diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7986 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.996608253250424,
+  "eval_steps": 500,
+  "global_step": 2652,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0,
+      "loss": 13.4718,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5e-08,
+      "loss": 13.504,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-08,
+      "loss": 13.7237,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-07,
+      "loss": 13.4448,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5e-07,
+      "loss": 13.4276,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-07,
+      "loss": 13.4904,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.5e-07,
+      "loss": 13.5499,
+      "step": 14
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3e-07,
+      "loss": 13.469,
+      "step": 16
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.5e-07,
+      "loss": 13.3643,
+      "step": 18
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4e-07,
+      "loss": 13.1439,
+      "step": 20
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.5e-07,
+      "loss": 12.9525,
+      "step": 22
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-07,
+      "loss": 13.0609,
+      "step": 24
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.5e-07,
+      "loss": 12.9562,
+      "step": 26
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 6e-07,
+      "loss": 12.6147,
+      "step": 28
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.5e-07,
+      "loss": 12.0172,
+      "step": 30
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 7e-07,
+      "loss": 11.5616,
+      "step": 32
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 7.5e-07,
+      "loss": 11.5432,
+      "step": 34
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 8e-07,
+      "loss": 11.1458,
+      "step": 36
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.499999999999999e-07,
+      "loss": 10.7621,
+      "step": 38
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9e-07,
+      "loss": 10.9786,
+      "step": 40
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.499999999999999e-07,
+      "loss": 10.2569,
+      "step": 42
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-06,
+      "loss": 9.8751,
+      "step": 44
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.05e-06,
+      "loss": 9.2012,
+      "step": 46
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.1e-06,
+      "loss": 9.7962,
+      "step": 48
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.1499999999999998e-06,
+      "loss": 9.4366,
+      "step": 50
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.2e-06,
+      "loss": 8.2082,
+      "step": 52
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.2499999999999999e-06,
+      "loss": 8.3599,
+      "step": 54
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.3e-06,
+      "loss": 7.7925,
+      "step": 56
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.35e-06,
+      "loss": 7.4611,
+      "step": 58
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.4e-06,
+      "loss": 7.4838,
+      "step": 60
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.4499999999999999e-06,
+      "loss": 7.1745,
+      "step": 62
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.5e-06,
+      "loss": 6.8636,
+      "step": 64
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.55e-06,
+      "loss": 6.328,
+      "step": 66
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.6e-06,
+      "loss": 6.1617,
+      "step": 68
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.6499999999999999e-06,
+      "loss": 5.541,
+      "step": 70
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.6999999999999998e-06,
+      "loss": 5.0421,
+      "step": 72
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.75e-06,
+      "loss": 5.902,
+      "step": 74
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8e-06,
+      "loss": 5.2086,
+      "step": 76
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.85e-06,
+      "loss": 4.6743,
+      "step": 78
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8999999999999998e-06,
+      "loss": 5.2676,
+      "step": 80
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.95e-06,
+      "loss": 4.601,
+      "step": 82
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-06,
+      "loss": 4.5361,
+      "step": 84
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.999997016077106e-06,
+      "loss": 4.8133,
+      "step": 86
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9999880643262316e-06,
+      "loss": 4.6242,
+      "step": 88
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9999731448007997e-06,
+      "loss": 4.2251,
+      "step": 90
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9999522575898473e-06,
+      "loss": 4.7076,
+      "step": 92
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9999254028180264e-06,
+      "loss": 5.183,
+      "step": 94
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.999892580645602e-06,
+      "loss": 4.759,
+      "step": 96
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9998537912684517e-06,
+      "loss": 4.6168,
+      "step": 98
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9998090349180645e-06,
+      "loss": 4.3821,
+      "step": 100
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.99975831186154e-06,
+      "loss": 4.39,
+      "step": 102
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.999701622401585e-06,
+      "loss": 3.8744,
+      "step": 104
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.999638966876513e-06,
+      "loss": 4.2736,
+      "step": 106
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9995703456602437e-06,
+      "loss": 4.2618,
+      "step": 108
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.999495759162297e-06,
+      "loss": 4.6974,
+      "step": 110
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.999415207827794e-06,
+      "loss": 4.3729,
+      "step": 112
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9993286921374522e-06,
+      "loss": 4.1522,
+      "step": 114
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9992362126075847e-06,
+      "loss": 4.41,
+      "step": 116
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9991377697900943e-06,
+      "loss": 3.8781,
+      "step": 118
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9990333642724733e-06,
+      "loss": 4.3471,
+      "step": 120
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.998922996677797e-06,
+      "loss": 3.8304,
+      "step": 122
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.998806667664722e-06,
+      "loss": 4.6405,
+      "step": 124
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9986843779274833e-06,
+      "loss": 4.5621,
+      "step": 126
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.998556128195886e-06,
