+      "step": 366
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6253627279137567e-05,
+      "loss": 0.679,
+      "step": 367
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6231772986941593e-05,
+      "loss": 0.6765,
+      "step": 368
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6209869924878315e-05,
+      "loss": 0.6819,
+      "step": 369
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6187918264361135e-05,
+      "loss": 0.677,
+      "step": 370
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.616591817718379e-05,
+      "loss": 0.6768,
+      "step": 371
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6143869835519003e-05,
+      "loss": 0.6746,
+      "step": 372
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.612177341191714e-05,
+      "loss": 0.6851,
+      "step": 373
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.609962907930486e-05,
+      "loss": 0.6794,
+      "step": 374
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.607743701098374e-05,
+      "loss": 0.6792,
+      "step": 375
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.6055197380628963e-05,
+      "loss": 0.6836,
+      "step": 376
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.603291036228791e-05,
+      "loss": 0.6726,
+      "step": 377
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6010576130378828e-05,
+      "loss": 0.677,
+      "step": 378
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5988194859689463e-05,
+      "loss": 0.675,
+      "step": 379
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5965766725375678e-05,
+      "loss": 0.6724,
+      "step": 380
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.59432919029601e-05,
+      "loss": 0.6758,
+      "step": 381
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.592077056833073e-05,
+      "loss": 0.6689,
+      "step": 382
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5898202897739574e-05,
+      "loss": 0.6736,
+      "step": 383
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5875589067801268e-05,
+      "loss": 0.6724,
+      "step": 384
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5852929255491688e-05,
+      "loss": 0.6743,
+      "step": 385
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5830223638146565e-05,
+      "loss": 0.6672,
+      "step": 386
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5807472393460107e-05,
+      "loss": 0.6863,
+      "step": 387
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.57846756994836e-05,
+      "loss": 0.6675,
+      "step": 388
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5761833734624007e-05,
+      "loss": 0.6765,
+      "step": 389
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5738946677642596e-05,
+      "loss": 0.6799,
+      "step": 390
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5716014707653518e-05,
+      "loss": 0.6418,
+      "step": 391
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5693038004122415e-05,
+      "loss": 0.6272,
+      "step": 392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5670016746865007e-05,
+      "loss": 0.6326,
+      "step": 393
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.56469511160457e-05,
+      "loss": 0.6294,
+      "step": 394
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5623841292176168e-05,
+      "loss": 0.6216,
+      "step": 395
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5600687456113936e-05,
+      "loss": 0.6311,
+      "step": 396
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5577489789060963e-05,
+      "loss": 0.6304,
+      "step": 397
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5554248472562245e-05,
+      "loss": 0.6321,
+      "step": 398
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.553096368850436e-05,
+      "loss": 0.6265,
+      "step": 399
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.550763561911408e-05,
+      "loss": 0.6284,
+      "step": 400
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5484264446956916e-05,
+      "loss": 0.6287,
+      "step": 401
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.546085035493571e-05,
+      "loss": 0.6262,
+      "step": 402
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5437393526289187e-05,
+      "loss": 0.6355,
+      "step": 403
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5413894144590532e-05,
+      "loss": 0.6272,
+      "step": 404
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.539035239374596e-05,
+      "loss": 0.6287,
+      "step": 405
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5366768457993248e-05,
+      "loss": 0.6218,
+      "step": 406
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5343142521900323e-05,
+      "loss": 0.6301,
+      "step": 407
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5319474770363814e-05,
+      "loss": 0.6226,
+      "step": 408
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.529576538860758e-05,
+      "loss": 0.6284,
+      "step": 409
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5272014562181294e-05,
+      "loss": 0.6257,
+      "step": 410
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5248222476958972e-05,
+      "loss": 0.6238,
+      "step": 411
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.522438931913751e-05,
+      "loss": 0.625,
+      "step": 412
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5200515275235254e-05,
+      "loss": 0.6257,
+      "step": 413
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.517660053209051e-05,
+      "loss": 0.6343,
+      "step": 414
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5152645276860111e-05,
+      "loss": 0.6333,
+      "step": 415
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5128649697017925e-05,
+      "loss": 0.6204,
+      "step": 416
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5104613980353407e-05,
+      "loss": 0.6304,
+      "step": 417
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5080538314970123e-05,
+      "loss": 0.6377,
+      "step": 418
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.5056422889284276e-05,
+      "loss": 0.6226,
+      "step": 419
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.5032267892023234e-05,
+      "loss": 0.6301,
+      "step": 420
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.5008073512224046e-05,
+      "loss": 0.6345,
+      "step": 421
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4983839939231986e-05,
+      "loss": 0.634,
+      "step": 422
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.495956736269903e-05,
+      "loss": 0.627,
+      "step": 423
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4935255972582415e-05,
+      "loss": 0.6377,
+      "step": 424
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4910905959143125e-05,
+      "loss": 0.6379,
+      "step": 425
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4886517512944413e-05,
+      "loss": 0.6311,
+      "step": 426
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4862090824850305e-05,
+      "loss": 0.6304,
+      "step": 427
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4837626086024113e-05,
+      "loss": 0.6304,
+      "step": 428
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.481312348792692e-05,
+      "loss": 0.6335,
+      "step": 429
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4788583222316112e-05,
+      "loss": 0.6184,
+      "step": 430
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4764005481243854e-05,
+      "loss": 0.6262,
+      "step": 431
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4739390457055593e-05,
+      "loss": 0.6274,
+      "step": 432
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.471473834238856e-05,
+      "loss": 0.625,
+      "step": 433
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4690049330170242e-05,
+      "loss": 0.6299,
+      "step": 434
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4665323613616904e-05,
+      "loss": 0.6194,
+      "step": 435
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4640561386232054e-05,
+      "loss": 0.6204,
+      "step": 436
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4615762841804922e-05,
+      "loss": 0.6252,
+      "step": 437
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4590928174408966e-05,
+      "loss": 0.6292,
+      "step": 438
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4566057578400341e-05,
+      "loss": 0.6235,
+      "step": 439
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4541151248416384e-05,
+      "loss": 0.6248,
+      "step": 440
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4516209379374076e-05,
+      "loss": 0.6296,
+      "step": 441
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4491232166468532e-05,
+      "loss": 0.6279,
+      "step": 442
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.446621980517147e-05,
+      "loss": 0.6323,
+      "step": 443
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4441172491229677e-05,
+      "loss": 0.6191,
+      "step": 444
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4416090420663475e-05,
+      "loss": 0.636,
+      "step": 445
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4390973789765203e-05,
+      "loss": 0.6274,
+      "step": 446
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4365822795097662e-05,
+      "loss": 0.6338,
+      "step": 447
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4340637633492578e-05,
+      "loss": 0.6267,
+      "step": 448
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4315418502049079e-05,
+      "loss": 0.6362,
+      "step": 449
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4290165598132135e-05,
+      "loss": 0.6257,
+      "step": 450
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.426487911937102e-05,
+      "loss": 0.6255,
+      "step": 451
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4239559263657764e-05,
+      "loss": 0.6262,
+      "step": 452
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.421420622914561e-05,
+      "loss": 0.6299,
+      "step": 453
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4188820214247452e-05,
+      "loss": 0.6331,
+      "step": 454
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.4163401417634289e-05,
+      "loss": 0.626,
+      "step": 455
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.4137950038233672e-05,
+      "loss": 0.626,
+      "step": 456
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.411246627522815e-05,
+      "loss": 0.6267,
+      "step": 457
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.4086950328053696e-05,
+      "loss": 0.6328,
+      "step": 458
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.4061402396398163e-05,
+      "loss": 0.6299,
+      "step": 459
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.4035822680199716e-05,
+      "loss": 0.6282,
+      "step": 460
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.4010211379645258e-05,
+      "loss": 0.6331,
+      "step": 461
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3984568695168877e-05,
+      "loss": 0.6326,
+      "step": 462
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3958894827450272e-05,
+      "loss": 0.6292,
+      "step": 463
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3933189977413182e-05,
+      "loss": 0.6326,
+      "step": 464
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3907454346223807e-05,
+      "loss": 0.6311,
+      "step": 465
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3881688135289253e-05,
+      "loss": 0.6326,
+      "step": 466
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3855891546255935e-05,
+      "loss": 0.6301,
+      "step": 467