+      "loss": 4.3485,
+      "step": 128
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9984219192353047e-06,
+      "loss": 4.5476,
+      "step": 130
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9982817518466784e-06,
+      "loss": 4.1366,
+      "step": 132
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.998135626866504e-06,
+      "loss": 3.8135,
+      "step": 134
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.997983545166833e-06,
+      "loss": 4.8264,
+      "step": 136
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9978255076552656e-06,
+      "loss": 4.5821,
+      "step": 138
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.997661515274945e-06,
+      "loss": 5.1009,
+      "step": 140
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.997491569004553e-06,
+      "loss": 3.5422,
+      "step": 142
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9973156698583026e-06,
+      "loss": 3.8344,
+      "step": 144
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.997133818885932e-06,
+      "loss": 4.4703,
+      "step": 146
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9969460171727005e-06,
+      "loss": 4.1585,
+      "step": 148
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9967522658393796e-06,
+      "loss": 3.9991,
+      "step": 150
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9965525660422476e-06,
+      "loss": 4.329,
+      "step": 152
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.996346918973082e-06,
+      "loss": 4.0715,
+      "step": 154
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9961353258591523e-06,
+      "loss": 4.1138,
+      "step": 156
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.995917787963214e-06,
+      "loss": 4.3047,
+      "step": 158
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9956943065834995e-06,
+      "loss": 4.166,
+      "step": 160
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9954648830537116e-06,
+      "loss": 4.7719,
+      "step": 162
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9952295187430144e-06,
+      "loss": 3.9149,
+      "step": 164
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9949882150560254e-06,
+      "loss": 4.4385,
+      "step": 166
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.994740973432809e-06,
+      "loss": 4.3416,
+      "step": 168
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.994487795348863e-06,
+      "loss": 4.5966,
+      "step": 170
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.994228682315117e-06,
+      "loss": 4.2079,
+      "step": 172
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.993963635877917e-06,
+      "loss": 4.3397,
+      "step": 174
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.993692657619019e-06,
+      "loss": 4.2443,
+      "step": 176
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9934157491555797e-06,
+      "loss": 3.8927,
+      "step": 178
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9931329121401462e-06,
+      "loss": 4.0237,
+      "step": 180
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9928441482606463e-06,
+      "loss": 4.3192,
+      "step": 182
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.992549459240378e-06,
+      "loss": 3.5739,
+      "step": 184
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.992248846838e-06,
+      "loss": 4.0837,
+      "step": 186
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9919423128475205e-06,
+      "loss": 3.9352,
+      "step": 188
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9916298590982877e-06,
+      "loss": 4.2776,
+      "step": 190
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.991311487454977e-06,
+      "loss": 3.9825,
+      "step": 192
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9909871998175815e-06,
+      "loss": 3.5599,
+      "step": 194
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9906569981213996e-06,
+      "loss": 3.8652,
+      "step": 196
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.990320884337024e-06,
+      "loss": 4.0849,
+      "step": 198
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9899788604703303e-06,
+      "loss": 3.9656,
+      "step": 200
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.989630928562464e-06,
+      "loss": 3.6045,
+      "step": 202
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.989277090689829e-06,
+      "loss": 3.4662,
+      "step": 204
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.988917348964075e-06,
+      "loss": 3.6883,
+      "step": 206
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9885517055320857e-06,
+      "loss": 4.3033,
+      "step": 208
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.988180162575964e-06,
+      "loss": 3.9864,
+      "step": 210
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9878027223130216e-06,
+      "loss": 3.7801,
+      "step": 212
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9874193869957635e-06,
+      "loss": 3.8201,
+      "step": 214
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9870301589118756e-06,
+      "loss": 3.5885,
+      "step": 216
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.986635040384211e-06,
+      "loss": 3.4567,
+      "step": 218
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.986234033770777e-06,
+      "loss": 3.8317,
+      "step": 220
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9858271414647185e-06,
+      "loss": 3.8597,
+      "step": 222
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.985414365894306e-06,
+      "loss": 3.9962,
+      "step": 224
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9849957095229205e-06,
+      "loss": 3.8344,
+      "step": 226
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.984571174849039e-06,
+      "loss": 4.2405,
+      "step": 228
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9841407644062186e-06,
+      "loss": 3.6655,
+      "step": 230
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9837044807630826e-06,
+      "loss": 4.4589,
+      "step": 232
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9832623265233046e-06,
+      "loss": 3.8763,
+      "step": 234
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9828143043255924e-06,
+      "loss": 4.6417,
+      "step": 236
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.9823604168436743e-06,
+      "loss": 4.3164,
+      "step": 238
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.98190066678628e-06,
+      "loss": 3.9327,
+      "step": 240
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.9814350568971276e-06,
+      "loss": 3.7355,
+      "step": 242
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.9809635899549044e-06,
+      "loss": 3.8522,
+      "step": 244
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.9804862687732527e-06,
+      "loss": 3.8598,
+      "step": 246
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.980003096200752e-06,
+      "loss": 3.558,
+      "step": 248
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.9795140751209014e-06,
+      "loss": 3.4079,
+      "step": 250
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.9790192084521032e-06,
+      "loss": 4.322,
+      "step": 252
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.9785184991476455e-06,
+      "loss": 3.9125,
+      "step": 254
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.978011950195685e-06,
+      "loss": 3.7043,
+      "step": 256
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.977499564619226e-06,
+      "loss": 3.6562,
+      "step": 258
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9769813454761078e-06,
+      "loss": 3.8608,
+      "step": 260
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.976457295858982e-06,
+      "loss": 4.3005,
+      "step": 262
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.975927418895296e-06,
+      "loss": 3.3468,
+      "step": 264
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.975391717747274e-06,
+      "loss": 3.7217,
+      "step": 266
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9748501956118977e-06,
+      "loss": 3.1169,
+      "step": 268
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9743028557208875e-06,
+      "loss": 3.6231,
+      "step": 270
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.973749701340684e-06,
+      "loss": 4.0356,
+      "step": 272
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9731907357724263e-06,
+      "loss": 3.749,
+      "step": 274
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9726259623519353e-06,
+      "loss": 3.6537,
+      "step": 276
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9720553844496916e-06,
+      "loss": 3.4349,
+      "step": 278
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9714790054708166e-06,
+      "loss": 4.0901,
+      "step": 280
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9708968288550505e-06,
+      "loss": 3.46,
+      "step": 282
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9703088580767337e-06,
+      "loss": 4.2503,
+      "step": 284
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.969715096644785e-06,
+      "loss": 4.1053,
+      "step": 286
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9691155481026816e-06,
+      "loss": 4.0205,
+      "step": 288
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9685102160284363e-06,
+      "loss": 3.708,
+      "step": 290
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9678991040345775e-06,
+      "loss": 3.3928,
+      "step": 292
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9672822157681278e-06,
+      "loss": 4.2508,
+      "step": 294
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.966659554910581e-06,
+      "loss": 3.5589,
+      "step": 296
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.9660311251778808e-06,
+      "loss": 4.1245,
+      "step": 298
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9653969303203995e-06,
+      "loss": 3.8555,
+      "step": 300
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.964756974122913e-06,
+      "loss": 3.594,
+      "step": 302
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.964111260404583e-06,
+      "loss": 3.627,
+      "step": 304
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9634597930189283e-06,
+      "loss": 3.5641,
+      "step": 306
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.962802575853806e-06,
+      "loss": 3.8781,
+      "step": 308
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.962139612831387e-06,
+      "loss": 3.3285,
+      "step": 310
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9614709079081313e-06,
+      "loss": 3.5192,
+      "step": 312
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9607964650747686e-06,
+      "loss": 3.9396,
+      "step": 314
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.960116288356268e-06,
+      "loss": 3.5475,
+      "step": 316
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9594303818118203e-06,
+      "loss": 3.8398,
+      "step": 318
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9587387495348097e-06,
+      "loss": 3.7307,
+      "step": 320
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.958041395652791e-06,
+      "loss": 3.5379,
+      "step": 322
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.957338324327464e-06,
+      "loss": 4.157,
+      "step": 324
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.956629539754651e-06,
+      "loss": 3.4495,
+      "step": 326
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.9559150461642684e-06,
+      "loss": 3.6444,
+      "step": 328
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9551948478203044e-06,
+      "loss": 3.3748,
+      "step": 330
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9544689490207904e-06,
+      "loss": 3.8947,
+      "step": 332
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9537373540977796e-06,
+      "loss": 3.3734,
+      "step": 334
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9530000674173176e-06,
+      "loss": 3.5926,
+      "step": 336
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.952257093379417e-06,
+      "loss": 3.706,
+      "step": 338
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9515084364180328e-06,
+      "loss": 3.6486,
+      "step": 340
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.950754101001034e-06,
+      "loss": 3.7235,
+      "step": 342
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.9499940916301783e-06,
+      "loss": 4.0214,
+      "step": 344
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.9492284128410843e-06,
+      "loss": 3.2606,
+      "step": 346
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9484570692032046e-06,
+      "loss": 3.3426,
+      "step": 348
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9476800653197994e-06,
+      "loss": 3.8355,
+      "step": 350
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9468974058279084e-06,
+      "loss": 3.0529,
+      "step": 352
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9461090953983223e-06,
+      "loss": 3.5752,
+      "step": 354
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.945315138735556e-06,
+      "loss": 4.0122,
+      "step": 356
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.944515540577821e-06,
+      "loss": 2.9584,
+      "step": 358
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.9437103056969956e-06,
+      "loss": 3.0343,
+      "step": 360
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.9428994388985973e-06,
+      "loss": 3.2726,
+      "step": 362
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.942082945021754e-06,
+      "loss": 3.6547,
+      "step": 364
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.941260828939175e-06,
+      "loss": 3.528,
+      "step": 366