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3830064781008006e-05,
+      "loss": 0.6226,
+      "step": 468
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3804208041665777e-05,
+      "loss": 0.635,
+      "step": 469
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3778321530584143e-05,
+      "loss": 0.6311,
+      "step": 470
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.375240545035099e-05,
+      "loss": 0.623,
+      "step": 471
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3726460003785609e-05,
+      "loss": 0.626,
+      "step": 472
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3700485393937117e-05,
+      "loss": 0.6279,
+      "step": 473
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.367448182408286e-05,
+      "loss": 0.6348,
+      "step": 474
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3648449497726823e-05,
+      "loss": 0.6206,
+      "step": 475
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3622388618598045e-05,
+      "loss": 0.6326,
+      "step": 476
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3596299390649013e-05,
+      "loss": 0.6255,
+      "step": 477
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3570182018054077e-05,
+      "loss": 0.6252,
+      "step": 478
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3544036705207844e-05,
+      "loss": 0.6301,
+      "step": 479
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.351786365672358e-05,
+      "loss": 0.6306,
+      "step": 480
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3491663077431614e-05,
+      "loss": 0.6318,
+      "step": 481
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3465435172377733e-05,
+      "loss": 0.6301,
+      "step": 482
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.343918014682157e-05,
+      "loss": 0.6255,
+      "step": 483
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3412898206235e-05,
+      "loss": 0.6274,
+      "step": 484
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3386589556300552e-05,
+      "loss": 0.6233,
+      "step": 485
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3360254402909755e-05,
+      "loss": 0.6299,
+      "step": 486
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3333892952161582e-05,
+      "loss": 0.6243,
+      "step": 487
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3307505410360793e-05,
+      "loss": 0.6289,
+      "step": 488
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3281091984016343e-05,
+      "loss": 0.6396,
+      "step": 489
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.325465287983975e-05,
+      "loss": 0.6313,
+      "step": 490
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3228188304743505e-05,
+      "loss": 0.6399,
+      "step": 491
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3201698465839414e-05,
+      "loss": 0.6274,
+      "step": 492
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3175183570437009e-05,
+      "loss": 0.6252,
+      "step": 493
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.3148643826041911e-05,
+      "loss": 0.6396,
+      "step": 494
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.3122079440354213e-05,
+      "loss": 0.6418,
+      "step": 495
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.309549062126685e-05,
+      "loss": 0.6367,
+      "step": 496
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.3068877576863967e-05,
+      "loss": 0.625,
+      "step": 497
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.30422405154193e-05,
+      "loss": 0.6348,
+      "step": 498
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.3015579645394549e-05,
+      "loss": 0.6238,
+      "step": 499
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2988895175437728e-05,
+      "loss": 0.6172,
+      "step": 500
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2962187314381549e-05,
+      "loss": 0.6262,
+      "step": 501
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2935456271241784e-05,
+      "loss": 0.6335,
+      "step": 502
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2908702255215634e-05,
+      "loss": 0.6292,
+      "step": 503
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2881925475680076e-05,
+      "loss": 0.6296,
+      "step": 504
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2855126142190231e-05,
+      "loss": 0.6313,
+      "step": 505
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2828304464477745e-05,
+      "loss": 0.6311,
+      "step": 506
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2801460652449113e-05,
+      "loss": 0.6233,
+      "step": 507
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.277459491618406e-05,
+      "loss": 0.6238,
+      "step": 508
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2747707465933889e-05,
+      "loss": 0.6184,
+      "step": 509
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2720798512119838e-05,
+      "loss": 0.6274,
+      "step": 510
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2693868265331432e-05,
+      "loss": 0.6216,
+      "step": 511
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.266691693632483e-05,
+      "loss": 0.6292,
+      "step": 512
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2639944736021193e-05,
+      "loss": 0.6235,
+      "step": 513
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2612951875505e-05,
+      "loss": 0.625,
+      "step": 514
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2585938566022438e-05,
+      "loss": 0.6208,
+      "step": 515
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2558905018979713e-05,
+      "loss": 0.6213,
+      "step": 516
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.253185144594142e-05,
+      "loss": 0.6208,
+      "step": 517
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2504778058628872e-05,
+      "loss": 0.624,
+      "step": 518
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2477685068918448e-05,
+      "loss": 0.6216,
+      "step": 519
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2450572688839941e-05,
+      "loss": 0.6316,
+      "step": 520
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2423441130574887e-05,
+      "loss": 0.6306,
+      "step": 521
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2396290606454912e-05,
+      "loss": 0.6309,
+      "step": 522
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2369121328960072e-05,
+      "loss": 0.6294,
+      "step": 523
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2341933510717184e-05,
+      "loss": 0.6309,
+      "step": 524
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2314727364498162e-05,
+      "loss": 0.6211,
+      "step": 525
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2287503103218363e-05,
+      "loss": 0.6235,
+      "step": 526
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2260260939934907e-05,
+      "loss": 0.6323,
+      "step": 527
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2233001087845013e-05,
+      "loss": 0.6367,
+      "step": 528
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2205723760284338e-05,
+      "loss": 0.6277,
+      "step": 529
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.21784291707253e-05,
+      "loss": 0.6277,
+      "step": 530
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2151117532775409e-05,
+      "loss": 0.6248,
+      "step": 531
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2123789060175593e-05,
+      "loss": 0.6318,
+      "step": 532
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2096443966798533e-05,
+      "loss": 0.6272,
+      "step": 533
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2069082466646981e-05,
+      "loss": 0.6248,
+      "step": 534
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2041704773852092e-05,
+      "loss": 0.6306,
+      "step": 535
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2014311102671738e-05,
+      "loss": 0.624,
+      "step": 536
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1986901667488849e-05,
+      "loss": 0.6296,
+      "step": 537
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1959476682809707e-05,
+      "loss": 0.625,
+      "step": 538
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1932036363262304e-05,
+      "loss": 0.6223,
+      "step": 539
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1904580923594629e-05,
+      "loss": 0.6335,
+      "step": 540
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1877110578673005e-05,
+      "loss": 0.6294,
+      "step": 541
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1849625543480402e-05,
+      "loss": 0.6257,
+      "step": 542
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1822126033114763e-05,
+      "loss": 0.635,
+      "step": 543
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1794612262787308e-05,
+      "loss": 0.6262,
+      "step": 544
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1767084447820852e-05,
+      "loss": 0.6326,
+      "step": 545
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.173954280364813e-05,
+      "loss": 0.6367,
+      "step": 546
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1711987545810101e-05,
+      "loss": 0.6323,
+      "step": 547
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1684418889954267e-05,
+      "loss": 0.6274,
+      "step": 548
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1656837051832981e-05,
+      "loss": 0.6267,
+      "step": 549
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1629242247301759e-05,
+      "loss": 0.6277,
+      "step": 550
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1601634692317598e-05,
+      "loss": 0.6379,
+      "step": 551
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1574014602937274e-05,
+      "loss": 0.6287,
+      "step": 552
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1546382195315654e-05,
+      "loss": 0.6265,
+      "step": 553
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1518737685704023e-05,
+      "loss": 0.6252,
+      "step": 554
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.149108129044836e-05,
+      "loss": 0.6235,
+      "step": 555
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1463413225987667e-05,
+      "loss": 0.6243,
+      "step": 556
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.143573370885227e-05,
+      "loss": 0.623,
+      "step": 557
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1408042955662124e-05,
+      "loss": 0.6272,
+      "step": 558
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1380341183125114e-05,
+      "loss": 0.6172,
+      "step": 559
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1352628608035368e-05,
+      "loss": 0.6299,
+      "step": 560
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1324905447271551e-05,
+      "loss": 0.6208,
+      "step": 561
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.129717191779517e-05,
+      "loss": 0.6169,
+      "step": 562
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1269428236648884e-05,
+      "loss": 0.6252,
+      "step": 563
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1241674620954785e-05,
+      "loss": 0.6333,
+      "step": 564
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1213911287912736e-05,
+      "loss": 0.6255,
+      "step": 565
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1186138454798624e-05,
+      "loss": 0.6248,
+      "step": 566
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1158356338962697e-05,
+      "loss": 0.625,
+      "step": 567
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.1130565157827848e-05,
+      "loss": 0.6289,
+      "step": 568