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.940433095557123e-06,
+      "loss": 3.4337,
+      "step": 368
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.939599749815383e-06,
+      "loss": 3.5137,
+      "step": 370
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.938760796687234e-06,
+      "loss": 3.4036,
+      "step": 372
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.9379162411794177e-06,
+      "loss": 3.1954,
+      "step": 374
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.9370660883321126e-06,
+      "loss": 3.665,
+      "step": 376
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.936210343218899e-06,
+      "loss": 3.2216,
+      "step": 378
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9353490109467317e-06,
+      "loss": 3.5219,
+      "step": 380
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.93448209665591e-06,
+      "loss": 3.4054,
+      "step": 382
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.933609605520043e-06,
+      "loss": 3.2842,
+      "step": 384
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.9327315427460243e-06,
+      "loss": 3.6067,
+      "step": 386
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.9318479135739977e-06,
+      "loss": 2.9899,
+      "step": 388
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.930958723277324e-06,
+      "loss": 3.375,
+      "step": 390
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.9300639771625554e-06,
+      "loss": 3.213,
+      "step": 392
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.929163680569398e-06,
+      "loss": 3.3821,
+      "step": 394
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9282578388706834e-06,
+      "loss": 3.3247,
+      "step": 396
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.927346457472335e-06,
+      "loss": 3.6484,
+      "step": 398
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.926429541813336e-06,
+      "loss": 2.9465,
+      "step": 400
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9255070973656983e-06,
+      "loss": 4.158,
+      "step": 402
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.924579129634428e-06,
+      "loss": 3.193,
+      "step": 404
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.9236456441574927e-06,
+      "loss": 3.3612,
+      "step": 406
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.922706646505791e-06,
+      "loss": 3.2942,
+      "step": 408
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.9217621422831153e-06,
+      "loss": 3.1285,
+      "step": 410
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.9208121371261213e-06,
+      "loss": 3.6443,
+      "step": 412
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.919856636704293e-06,
+      "loss": 3.7119,
+      "step": 414
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.91889564671991e-06,
+      "loss": 3.3598,
+      "step": 416
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.917929172908012e-06,
+      "loss": 3.5273,
+      "step": 418
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.916957221036366e-06,
+      "loss": 3.6511,
+      "step": 420
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.9159797969054307e-06,
+      "loss": 3.0714,
+      "step": 422
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9149969063483223e-06,
+      "loss": 3.094,
+      "step": 424
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.914008555230781e-06,
+      "loss": 3.3933,
+      "step": 426
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.9130147494511325e-06,
+      "loss": 3.3078,
+      "step": 428
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.9120154949402576e-06,
+      "loss": 3.474,
+      "step": 430
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.9110107976615527e-06,
+      "loss": 3.2673,
+      "step": 432
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.910000663610896e-06,
+      "loss": 3.2293,
+      "step": 434
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.908985098816612e-06,
+      "loss": 3.6537,
+      "step": 436
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.9079641093394344e-06,
+      "loss": 3.3056,
+      "step": 438
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.9069377012724716e-06,
+      "loss": 3.0124,
+      "step": 440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9059058807411686e-06,
+      "loss": 3.5076,
+      "step": 442
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9048686539032705e-06,
+      "loss": 2.8775,
+      "step": 444
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9038260269487875e-06,
+      "loss": 3.7755,
+      "step": 446
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.902778006099957e-06,
+      "loss": 3.6773,
+      "step": 448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9017245976112051e-06,
+      "loss": 3.3806,
+      "step": 450
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.900665807769111e-06,
+      "loss": 3.0561,
+      "step": 452
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.8996016428923704e-06,
+      "loss": 3.4006,
+      "step": 454
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.898532109331754e-06,
+      "loss": 3.1419,
+      "step": 456
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.8974572134700738e-06,
+      "loss": 3.3714,
+      "step": 458
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.896376961722142e-06,
+      "loss": 3.3205,
+      "step": 460
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.8952913605347352e-06,
+      "loss": 3.2149,
+      "step": 462
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.894200416386553e-06,
+      "loss": 3.4137,
+      "step": 464
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.8931041357881826e-06,
+      "loss": 3.5795,
+      "step": 466
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.892002525282057e-06,
+      "loss": 3.3404,
+      "step": 468
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.8908955914424182e-06,
+      "loss": 3.477,
+      "step": 470
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.8897833408752766e-06,
+      "loss": 3.4584,
+      "step": 472
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.8886657802183718e-06,
+      "loss": 3.1697,
+      "step": 474
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.8875429161411339e-06,
+      "loss": 3.2975,
+      "step": 476
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.886414755344642e-06,
+      "loss": 3.3685,
+      "step": 478
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.8852813045615861e-06,
+      "loss": 3.0643,
+      "step": 480
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.8841425705562255e-06,
+      "loss": 2.9802,
+      "step": 482
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.8829985601243494e-06,
+      "loss": 3.6372,
+      "step": 484
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.8818492800932354e-06,