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.1102765128887915e-05,
+      "loss": 0.6255,
+      "step": 569
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.1074956469705976e-05,
+      "loss": 0.6289,
+      "step": 570
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.1047139397912648e-05,
+      "loss": 0.6243,
+      "step": 571
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.1019314131204394e-05,
+      "loss": 0.6313,
+      "step": 572
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0991480887341795e-05,
+      "loss": 0.6294,
+      "step": 573
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0963639884147879e-05,
+      "loss": 0.6309,
+      "step": 574
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0935791339506387e-05,
+      "loss": 0.6243,
+      "step": 575
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0907935471360088e-05,
+      "loss": 0.6255,
+      "step": 576
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.088007249770905e-05,
+      "loss": 0.6216,
+      "step": 577
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0852202636608963e-05,
+      "loss": 0.6279,
+      "step": 578
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0824326106169419e-05,
+      "loss": 0.6218,
+      "step": 579
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0796443124552185e-05,
+      "loss": 0.6216,
+      "step": 580
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0768553909969539e-05,
+      "loss": 0.6169,
+      "step": 581
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0740658680682523e-05,
+      "loss": 0.6228,
+      "step": 582
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.071275765499926e-05,
+      "loss": 0.6274,
+      "step": 583
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0684851051273221e-05,
+      "loss": 0.6338,
+      "step": 584
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0656939087901552e-05,
+      "loss": 0.6323,
+      "step": 585
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0629021983323324e-05,
+      "loss": 0.6257,
+      "step": 586
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.060109995601785e-05,
+      "loss": 0.6262,
+      "step": 587
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.057317322450297e-05,
+      "loss": 0.6321,
+      "step": 588
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0545242007333342e-05,
+      "loss": 0.6245,
+      "step": 589
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0517306523098725e-05,
+      "loss": 0.6233,
+      "step": 590
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0489366990422264e-05,
+      "loss": 0.6257,
+      "step": 591
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0461423627958803e-05,
+      "loss": 0.6282,
+      "step": 592
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0433476654393149e-05,
+      "loss": 0.6272,
+      "step": 593
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.040552628843837e-05,
+      "loss": 0.6245,
+      "step": 594
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0377572748834079e-05,
+      "loss": 0.6272,
+      "step": 595
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0349616254344735e-05,
+      "loss": 0.6353,
+      "step": 596
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0321657023757918e-05,
+      "loss": 0.637,
+      "step": 597
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0293695275882616e-05,
+      "loss": 0.6282,
+      "step": 598
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0265731229547525e-05,
+      "loss": 0.6274,
+      "step": 599
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0237765103599324e-05,
+      "loss": 0.6248,
+      "step": 600
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0209797116900969e-05,
+      "loss": 0.6233,
+      "step": 601
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0181827488329975e-05,
+      "loss": 0.6255,
+      "step": 602
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.0153856436776706e-05,
+      "loss": 0.626,
+      "step": 603
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.0125884181142673e-05,
+      "loss": 0.6262,
+      "step": 604
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.0097910940338796e-05,
+      "loss": 0.6243,
+      "step": 605
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.0069936933283712e-05,
+      "loss": 0.6238,
+      "step": 606
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.0041962378902056e-05,
+      "loss": 0.6218,
+      "step": 607
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.0013987496122744e-05,
+      "loss": 0.6313,
+      "step": 608
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.98601250387726e-06,
+      "loss": 0.6326,
+      "step": 609
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.958037621097947e-06,
+      "loss": 0.6206,
+      "step": 610
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.93006306671629e-06,
+      "loss": 0.6287,
+      "step": 611
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.902089059661206e-06,
+      "loss": 0.6277,
+      "step": 612
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.87411581885733e-06,
+      "loss": 0.6228,
+      "step": 613
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.846143563223294e-06,
+      "loss": 0.6223,
+      "step": 614
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.81817251167003e-06,
+      "loss": 0.6311,
+      "step": 615
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.790202883099036e-06,
+      "loss": 0.6206,
+      "step": 616
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.762234896400678e-06,
+      "loss": 0.6204,
+      "step": 617
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.734268770452478e-06,
+      "loss": 0.623,
+      "step": 618
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.706304724117386e-06,
+      "loss": 0.6292,
+      "step": 619
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.678342976242084e-06,
+      "loss": 0.6277,
+      "step": 620
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.650383745655265e-06,
+      "loss": 0.6228,
+      "step": 621
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.622427251165921e-06,
+      "loss": 0.625,
+      "step": 622
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.594473711561636e-06,
+      "loss": 0.6316,
+      "step": 623
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.566523345606854e-06,
+      "loss": 0.6292,
+      "step": 624
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.538576372041199e-06,
+      "loss": 0.6282,
+      "step": 625
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.510633009577737e-06,
+      "loss": 0.6262,
+      "step": 626
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.48269347690128e-06,
+      "loss": 0.6238,
+      "step": 627
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.45475799266666e-06,
+      "loss": 0.6177,
+      "step": 628
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.426826775497029e-06,
+      "loss": 0.6223,
+      "step": 629
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.398900043982156e-06,
+      "loss": 0.625,
+      "step": 630
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.37097801667668e-06,
+      "loss": 0.616,
+      "step": 631
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.34306091209845e-06,
+      "loss": 0.6233,
+      "step": 632
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.31514894872678e-06,
+      "loss": 0.6228,
+      "step": 633
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.287242345000744e-06,
+      "loss": 0.6333,
+      "step": 634
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.259341319317477e-06,
+      "loss": 0.6321,
+      "step": 635
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.231446090030463e-06,
+      "loss": 0.6174,
+      "step": 636
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.20355687544782e-06,
+      "loss": 0.6245,
+      "step": 637
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.175673893830588e-06,
+      "loss": 0.6184,
+      "step": 638
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.147797363391038e-06,
+      "loss": 0.6226,
+      "step": 639
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.119927502290954e-06,
+      "loss": 0.627,
+      "step": 640
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.092064528639917e-06,
+      "loss": 0.6255,
+      "step": 641
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.064208660493614e-06,
+      "loss": 0.6216,
+      "step": 642
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.036360115852123e-06,
+      "loss": 0.6218,
+      "step": 643
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.00851911265821e-06,
+      "loss": 0.6267,
+      "step": 644
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.980685868795613e-06,
+      "loss": 0.6321,
+      "step": 645
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.952860602087356e-06,
+      "loss": 0.6248,
+      "step": 646
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.925043530294027e-06,
+      "loss": 0.6287,
+      "step": 647
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.897234871112088e-06,
+      "loss": 0.6177,
+      "step": 648
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.869434842172152e-06,
+      "loss": 0.6267,
+      "step": 649
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.841643661037303e-06,
+      "loss": 0.6245,
+      "step": 650
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.813861545201381e-06,
+      "loss": 0.6245,
+      "step": 651
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.786088712087269e-06,
+      "loss": 0.6282,
+      "step": 652
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.758325379045217e-06,
+      "loss": 0.6255,
+      "step": 653
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.730571763351121e-06,
+      "loss": 0.6201,
+      "step": 654
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.702828082204831e-06,
+      "loss": 0.6213,
+      "step": 655
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.67509455272845e-06,
+      "loss": 0.6218,
+      "step": 656
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.647371391964632e-06,
+      "loss": 0.6223,
+      "step": 657
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.619658816874891e-06,
+      "loss": 0.6243,
+      "step": 658
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.591957044337881e-06,
+      "loss": 0.6287,
+      "step": 659
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.564266291147734e-06,
+      "loss": 0.6221,
+      "step": 660
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.536586774012336e-06,
+      "loss": 0.624,
+      "step": 661
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.508918709551643e-06,
+      "loss": 0.6233,
+      "step": 662
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.481262314295979e-06,
+      "loss": 0.6292,
+      "step": 663
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.453617804684347e-06,
+      "loss": 0.6226,
+      "step": 664
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.425985397062733e-06,
+      "loss": 0.6138,
+      "step": 665
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.398365307682406e-06,
+      "loss": 0.6238,
+      "step": 666
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.370757752698243e-06,
+      "loss": 0.6235,
+      "step": 667
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.343162948167022e-06,
+      "loss": 0.625,
+      "step": 668
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.315581110045735e-06,
+      "loss": 0.634,
+      "step": 669