+      "loss": 3.0764,
+      "step": 486
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.8806947373216099e-06,
+      "loss": 3.1796,
+      "step": 488
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.8795349386996057e-06,
+      "loss": 3.0618,
+      "step": 490
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.8783698911487219e-06,
+      "loss": 3.275,
+      "step": 492
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.877199601621783e-06,
+      "loss": 3.2729,
+      "step": 494
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.8760240771028966e-06,
+      "loss": 3.0195,
+      "step": 496
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.8748433246074114e-06,
+      "loss": 3.0746,
+      "step": 498
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.8736573511818763e-06,
+      "loss": 3.3867,
+      "step": 500
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.872466163903998e-06,
+      "loss": 3.4967,
+      "step": 502
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.871269769882598e-06,
+      "loss": 3.3825,
+      "step": 504
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.8700681762575718e-06,
+      "loss": 3.0541,
+      "step": 506
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.8688613901998449e-06,
+      "loss": 2.9684,
+      "step": 508
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.8676494189113302e-06,
+      "loss": 3.6065,
+      "step": 510
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.866432269624885e-06,
+      "loss": 2.6745,
+      "step": 512
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.8652099496042692e-06,
+      "loss": 2.8921,
+      "step": 514
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.8639824661441002e-06,
+      "loss": 3.4221,
+      "step": 516
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.8627498265698093e-06,
+      "loss": 3.1598,
+      "step": 518
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.8615120382375999e-06,
+      "loss": 3.304,
+      "step": 520
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.8602691085344022e-06,
+      "loss": 3.6469,
+      "step": 522
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.8590210448778285e-06,
+      "loss": 3.029,
+      "step": 524
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.8577678547161302e-06,
+      "loss": 3.1291,
+      "step": 526
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.8565095455281531e-06,
+      "loss": 3.1253,
+      "step": 528
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.8552461248232926e-06,
+      "loss": 3.2768,
+      "step": 530
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.8539776001414484e-06,
+      "loss": 3.3853,
+      "step": 532
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.85270397905298e-06,
+      "loss": 2.9475,
+      "step": 534
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.8514252691586618e-06,
+      "loss": 3.0003,
+      "step": 536
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.8501414780896375e-06,
+      "loss": 3.3764,
+      "step": 538
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.8488526135073742e-06,
+      "loss": 3.1691,
+      "step": 540
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.8475586831036163e-06,
+      "loss": 2.9825,
+      "step": 542
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.8462596946003415e-06,
+      "loss": 3.3984,
+      "step": 544
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.844955655749713e-06,
+      "loss": 3.0431,
+      "step": 546
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.8436465743340334e-06,
+      "loss": 3.3784,
+      "step": 548
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.8423324581656982e-06,
+      "loss": 3.6729,
+      "step": 550
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8410133150871507e-06,
+      "loss": 3.1924,
+      "step": 552
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8396891529708331e-06,
+      "loss": 3.4732,
+      "step": 554
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.838359979719141e-06,
+      "loss": 3.0055,
+      "step": 556
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8370258032643751e-06,
+      "loss": 3.1568,
+      "step": 558
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8356866315686945e-06,
+      "loss": 3.1793,
+      "step": 560
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.83434247262407e-06,
+      "loss": 3.0726,
+      "step": 562
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.832993334452234e-06,
+      "loss": 3.269,
+      "step": 564
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8316392251046362e-06,
+      "loss": 3.0972,
+      "step": 566
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8302801526623915e-06,
+      "loss": 3.6254,
+      "step": 568
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8289161252362349e-06,
+      "loss": 3.1806,
+      "step": 570
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8275471509664718e-06,
+      "loss": 3.1242,
+      "step": 572
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8261732380229295e-06,
+      "loss": 3.3653,
+      "step": 574
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8247943946049084e-06,
+      "loss": 2.9998,
+      "step": 576
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.8234106289411336e-06,
+      "loss": 2.9557,
+      "step": 578
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.8220219492897052e-06,
+      "loss": 2.9404,
+      "step": 580
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.820628363938049e-06,
+      "loss": 2.9931,
+      "step": 582
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.8192298812028677e-06,
+      "loss": 2.8562,
+      "step": 584
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.8178265094300904e-06,
+      "loss": 3.688,
+      "step": 586
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.8164182569948237e-06,
+      "loss": 3.6602,
+      "step": 588
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.8150051323013006e-06,
+      "loss": 3.111,
+      "step": 590
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.8135871437828316e-06,
+      "loss": 3.4345,
+      "step": 592
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.8121642999017534e-06,
+      "loss": 2.7353,
+      "step": 594
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.8107366091493791e-06,
+      "loss": 2.9942,
+      "step": 596
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.8100209488579017e-06,
+      "loss": 3.6281,
+      "step": 598
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.8085860037830547e-06,
+      "loss": 2.9742,
+      "step": 600
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.807146233191631e-06,
+      "loss": 3.1228,
+      "step": 602
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.8057016456759593e-06,
+      "loss": 2.8459,