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.2880124541899e-06,
+      "loss": 0.6157,
+      "step": 670
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.260457196351872e-06,
+      "loss": 0.626,
+      "step": 671
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.232915552179152e-06,
+      "loss": 0.624,
+      "step": 672
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.205387737212699e-06,
+      "loss": 0.6064,
+      "step": 673
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.177873966885239e-06,
+      "loss": 0.6179,
+      "step": 674
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.1503744565196e-06,
+      "loss": 0.6172,
+      "step": 675
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.122889421326996e-06,
+      "loss": 0.626,
+      "step": 676
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.095419076405373e-06,
+      "loss": 0.6245,
+      "step": 677
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.067963636737697e-06,
+      "loss": 0.6287,
+      "step": 678
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.040523317190295e-06,
+      "loss": 0.626,
+      "step": 679
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 8.013098332511155e-06,
+      "loss": 0.6182,
+      "step": 680
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.985688897328264e-06,
+      "loss": 0.6248,
+      "step": 681
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.958295226147911e-06,
+      "loss": 0.623,
+      "step": 682
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.93091753335302e-06,
+      "loss": 0.6333,
+      "step": 683
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.903556033201469e-06,
+      "loss": 0.6248,
+      "step": 684
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.876210939824409e-06,
+      "loss": 0.6221,
+      "step": 685
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.848882467224596e-06,
+      "loss": 0.6243,
+      "step": 686
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.821570829274703e-06,
+      "loss": 0.6201,
+      "step": 687
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.794276239715666e-06,
+      "loss": 0.6252,
+      "step": 688
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.76699891215499e-06,
+      "loss": 0.6143,
+      "step": 689
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.739739060065095e-06,
+      "loss": 0.6318,
+      "step": 690
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.712496896781637e-06,
+      "loss": 0.6201,
+      "step": 691
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.685272635501836e-06,
+      "loss": 0.6245,
+      "step": 692
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.658066489282821e-06,
+      "loss": 0.6208,
+      "step": 693
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.630878671039931e-06,
+      "loss": 0.6238,
+      "step": 694
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.603709393545089e-06,
+      "loss": 0.6235,
+      "step": 695
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.576558869425116e-06,
+      "loss": 0.613,
+      "step": 696
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.54942731116006e-06,
+      "loss": 0.6287,
+      "step": 697
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.522314931081552e-06,
+      "loss": 0.6228,
+      "step": 698
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.495221941371129e-06,
+      "loss": 0.6216,
+      "step": 699
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.4681485540585845e-06,
+      "loss": 0.6248,
+      "step": 700
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.441094981020291e-06,
+      "loss": 0.614,
+      "step": 701
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.414061433977566e-06,
+      "loss": 0.6226,
+      "step": 702
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.387048124495005e-06,
+      "loss": 0.6235,
+      "step": 703
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.360055263978811e-06,
+      "loss": 0.6213,
+      "step": 704
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.33308306367517e-06,
+      "loss": 0.6179,
+      "step": 705
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.3061317346685686e-06,
+      "loss": 0.6111,
+      "step": 706
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.279201487880165e-06,
+      "loss": 0.6165,
+      "step": 707
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.2522925340661145e-06,
+      "loss": 0.6238,
+      "step": 708
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.225405083815943e-06,
+      "loss": 0.6243,
+      "step": 709
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.198539347550888e-06,
+      "loss": 0.6306,
+      "step": 710
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.171695535522258e-06,
+      "loss": 0.627,
+      "step": 711
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.144873857809769e-06,
+      "loss": 0.6301,
+      "step": 712
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.118074524319927e-06,
+      "loss": 0.6211,
+      "step": 713
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.09129774478437e-06,
+      "loss": 0.6223,
+      "step": 714
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.064543728758217e-06,
+      "loss": 0.6284,
+      "step": 715
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.037812685618455e-06,
+      "loss": 0.623,
+      "step": 716
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 7.011104824562276e-06,
+      "loss": 0.6221,
+      "step": 717
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.984420354605454e-06,
+      "loss": 0.6194,
+      "step": 718
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.957759484580699e-06,
+      "loss": 0.6204,
+      "step": 719
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.9311224231360345e-06,
+      "loss": 0.6235,
+      "step": 720
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.904509378733154e-06,
+      "loss": 0.6277,
+      "step": 721
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.8779205596457885e-06,
+      "loss": 0.6257,
+      "step": 722
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.851356173958091e-06,
+      "loss": 0.6294,
+      "step": 723
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.8248164295629945e-06,
+      "loss": 0.6167,
+      "step": 724
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.798301534160589e-06,
+      "loss": 0.623,
+      "step": 725
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.771811695256498e-06,
+      "loss": 0.6187,
+      "step": 726
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.745347120160248e-06,
+      "loss": 0.6172,
+      "step": 727
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.718908015983663e-06,
+      "loss": 0.6238,
+      "step": 728
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.692494589639209e-06,
+      "loss": 0.6169,
+      "step": 729
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.66610704783842e-06,
+      "loss": 0.6194,
+      "step": 730
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.639745597090246e-06,
+      "loss": 0.6162,
+      "step": 731
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.613410443699453e-06,
+      "loss": 0.6162,
+      "step": 732
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.587101793765e-06,
+      "loss": 0.626,
+      "step": 733
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.560819853178432e-06,
+      "loss": 0.6223,
+      "step": 734
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.53456482762227e-06,
+      "loss": 0.6133,
+      "step": 735
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.508336922568389e-06,
+      "loss": 0.6167,
+      "step": 736
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.482136343276424e-06,
+      "loss": 0.6228,
+      "step": 737
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.45596329479216e-06,
+      "loss": 0.6204,
+      "step": 738
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.429817981945925e-06,
+      "loss": 0.615,
+      "step": 739
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.403700609350988e-06,
+      "loss": 0.624,
+      "step": 740
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.377611381401958e-06,
+      "loss": 0.6169,
+      "step": 741
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.3515505022731805e-06,
+      "loss": 0.6187,
+      "step": 742
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.325518175917145e-06,
+      "loss": 0.6218,
+      "step": 743
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.2995146060628865e-06,
+      "loss": 0.6211,
+      "step": 744
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.2735399962143925e-06,
+      "loss": 0.6323,
+      "step": 745
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.247594549649012e-06,
+      "loss": 0.6226,
+      "step": 746
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.221678469415858e-06,
+      "loss": 0.616,
+      "step": 747
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.1957919583342255e-06,
+      "loss": 0.6216,
+      "step": 748
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.169935218992001e-06,
+      "loss": 0.6304,
+      "step": 749
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.144108453744069e-06,
+      "loss": 0.6267,
+      "step": 750
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.118311864710749e-06,
+      "loss": 0.6223,
+      "step": 751
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.092545653776194e-06,
+      "loss": 0.6287,
+      "step": 752
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.066810022586821e-06,
+      "loss": 0.615,
+      "step": 753
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.04110517254973e-06,
+      "loss": 0.6157,
+      "step": 754
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.0154313048311275e-06,
+      "loss": 0.6289,
+      "step": 755
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.989788620354746e-06,
+      "loss": 0.6211,
+      "step": 756
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.964177319800287e-06,
+      "loss": 0.6228,
+      "step": 757
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.938597603601839e-06,
+      "loss": 0.6223,
+      "step": 758
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.913049671946306e-06,
+      "loss": 0.6162,
+      "step": 759
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.887533724771852e-06,
+      "loss": 0.624,
+      "step": 760
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.862049961766327e-06,
+      "loss": 0.6187,
+      "step": 761
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.836598582365717e-06,
+      "loss": 0.6155,
+      "step": 762
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.811179785752555e-06,
+      "loss": 0.6189,
+      "step": 763
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.7857937708543935e-06,
+      "loss": 0.6311,
+      "step": 764
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.760440736342239e-06,
+      "loss": 0.6104,
+      "step": 765
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.73512088062898e-06,
+      "loss": 0.6213,
+      "step": 766
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.709834401867866e-06,
+      "loss": 0.6157,
+      "step": 767
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.6845814979509206e-06,
+      "loss": 0.6208,
+      "step": 768
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.659362366507426e-06,
+      "loss": 0.6187,
+      "step": 769
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.634177204902341e-06,
+      "loss": 0.6106,
+      "step": 770
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.609026210234799e-06,
+      "loss": 0.6245,
+      "step": 771
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.583909579336528e-06,
+      "loss": 0.6228,
+      "step": 772
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.558827508770326e-06,
+      "loss": 0.6174,
+      "step": 773
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.533780194828533e-06,
+      "loss": 0.616,
+      "step": 774
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.508767833531468e-06,
+      "loss": 0.6274,
+      "step": 775
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.48379062062593e-06,
+      "loss": 0.6221,
+      "step": 776
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.458848751583619e-06,
+      "loss": 0.6143,
+      "step": 777
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.433942421599661e-06,
+      "loss": 0.6162,
+      "step": 778
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.4090718255910395e-06,
+      "loss": 0.6179,
+      "step": 779
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.384237158195082e-06,
+      "loss": 0.6208,
+      "step": 780
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.359438613767951e-06,
+      "loss": 0.6018,
+      "step": 781
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.334676386383093e-06,
+      "loss": 0.5769,
+      "step": 782
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.3099506698297624e-06,
+      "loss": 0.5691,
+      "step": 783
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.2852616576114445e-06,
+      "loss": 0.5781,
+      "step": 784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.2606095429444106e-06,
+      "loss": 0.5745,
+      "step": 785
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.2359945187561516e-06,
+      "loss": 0.5754,
+      "step": 786
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.21141677768389e-06,
+      "loss": 0.5747,
+      "step": 787
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.1868765120730835e-06,
+      "loss": 0.5732,
+      "step": 788
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.162373913975889e-06,
+      "loss": 0.5703,
+      "step": 789
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.137909175149698e-06,
+      "loss": 0.5657,
+      "step": 790
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.113482487055587e-06,
+      "loss": 0.5652,
+      "step": 791
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.089094040856877e-06,
+      "loss": 0.5786,
+      "step": 792
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.064744027417586e-06,
+      "loss": 0.5664,
+      "step": 793
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.0404326373009734e-06,
+      "loss": 0.5698,
+      "step": 794
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.016160060768021e-06,
+      "loss": 0.5708,
+      "step": 795
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.991926487775953e-06,
+      "loss": 0.5752,
+      "step": 796
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.9677321079767726e-06,
+      "loss": 0.5662,
+      "step": 797
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.943577110715726e-06,
+      "loss": 0.5674,
+      "step": 798
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.919461685029879e-06,
+      "loss": 0.5662,
+      "step": 799
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.895386019646593e-06,
+      "loss": 0.5576,
+      "step": 800
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.871350302982077e-06,
+      "loss": 0.5581,
+      "step": 801
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.847354723139893e-06,
+      "loss": 0.5681,
+      "step": 802
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.82339946790949e-06,
+      "loss": 0.5669,
+      "step": 803
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.799484724764753e-06,
+      "loss": 0.5688,
+      "step": 804
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.775610680862492e-06,
+      "loss": 0.5635,
+      "step": 805
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.751777523041033e-06,
+      "loss": 0.5659,
+      "step": 806
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.7279854378187055e-06,
+      "loss": 0.5674,
+      "step": 807
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.704234611392422e-06,
+      "loss": 0.5664,
+      "step": 808
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.6805252296361915e-06,
+      "loss": 0.5813,
+      "step": 809
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.6568574780996775e-06,
+      "loss": 0.564,
+      "step": 810
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.633231542006761e-06,
+      "loss": 0.563,
+      "step": 811
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.609647606254046e-06,
+      "loss": 0.5615,
+      "step": 812
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.586105855409472e-06,
+      "loss": 0.5625,
+      "step": 813
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.562606473710817e-06,
+      "loss": 0.5613,
+      "step": 814
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.539149645064295e-06,
+      "loss": 0.5667,
+      "step": 815
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.515735553043088e-06,
+      "loss": 0.5686,
+      "step": 816
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.492364380885923e-06,
+      "loss": 0.5674,
+      "step": 817
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.469036311495642e-06,
+      "loss": 0.5703,
+      "step": 818
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.445751527437761e-06,
+      "loss": 0.573,
+      "step": 819
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.42251021093904e-06,
+      "loss": 0.5623,
+      "step": 820
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.3993125438860674e-06,
+      "loss": 0.572,
+      "step": 821
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.3761587078238335e-06,
+      "loss": 0.5671,
+      "step": 822
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.353048883954299e-06,
+      "loss": 0.575,
+      "step": 823
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.3299832531349935e-06,
+      "loss": 0.571,
+      "step": 824
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.306961995877588e-06,
+      "loss": 0.5764,
+      "step": 825
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.2839852923464835e-06,
+      "loss": 0.5764,
+      "step": 826
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.261053322357406e-06,
+      "loss": 0.5708,
+      "step": 827
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.238166265375994e-06,
+      "loss": 0.5693,
+      "step": 828
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.215324300516404e-06,
+      "loss": 0.5583,
+      "step": 829
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.192527606539893e-06,
+      "loss": 0.5696,
+      "step": 830
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.169776361853437e-06,
+      "loss": 0.5601,
+      "step": 831
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.147070744508316e-06,
+      "loss": 0.5688,
+      "step": 832
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.124410932198736e-06,
+      "loss": 0.5647,
+      "step": 833
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.10179710226043e-06,
+      "loss": 0.5688,
+      "step": 834
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.0792294316692725e-06,
+      "loss": 0.574,
+      "step": 835
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.056708097039902e-06,
+      "loss": 0.5669,
+      "step": 836
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.034233274624322e-06,
+      "loss": 0.5618,
+      "step": 837
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.01180514031054e-06,
+      "loss": 0.571,
+      "step": 838
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.989423869621175e-06,
+      "loss": 0.574,
+      "step": 839
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.967089637712095e-06,
+      "loss": 0.5659,
+      "step": 840
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.944802619371041e-06,
+      "loss": 0.5698,
+      "step": 841
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.922562989016258e-06,
+      "loss": 0.5718,
+      "step": 842
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.9003709206951425e-06,
+      "loss": 0.5623,
+      "step": 843
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.8782265880828575e-06,
+      "loss": 0.5662,
+      "step": 844
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.856130164480998e-06,
+      "loss": 0.5671,
+      "step": 845
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.834081822816213e-06,
+      "loss": 0.5715,
+      "step": 846
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.8120817356388693e-06,
+      "loss": 0.571,
+      "step": 847
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7901300751216863e-06,
+      "loss": 0.5637,
+      "step": 848
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7682270130584087e-06,
+      "loss": 0.5713,
+      "step": 849
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.7463727208624366e-06,
+      "loss": 0.5645,
+      "step": 850
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.7245673695655016e-06,
+      "loss": 0.5681,
+      "step": 851
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.7028111298163292e-06,
+      "loss": 0.5669,
+      "step": 852
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6811041718792916e-06,
+      "loss": 0.5642,
+      "step": 853
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.659446665633083e-06,
+      "loss": 0.5657,
+      "step": 854
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.637838780569384e-06,
+      "loss": 0.5703,
+      "step": 855
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.6162806857915524e-06,
+      "loss": 0.5593,
+      "step": 856
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5947725500132747e-06,
+      "loss": 0.5747,
+      "step": 857
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5733145415572603e-06,
+      "loss": 0.5669,
+      "step": 858
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.55190682835393e-06,
+      "loss": 0.5718,
+      "step": 859
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.530549577940087e-06,
+      "loss": 0.5786,
+      "step": 860
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.5092429574576113e-06,
+      "loss": 0.5698,
+      "step": 861
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.487987133652154e-06,
+      "loss": 0.5623,
+      "step": 862
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.466782272871838e-06,
+      "loss": 0.5681,
+      "step": 863
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.445628541065943e-06,
+      "loss": 0.5637,
+      "step": 864
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.4245261037836087e-06,
+      "loss": 0.5598,
+      "step": 865
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.4034751261725583e-06,
+      "loss": 0.5691,
+      "step": 866
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.382475772977778e-06,
+      "loss": 0.5693,
+      "step": 867
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.361528208540248e-06,
+      "loss": 0.5566,
+      "step": 868
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.340632596795642e-06,
+      "loss": 0.5635,
+      "step": 869
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.3197891012730653e-06,
+      "loss": 0.5654,
+      "step": 870
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2989978850937474e-06,
+      "loss": 0.5693,
+      "step": 871
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2782591109697914e-06,