+      "step": 604
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.804252249857115e-06,
+      "loss": 3.2035,
+      "step": 606
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.802798054384869e-06,
+      "loss": 2.8832,
+      "step": 608
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8013390679376354e-06,
+      "loss": 3.0979,
+      "step": 610
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.7998752992224207e-06,
+      "loss": 3.3985,
+      "step": 612
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.7984067569747706e-06,
+      "loss": 2.9677,
+      "step": 614
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.7969334499587187e-06,
+      "loss": 3.2882,
+      "step": 616
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.7954553869667341e-06,
+      "loss": 2.8713,
+      "step": 618
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.7939725768196693e-06,
+      "loss": 3.0123,
+      "step": 620
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.7924850283667059e-06,
+      "loss": 3.2748,
+      "step": 622
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.7909927504853038e-06,
+      "loss": 3.1904,
+      "step": 624
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.7894957520811475e-06,
+      "loss": 3.0524,
+      "step": 626
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.7879940420880928e-06,
+      "loss": 3.4434,
+      "step": 628
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.7864876294681128e-06,
+      "loss": 3.3113,
+      "step": 630
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.7849765232112461e-06,
+      "loss": 2.9968,
+      "step": 632
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.783460732335542e-06,
+      "loss": 3.321,
+      "step": 634
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.7819402658870062e-06,
+      "loss": 2.4748,
+      "step": 636
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.7804151329395478e-06,
+      "loss": 3.3042,
+      "step": 638
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.7788853425949259e-06,
+      "loss": 3.0145,
+      "step": 640
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.777350903982693e-06,
+      "loss": 2.8297,
+      "step": 642
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.7758118262601415e-06,
+      "loss": 2.8803,
+      "step": 644
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.7742681186122507e-06,
+      "loss": 3.0037,
+      "step": 646
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.7727197902516295e-06,
+      "loss": 3.3117,
+      "step": 648
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.7711668504184628e-06,
+      "loss": 3.0009,
+      "step": 650
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.769609308380456e-06,
+      "loss": 3.4707,
+      "step": 652
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.7680471734327798e-06,
+      "loss": 2.9016,
+      "step": 654
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.7664804548980148e-06,
+      "loss": 2.9012,
+      "step": 656
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.7649091621260955e-06,
+      "loss": 3.3625,
+      "step": 658
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.7633333044942543e-06,
+      "loss": 2.8827,
+      "step": 660
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.7617528914069677e-06,
+      "loss": 3.0527,
+      "step": 662
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.7601679322958964e-06,
+      "loss": 3.2379,
+      "step": 664
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.7585784366198321e-06,
+      "loss": 2.7934,
+      "step": 666
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.7569844138646404e-06,
+      "loss": 3.009,
+      "step": 668
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.7553858735432025e-06,
+      "loss": 3.4461,
+      "step": 670
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.753782825195361e-06,
+      "loss": 2.8499,
+      "step": 672
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.752175278387861e-06,
+      "loss": 2.9445,
+      "step": 674
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.7505632427142945e-06,
+      "loss": 3.0315,
+      "step": 676
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.7489467277950409e-06,
+      "loss": 3.0093,
+      "step": 678
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.7473257432772126e-06,
+      "loss": 3.1375,
+      "step": 680
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.745700298834595e-06,
+      "loss": 2.9345,
+      "step": 682
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.7440704041675903e-06,
+      "loss": 3.6223,
+      "step": 684
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.7424360690031576e-06,
+      "loss": 3.0581,
+      "step": 686
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.7407973030947578e-06,
+      "loss": 3.6681,
+      "step": 688
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.7391541162222932e-06,
+      "loss": 3.0931,
+      "step": 690
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.7375065181920495e-06,
+      "loss": 2.9938,
+      "step": 692
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.7358545188366373e-06,
+      "loss": 3.0917,
+      "step": 694
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.7341981280149347e-06,
+      "loss": 3.0784,
+      "step": 696
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.732537355612026e-06,
+      "loss": 3.2368,
+      "step": 698
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.7308722115391453e-06,
+      "loss": 2.9081,
+      "step": 700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.7292027057336153e-06,
+      "loss": 3.2202,
+      "step": 702
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.7275288481587893e-06,
+      "loss": 3.2725,
+      "step": 704
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.7258506488039912e-06,
+      "loss": 2.7724,
+      "step": 706
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.724168117684456e-06,
+      "loss": 3.2317,
+      "step": 708
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.72248126484127e-06,
+      "loss": 3.4725,
+      "step": 710
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.7207901003413109e-06,
+      "loss": 3.0466,
+      "step": 712
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.719094634277187e-06,
+      "loss": 3.1728,
+      "step": 714
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.717394876767179e-06,
+      "loss": 3.0119,
+      "step": 716
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.7156908379551775e-06,
+      "loss": 3.1355,
+      "step": 718
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.7139825280106232e-06,
+      "loss": 2.7676,
+      "step": 720
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.7122699571284463e-06,
+      "loss": 3.3676,
+      "step": 722