+      "loss": 0.5674,
+      "step": 872
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.2575729412028813e-06,
+      "loss": 0.5732,
+      "step": 873
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.2369395376830203e-06,
+      "loss": 0.5742,
+      "step": 874
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.216359061887263e-06,
+      "loss": 0.5696,
+      "step": 875
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1958316748784514e-06,
+      "loss": 0.5686,
+      "step": 876
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.17535753730396e-06,
+      "loss": 0.5603,
+      "step": 877
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.1549368093944222e-06,
+      "loss": 0.5691,
+      "step": 878
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.134569650962501e-06,
+      "loss": 0.5698,
+      "step": 879
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.1142562214016146e-06,
+      "loss": 0.5691,
+      "step": 880
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0939966796846998e-06,
+      "loss": 0.5671,
+      "step": 881
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0737911843629688e-06,
+      "loss": 0.5679,
+      "step": 882
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.053639893564665e-06,
+      "loss": 0.563,
+      "step": 883
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.033542964993834e-06,
+      "loss": 0.5732,
+      "step": 884
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 3.0135005559290707e-06,
+      "loss": 0.5657,
+      "step": 885
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9935128232223142e-06,
+      "loss": 0.5664,
+      "step": 886
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9735799232975948e-06,
+      "loss": 0.561,
+      "step": 887
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.953702012149826e-06,
+      "loss": 0.5674,
+      "step": 888
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.933879245343578e-06,
+      "loss": 0.5647,
+      "step": 889
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.9141117780118575e-06,
+      "loss": 0.5654,
+      "step": 890
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8943997648549105e-06,
+      "loss": 0.5664,
+      "step": 891
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.874743360138982e-06,
+      "loss": 0.5579,
+      "step": 892
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8551427176951373e-06,
+      "loss": 0.5669,
+      "step": 893
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.835597990918042e-06,
+      "loss": 0.5654,
+      "step": 894
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8161093327647627e-06,
+      "loss": 0.5654,
+      "step": 895
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.796676895753573e-06,
+      "loss": 0.5649,
+      "step": 896
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.777300831962768e-06,
+      "loss": 0.5686,
+      "step": 897
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7579812930294567e-06,
+      "loss": 0.5647,
+      "step": 898
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.738718430148386e-06,
+      "loss": 0.5664,
+      "step": 899
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.7195123940707635e-06,
+      "loss": 0.5603,
+      "step": 900
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.700363335103057e-06,
+      "loss": 0.5686,
+      "step": 901
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.681271403105847e-06,
+      "loss": 0.5676,
+      "step": 902
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6622367474926205e-06,
+      "loss": 0.5659,
+      "step": 903
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.6432595172286347e-06,
+      "loss": 0.5676,
+      "step": 904
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.62433986082973e-06,
+      "loss": 0.5696,
+      "step": 905
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.6054779263611673e-06,
+      "loss": 0.5664,
+      "step": 906
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5866738614364906e-06,
+      "loss": 0.5664,
+      "step": 907
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.56792781321634e-06,
+      "loss": 0.572,
+      "step": 908
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.5492399284073353e-06,
+      "loss": 0.5713,
+      "step": 909
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.530610353260885e-06,
+      "loss": 0.5659,
+      "step": 910
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.51203923357209e-06,
+      "loss": 0.5701,
+      "step": 911
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4935267146785658e-06,
+      "loss": 0.5579,
+      "step": 912
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4750729414593166e-06,
+      "loss": 0.5647,
+      "step": 913
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4566780583336123e-06,
+      "loss": 0.5747,
+      "step": 914
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4383422092598398e-06,
+      "loss": 0.561,
+      "step": 915
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.4200655377343964e-06,
+      "loss": 0.5615,
+      "step": 916
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.4018481867905364e-06,
+      "loss": 0.564,
+      "step": 917
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3836902989972923e-06,
+      "loss": 0.5667,
+      "step": 918
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3655920164583246e-06,
+      "loss": 0.561,
+      "step": 919
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.347553480810827e-06,
+      "loss": 0.562,
+      "step": 920
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3295748332244173e-06,
+      "loss": 0.5696,
+      "step": 921
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.31165621440002e-06,
+      "loss": 0.5681,
+      "step": 922
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.29379776456879e-06,
+      "loss": 0.5686,
+      "step": 923
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2759996234909787e-06,
+      "loss": 0.5635,
+      "step": 924
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2582619304548794e-06,
+      "loss": 0.5728,
+      "step": 925
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2405848242757065e-06,
+      "loss": 0.5693,
+      "step": 926
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.222968443294533e-06,
+      "loss": 0.5652,
+      "step": 927
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.205412925377187e-06,
+      "loss": 0.5627,
+      "step": 928
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.187918407913183e-06,
+      "loss": 0.5645,
+      "step": 929
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.170485027814655e-06,
+      "loss": 0.5645,
+      "step": 930
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.153112921515259e-06,
+      "loss": 0.5691,
+      "step": 931
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1358022249691403e-06,
+      "loss": 0.574,
+      "step": 932
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.118553073649837e-06,
+      "loss": 0.5635,
+      "step": 933
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1013656025492457e-06,
+      "loss": 0.5674,
+      "step": 934
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0842399461765472e-06,
+      "loss": 0.5588,
+      "step": 935
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0671762385571557e-06,
+      "loss": 0.5698,
+      "step": 936
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.05017461323169e-06,
+      "loss": 0.5657,
+      "step": 937
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0332352032548906e-06,
+      "loss": 0.5684,
+      "step": 938
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.016358141194621e-06,
+      "loss": 0.5525,
+      "step": 939
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9995435591307944e-06,
+      "loss": 0.5696,
+      "step": 940
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.982791588654368e-06,
+      "loss": 0.5698,
+      "step": 941
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9661023608662934e-06,
+      "loss": 0.5615,
+      "step": 942
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.949476006376495e-06,
+      "loss": 0.5732,
+      "step": 943
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9329126553028643e-06,
+      "loss": 0.5635,
+      "step": 944
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9164124372702053e-06,
+      "loss": 0.5591,
+      "step": 945
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8999754814092663e-06,
+      "loss": 0.5637,
+      "step": 946
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8836019163556895e-06,
+      "loss": 0.5674,
+      "step": 947
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8672918702490305e-06,
+      "loss": 0.5684,
+      "step": 948
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8510454707317381e-06,
+      "loss": 0.5649,
+      "step": 949
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8348628449481688e-06,
+      "loss": 0.5647,
+      "step": 950
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.8187441195435818e-06,
+      "loss": 0.5598,
+      "step": 951
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.8026894206631517e-06,
+      "loss": 0.5596,
+      "step": 952
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7866988739509838e-06,
+      "loss": 0.5627,
+      "step": 953
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7707726045491247e-06,
+      "loss": 0.5635,
+      "step": 954
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7549107370965935e-06,
+      "loss": 0.5615,
+      "step": 955
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.73911339572839e-06,
+      "loss": 0.5632,
+      "step": 956
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7233807040745442e-06,
+      "loss": 0.5588,
+      "step": 957
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.707712785259128e-06,
+      "loss": 0.5627,
+      "step": 958
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6921097618993031e-06,
+      "loss": 0.5654,
+      "step": 959
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6765717561043614e-06,
+      "loss": 0.5706,
+      "step": 960
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6610988894747626e-06,
+      "loss": 0.5654,
+      "step": 961
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.645691283101196e-06,
+      "loss": 0.5676,
+      "step": 962
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6303490575636139e-06,
+      "loss": 0.5593,
+      "step": 963
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6150723329303065e-06,
+      "loss": 0.572,
+      "step": 964
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5998612287569481e-06,
+      "loss": 0.5613,
+      "step": 965
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5847158640856685e-06,
+      "loss": 0.5579,
+      "step": 966
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.569636357444122e-06,
+      "loss": 0.5605,
+      "step": 967
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.554622826844553e-06,
+      "loss": 0.5625,
+      "step": 968
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5396753897828853e-06,
+      "loss": 0.5608,
+      "step": 969
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.524794163237785e-06,
+      "loss": 0.5647,
+      "step": 970
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.5099792636697653e-06,
+      "loss": 0.564,
+      "step": 971
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4952308070202559e-06,
+      "loss": 0.5642,
+      "step": 972
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4805489087107062e-06,