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.710553135529006e-06,
+      "loss": 3.1025,
+      "step": 724
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.7088320734580287e-06,
+      "loss": 2.9067,
+      "step": 726
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.7071067811865474e-06,
+      "loss": 2.7302,
+      "step": 728
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.7053772690108406e-06,
+      "loss": 3.2192,
+      "step": 730
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.7036435472523696e-06,
+      "loss": 2.9505,
+      "step": 732
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.7019056262577193e-06,
+      "loss": 3.1462,
+      "step": 734
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.7001635163985335e-06,
+      "loss": 2.8909,
+      "step": 736
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.6984172280714554e-06,
+      "loss": 3.5523,
+      "step": 738
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.6966667716980649e-06,
+      "loss": 2.8366,
+      "step": 740
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.6949121577248154e-06,
+      "loss": 3.124,
+      "step": 742
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.6931533966229718e-06,
+      "loss": 2.9826,
+      "step": 744
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.6913904988885503e-06,
+      "loss": 2.8664,
+      "step": 746
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.6896234750422517e-06,
+      "loss": 2.9879,
+      "step": 748
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.6878523356294028e-06,
+      "loss": 2.722,
+      "step": 750
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.6860770912198898e-06,
+      "loss": 3.1467,
+      "step": 752
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.684297752408098e-06,
+      "loss": 2.9069,
+      "step": 754
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.6825143298128463e-06,
+      "loss": 3.0215,
+      "step": 756
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.6807268340773267e-06,
+      "loss": 3.0145,
+      "step": 758
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.6789352758690368e-06,
+      "loss": 3.1656,
+      "step": 760
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.6771396658797214e-06,
+      "loss": 3.0245,
+      "step": 762
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.6753400148253027e-06,
+      "loss": 3.1808,
+      "step": 764
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.6735363334458213e-06,
+      "loss": 3.6505,
+      "step": 766
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.6717286325053692e-06,
+      "loss": 3.3838,
+      "step": 768
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.669916922792027e-06,
+      "loss": 2.7696,
+      "step": 770
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.6681012151177986e-06,
+      "loss": 2.8441,
+      "step": 772
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.6662815203185478e-06,
+      "loss": 2.6182,
+      "step": 774
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.6644578492539322e-06,
+      "loss": 2.7576,
+      "step": 776
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.6626302128073402e-06,
+      "loss": 2.8683,
+      "step": 778
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.6607986218858233e-06,
+      "loss": 2.9898,
+      "step": 780
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.6589630874200343e-06,
+      "loss": 3.0682,
+      "step": 782
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.6571236203641595e-06,
+      "loss": 2.7996,
+      "step": 784
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.6552802316958543e-06,
+      "loss": 3.3134,
+      "step": 786
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.653432932416179e-06,
+      "loss": 3.1492,
+      "step": 788
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.65158173354953e-06,
+      "loss": 3.1127,
+      "step": 790
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.6497266461435775e-06,
+      "loss": 2.9384,
+      "step": 792
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.6478676812691959e-06,
+      "loss": 2.8595,
+      "step": 794
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.6460048500204016e-06,
+      "loss": 2.6861,
+      "step": 796
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.6441381635142842e-06,
+      "loss": 2.8855,
+      "step": 798
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.6422676328909408e-06,
+      "loss": 3.2246,
+      "step": 800
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.6403932693134096e-06,
+      "loss": 3.1151,
+      "step": 802
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.6385150839676039e-06,
+      "loss": 3.106,
+      "step": 804
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.6366330880622435e-06,
+      "loss": 2.726,
+      "step": 806
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.63474729282879e-06,
+      "loss": 3.0861,
+      "step": 808
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.6328577095213785e-06,
+      "loss": 2.9671,
+      "step": 810
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.6309643494167509e-06,
+      "loss": 2.8445,
+      "step": 812
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.6290672238141882e-06,
+      "loss": 2.864,
+      "step": 814
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.627166344035444e-06,
+      "loss": 3.6071,
+      "step": 816
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.6252617214246747e-06,
+      "loss": 2.987,
+      "step": 818
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.623353367348375e-06,
+      "loss": 3.0725,
+      "step": 820
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.6214412931953074e-06,
+      "loss": 2.8785,
+      "step": 822
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.619525510376436e-06,
+      "loss": 2.6348,
+      "step": 824
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6176060303248569e-06,
+      "loss": 3.2379,
+      "step": 826
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6156828644957313e-06,
+      "loss": 3.2543,
+      "step": 828
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.6137560243662156e-06,
+      "loss": 3.1434,
+      "step": 830
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.6118255214353958e-06,
+      "loss": 3.2644,
+      "step": 832
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.6098913672242142e-06,
+      "loss": 3.0826,
+      "step": 834
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.607953573275406e-06,
+      "loss": 2.8396,
+      "step": 836
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.6060121511534261e-06,
+      "loss": 2.8427,
+      "step": 838
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.6040671124443824e-06,
+      "loss": 3.0612,
+      "step": 840
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.6021184687559663e-06,