+      "loss": 0.561,
+      "step": 973
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4659336836416771e-06,
+      "loss": 0.5635,
+      "step": 974
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4513852461919553e-06,
+      "loss": 0.5625,
+      "step": 975
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4369037102176352e-06,
+      "loss": 0.5684,
+      "step": 976
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4224891890512438e-06,
+      "loss": 0.5664,
+      "step": 977
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4081417955008559e-06,
+      "loss": 0.5623,
+      "step": 978
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.393861641849198e-06,
+      "loss": 0.5659,
+      "step": 979
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3796488398527808e-06,
+      "loss": 0.564,
+      "step": 980
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3655035007410178e-06,
+      "loss": 0.5654,
+      "step": 981
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3514257352153625e-06,
+      "loss": 0.5684,
+      "step": 982
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3374156534484352e-06,
+      "loss": 0.5601,
+      "step": 983
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3234733650831576e-06,
+      "loss": 0.5642,
+      "step": 984
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3095989792319118e-06,
+      "loss": 0.5681,
+      "step": 985
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2957926044756653e-06,
+      "loss": 0.5674,
+      "step": 986
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.282054348863132e-06,
+      "loss": 0.5615,
+      "step": 987
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2683843199099255e-06,
+      "loss": 0.5637,
+      "step": 988
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2547826245977235e-06,
+      "loss": 0.561,
+      "step": 989
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2412493693734163e-06,
+      "loss": 0.5515,
+      "step": 990
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2277846601482824e-06,
+      "loss": 0.5615,
+      "step": 991
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2143886022971674e-06,
+      "loss": 0.5608,
+      "step": 992
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2010613006576445e-06,
+      "loss": 0.5625,
+      "step": 993
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.187802859529199e-06,
+      "loss": 0.5654,
+      "step": 994
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.174613382672416e-06,
+      "loss": 0.5691,
+      "step": 995
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1614929733081714e-06,
+      "loss": 0.5688,
+      "step": 996
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1484417341168108e-06,
+      "loss": 0.5637,
+      "step": 997
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1354597672373558e-06,
+      "loss": 0.5618,
+      "step": 998
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.122547174266707e-06,
+      "loss": 0.5586,
+      "step": 999
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1097040562588412e-06,
+      "loss": 0.5627,
+      "step": 1000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0969305137240227e-06,
+      "loss": 0.5703,
+      "step": 1001
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0842266466280193e-06,
+      "loss": 0.552,
+      "step": 1002
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.071592554391322e-06,
+      "loss": 0.5671,
+      "step": 1003
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0590283358883568e-06,
+      "loss": 0.5659,
+      "step": 1004
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0465340894467279e-06,
+      "loss": 0.5645,
+      "step": 1005
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.034109912846427e-06,
+      "loss": 0.5669,
+      "step": 1006
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0217559033190849e-06,
+      "loss": 0.5693,
+      "step": 1007
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0094721575472055e-06,
+      "loss": 0.5608,
+      "step": 1008
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.97258771663402e-07,
+      "loss": 0.5654,
+      "step": 1009
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.85115841249661e-07,
+      "loss": 0.5701,
+      "step": 1010
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.730434613365747e-07,
+      "loss": 0.5569,
+      "step": 1011
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.61041726402616e-07,
+      "loss": 0.5662,
+      "step": 1012
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.491107303733826e-07,
+      "loss": 0.5505,
+      "step": 1013
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.372505666208731e-07,
+      "loss": 0.5652,
+      "step": 1014
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.254613279627522e-07,
+      "loss": 0.5632,
+      "step": 1015
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.137431066616198e-07,
+      "loss": 0.5598,
+      "step": 1016
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.020959944243046e-07,
+      "loss": 0.5652,
+      "step": 1017
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.90520082401124e-07,
+      "loss": 0.5686,
+      "step": 1018
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.790154611851919e-07,
+      "loss": 0.5605,
+      "step": 1019
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.675822208116947e-07,
+      "loss": 0.5605,
+      "step": 1020
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.562204507571914e-07,
+      "loss": 0.5664,
+      "step": 1021
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.449302399389181e-07,
+      "loss": 0.5667,
+      "step": 1022
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.33711676714084e-07,
+      "loss": 0.5615,
+      "step": 1023
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.225648488791893e-07,
+      "loss": 0.5662,
+      "step": 1024
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.114898436693264e-07,
+      "loss": 0.5535,
+      "step": 1025
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.004867477575118e-07,
+      "loss": 0.5583,
+      "step": 1026
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.895556472539922e-07,
+      "loss": 0.5674,
+      "step": 1027
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.786966277055818e-07,
+      "loss": 0.5632,
+      "step": 1028
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.67909774094987e-07,
+      "loss": 0.5625,
+      "step": 1029
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.571951708401471e-07,
+      "loss": 0.5652,
+      "step": 1030
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.465529017935658e-07,
+      "loss": 0.5574,
+      "step": 1031
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.35983050241661e-07,
+      "loss": 0.5625,
+      "step": 1032
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.254856989041126e-07,
+      "loss": 0.5625,
+      "step": 1033
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.150609299332123e-07,
+      "loss": 0.5586,
+      "step": 1034
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.047088249132206e-07,
+      "loss": 0.5691,
+      "step": 1035
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.944294648597294e-07,
+      "loss": 0.5598,
+      "step": 1036
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.842229302190351e-07,
+      "loss": 0.5757,
+      "step": 1037
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.740893008674943e-07,
+      "loss": 0.563,
+      "step": 1038
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.640286561109077e-07,
+      "loss": 0.5554,
+      "step": 1039
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.540410746839054e-07,
+      "loss": 0.564,
+      "step": 1040
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.441266347493158e-07,
+      "loss": 0.5745,
+      "step": 1041
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.34285413897564e-07,
+      "loss": 0.5554,
+      "step": 1042
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.245174891460637e-07,
+      "loss": 0.5535,
+      "step": 1043
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.148229369386138e-07,
+      "loss": 0.5647,
+      "step": 1044
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.052018331447984e-07,
+      "loss": 0.5671,
+      "step": 1045
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.956542530593923e-07,
+      "loss": 0.5698,
+      "step": 1046
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.861802714017783e-07,
+      "loss": 0.5676,
+      "step": 1047
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.767799623153525e-07,
+      "loss": 0.5679,
+      "step": 1048
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.674533993669529e-07,
+      "loss": 0.5657,
+      "step": 1049
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.582006555462782e-07,
+      "loss": 0.5635,
+      "step": 1050
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.490218032653205e-07,
+      "loss": 0.5649,
+      "step": 1051
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.399169143577943e-07,
+      "loss": 0.5635,
+      "step": 1052
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.30886060078577e-07,
+      "loss": 0.5598,
+      "step": 1053
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.219293111031531e-07,
+      "loss": 0.561,
+      "step": 1054
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.130467375270576e-07,
+      "loss": 0.5596,
+      "step": 1055
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.042384088653263e-07,
+      "loss": 0.5652,
+      "step": 1056
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.955043940519555e-07,
+      "loss": 0.5608,
+      "step": 1057
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.868447614393646e-07,
+      "loss": 0.5637,
+      "step": 1058
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.782595787978528e-07,
+      "loss": 0.5588,
+      "step": 1059
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.697489133150768e-07,
+      "loss": 0.564,
+      "step": 1060
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.6131283159552175e-07,
+      "loss": 0.572,
+      "step": 1061
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.5295139965997883e-07,
+      "loss": 0.5605,
+      "step": 1062
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.4466468294502963e-07,
+      "loss": 0.5547,
+      "step": 1063
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.364527463025336e-07,
+      "loss": 0.5676,
+      "step": 1064
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.283156539991251e-07,
+      "loss": 0.564,
+      "step": 1065
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.2025346971570346e-07,
+      "loss": 0.5698,
+      "step": 1066
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.1226625654693817e-07,
+      "loss": 0.5642,
+      "step": 1067
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.043540770007748e-07,
+      "loss": 0.5623,
+      "step": 1068
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.965169929979451e-07,
+      "loss": 0.5774,
+      "step": 1069
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.887550658714845e-07,
+      "loss": 0.5642,
+      "step": 1070
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.810683563662476e-07,
+      "loss": 0.5645,
+      "step": 1071
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.734569246384401e-07,
+      "loss": 0.563,
+      "step": 1072
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.659208302551387e-07,
+      "loss": 0.5623,
+      "step": 1073
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.5846013219383523e-07,
+      "loss": 0.5669,
+      "step": 1074
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.510748888419657e-07,
+      "loss": 0.5593,
+      "step": 1075
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.4376515799645736e-07,
+      "loss": 0.5627,
+      "step": 1076
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.365309968632802e-07,
+      "loss": 0.5569,
+      "step": 1077
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.2937246205699047e-07,
+      "loss": 0.5537,
+      "step": 1078
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.222896096002981e-07,
+      "loss": 0.5632,
+      "step": 1079
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.1528249492361663e-07,
+      "loss": 0.5681,
+      "step": 1080
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.0835117286464154e-07,
+      "loss": 0.5596,
+      "step": 1081
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.0149569766791174e-07,
+      "loss": 0.5615,
+      "step": 1082
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.947161229843909e-07,
+      "loss": 0.5679,
+      "step": 1083
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.8801250187104e-07,
+      "loss": 0.5706,
+      "step": 1084
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.8138488679041344e-07,
+      "loss": 0.5667,
+      "step": 1085
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.748333296102379e-07,
+      "loss": 0.5615,
+      "step": 1086
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.683578816030119e-07,
+      "loss": 0.563,
+      "step": 1087
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.61958593445607e-07,
+      "loss": 0.5657,
+      "step": 1088
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.556355152188605e-07,
+      "loss": 0.5659,
+      "step": 1089
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.493886964072001e-07,
+      "loss": 0.5574,
+      "step": 1090
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.432181858982352e-07,
+      "loss": 0.5586,
+      "step": 1091
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3712403198239753e-07,
+      "loss": 0.5574,
+      "step": 1092
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3110628235254673e-07,
+      "loss": 0.5681,
+      "step": 1093
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2516498410359966e-07,
+      "loss": 0.5596,
+      "step": 1094
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.1930018373217177e-07,
+      "loss": 0.5608,
+      "step": 1095
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.135119271361996e-07,
+      "loss": 0.5552,
+      "step": 1096
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0780025961459226e-07,
+      "loss": 0.563,
+      "step": 1097
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0216522586686827e-07,
+      "loss": 0.5671,
+      "step": 1098
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.9660686999281365e-07,
+      "loss": 0.5649,
+      "step": 1099
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.9112523549213336e-07,
+      "loss": 0.5596,
+      "step": 1100
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.857203652641071e-07,
+      "loss": 0.5623,
+      "step": 1101
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.8039230160726174e-07,
+      "loss": 0.5603,
+      "step": 1102
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.751410862190317e-07,
+      "loss": 0.5632,
+      "step": 1103
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.6996676019544134e-07,
+      "loss": 0.5566,
+      "step": 1104
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.6486936403077193e-07,
+      "loss": 0.5562,
+      "step": 1105
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.5984893761725962e-07,
+      "loss": 0.5615,
+      "step": 1106
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.5490552024476801e-07,
+      "loss": 0.5635,
+      "step": 1107
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.500391506004939e-07,
+      "loss": 0.5671,
+      "step": 1108
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4524986676865415e-07,
+      "loss": 0.5537,
+      "step": 1109
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4053770623019158e-07,
+      "loss": 0.5605,
+      "step": 1110
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3590270586248623e-07,
+      "loss": 0.5593,
+      "step": 1111
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.313449019390578e-07,
+      "loss": 0.5571,
+      "step": 1112
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2686433012928933e-07,
+      "loss": 0.5635,
+      "step": 1113
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2246102549814398e-07,
+      "loss": 0.5564,
+      "step": 1114
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1813502250589304e-07,
+      "loss": 0.5613,
+      "step": 1115
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1388635500784285e-07,
+      "loss": 0.5615,
+      "step": 1116
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.0971505625407498e-07,
+      "loss": 0.563,
+      "step": 1117
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.0562115888918089e-07,
+      "loss": 0.5625,
+      "step": 1118
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0160469495200887e-07,
+      "loss": 0.5627,
+      "step": 1119
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.766569587541297e-08,
+      "loss": 0.5637,
+      "step": 1120
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.38041924860078e-08,
+      "loss": 0.5603,
+      "step": 1121
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.002021500392644e-08,
+      "loss": 0.5649,
+      "step": 1122
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.631379304258391e-08,
+      "loss": 0.5669,
+      "step": 1123
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.268495560844636e-08,
+      "loss": 0.5686,
+      "step": 1124
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.913373110080113e-08,
+      "loss": 0.5706,
+      "step": 1125
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.566014731153814e-08,
+      "loss": 0.564,
+      "step": 1126
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.226423142493222e-08,
+      "loss": 0.5596,
+      "step": 1127
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.894601001742552e-08,
+      "loss": 0.5586,
+      "step": 1128
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.570550905742657e-08,
+      "loss": 0.5623,
+      "step": 1129
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.254275390510157e-08,
+      "loss": 0.5645,
+      "step": 1130
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.9457769312177835e-08,
+      "loss": 0.5562,
+      "step": 1131
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.645057942175291e-08,
+      "loss": 0.5684,
+      "step": 1132
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.3521207768097996e-08,
+      "loss": 0.5688,
+      "step": 1133
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.0669677276482535e-08,
+      "loss": 0.563,
+      "step": 1134
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.7896010262988845e-08,
+      "loss": 0.5486,
+      "step": 1135
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.520022843434113e-08,
+      "loss": 0.5605,
+      "step": 1136
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.258235288773227e-08,
+      "loss": 0.5598,
+      "step": 1137
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.004240411066285e-08,
+      "loss": 0.5637,
+      "step": 1138
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.7580401980775726e-08,
+      "loss": 0.5588,
+      "step": 1139
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.5196365765703954e-08,
+      "loss": 0.5559,
+      "step": 1140
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.2890314122919766e-08,
+      "loss": 0.5659,
+      "step": 1141
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.066226509958803e-08,
+      "loss": 0.5593,
+      "step": 1142
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.8512236132424155e-08,
+      "loss": 0.5632,
+      "step": 1143
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.6440244047559738e-08,
+      "loss": 0.5618,
+      "step": 1144
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.4446305060408236e-08,
+      "loss": 0.5681,
+      "step": 1145
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.253043477554062e-08,
+      "loss": 0.5645,
+      "step": 1146
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.0692648186559916e-08,
+      "loss": 0.5676,
+      "step": 1147
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.893295967598907e-08,
+      "loss": 0.5608,
+      "step": 1148
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.7251383015153277e-08,
+      "loss": 0.5635,
+      "step": 1149
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.5647931364074497e-08,
+      "loss": 0.5713,
+      "step": 1150
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.412261727136821e-08,
+      "loss": 0.562,
+      "step": 1151
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.2675452674145717e-08,
+      "loss": 0.5625,
+      "step": 1152
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.1306448897919764e-08,
+      "loss": 0.5605,
+      "step": 1153
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0015616656516846e-08,
+      "loss": 0.5583,
+      "step": 1154
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.802966051991713e-09,
+      "loss": 0.5625,
+      "step": 1155
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.66850657455076e-09,
+      "loss": 0.5645,
+      "step": 1156
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 6.612247102477653e-09,
+      "loss": 0.561,
+      "step": 1157
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.634195902057827e-09,
+      "loss": 0.5603,
+      "step": 1158
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.734360627525192e-09,
+      "loss": 0.5583,
+      "step": 1159
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.9127483209899786e-09,
+      "loss": 0.564,
+      "step": 1160
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.1693654123921003e-09,
+      "loss": 0.5583,
+      "step": 1161
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.5042177194478657e-09,
+      "loss": 0.5632,
+      "step": 1162
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.9173104476055694e-09,
+      "loss": 0.5566,
+      "step": 1163
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.4086481900010829e-09,
+      "loss": 0.5588,
+      "step": 1164
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.782349274256587e-10,
+      "loss": 0.5591,
+      "step": 1165
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 6.260740282970635e-10,
+      "loss": 0.5671,
+      "step": 1166
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.5216824862405184e-10,
+      "loss": 0.5696,
+      "step": 1167
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.5651973199526382e-10,
+      "loss": 0.5754,
+      "step": 1168
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.913000955702018e-11,
+      "loss": 0.562,
+      "step": 1169
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.564,
+      "step": 1170
+    },
+    {
+      "epoch": 3.0,
+      "step": 1170,
+      "total_flos": 3053244603432960.0,
+      "train_loss": 0.39684578659188036,
+      "train_runtime": 121926.3369,
+      "train_samples_per_second": 2.461,
+      "train_steps_per_second": 0.01
+    }
+  ],
+  "max_steps": 1170,
+  "num_train_epochs": 3,
+  "total_flos": 3053244603432960.0,
+  "trial_name": null,
+  "trial_params": null
+}