+      "loss": 2.8976,
+      "step": 842
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.6001662317173826e-06,
+      "loss": 3.0533,
+      "step": 844
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.598210412979281e-06,
+      "loss": 2.9777,
+      "step": 846
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.5962510242136853e-06,
+      "loss": 3.2888,
+      "step": 848
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.594288077113927e-06,
+      "loss": 3.1706,
+      "step": 850
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.5923215833945704e-06,
+      "loss": 2.8492,
+      "step": 852
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.5903515547913474e-06,
+      "loss": 3.0864,
+      "step": 854
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.5883780030610847e-06,
+      "loss": 3.1081,
+      "step": 856
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.5864009399816344e-06,
+      "loss": 3.5884,
+      "step": 858
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.5844203773518045e-06,
+      "loss": 2.8961,
+      "step": 860
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.5824363269912877e-06,
+      "loss": 3.2312,
+      "step": 862
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.5804488007405898e-06,
+      "loss": 2.6319,
+      "step": 864
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.5784578104609611e-06,
+      "loss": 3.2546,
+      "step": 866
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.5764633680343251e-06,
+      "loss": 3.0429,
+      "step": 868
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.574465485363206e-06,
+      "loss": 2.845,
+      "step": 870
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.5724641743706598e-06,
+      "loss": 2.6803,
+      "step": 872
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.570459447000202e-06,
+      "loss": 2.7251,
+      "step": 874
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.5684513152157361e-06,
+      "loss": 2.8006,
+      "step": 876
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.566439791001483e-06,
+      "loss": 2.8095,
+      "step": 878
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.5644248863619089e-06,
+      "loss": 3.0018,
+      "step": 880
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.5624066133216544e-06,
+      "loss": 2.8637,
+      "step": 882
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.5603849839254614e-06,
+      "loss": 3.4833,
+      "step": 884
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.5583600102381022e-06,
+      "loss": 2.8546,
+      "step": 886
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.5563317043443081e-06,
+      "loss": 2.832,
+      "step": 888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.5543000783486955e-06,
+      "loss": 2.9643,
+      "step": 890
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.5522651443756946e-06,
+      "loss": 3.2313,
+      "step": 892
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.5502269145694787e-06,
+      "loss": 3.019,
+      "step": 894
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.5481854010938878e-06,
+      "loss": 3.1545,
+      "step": 896
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.5461406161323598e-06,
+      "loss": 2.829,
+      "step": 898
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.5440925718878566e-06,
+      "loss": 3.1711,
+      "step": 900
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.5420412805827898e-06,
+      "loss": 2.9604,
+      "step": 902
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.5399867544589499e-06,
+      "loss": 2.7188,
+      "step": 904
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.5379290057774316e-06,
+      "loss": 2.9056,
+      "step": 906
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.5358680468185624e-06,
+      "loss": 3.2179,
+      "step": 908
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.5338038898818266e-06,
+      "loss": 2.8448,
+      "step": 910
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.5317365472857953e-06,
+      "loss": 2.9409,
+      "step": 912
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.52966603136805e-06,
+      "loss": 3.0139,
+      "step": 914
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.52759235448511e-06,
+      "loss": 3.0152,
+      "step": 916
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.52551552901236e-06,
+      "loss": 2.956,
+      "step": 918
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.5234355673439736e-06,
+      "loss": 3.0707,
+      "step": 920
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.521352481892841e-06,
+      "loss": 2.6729,
+      "step": 922
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.5192662850904955e-06,
+      "loss": 2.9812,
+      "step": 924
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.5171769893870376e-06,
+      "loss": 2.9229,
+      "step": 926
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.5150846072510623e-06,
+      "loss": 2.9769,
+      "step": 928
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.512989151169583e-06,
+      "loss": 3.3119,
+      "step": 930
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.5108906336479586e-06,
+      "loss": 2.8164,
+      "step": 932
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.5087890672098182e-06,
+      "loss": 2.8443,
+      "step": 934
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5066844643969858e-06,
+      "loss": 2.9107,
+      "step": 936
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.5045768377694067e-06,
+      "loss": 3.1381,
+      "step": 938
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.502466199905072e-06,
+      "loss": 2.9275,
+      "step": 940
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.5003525633999425e-06,
+      "loss": 2.986,
+      "step": 942
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.498235940867875e-06,
+      "loss": 2.5823,
+      "step": 944
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.4961163449405466e-06,
+      "loss": 2.6783,
+      "step": 946
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.4939937882673782e-06,
+      "loss": 3.398,
+      "step": 948
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4918682835154614e-06,
+      "loss": 3.247,
+      "step": 950
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4897398433694805e-06,
+      "loss": 2.377,
+      "step": 952
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4876084805316388e-06,
+      "loss": 3.2205,
+      "step": 954
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4854742077215797e-06,
+      "loss": 3.2777,
+      "step": 956
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.483337037676315e-06,
+      "loss": 2.829,
+      "step": 958
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.4811969831501458e-06,
+      "loss": 2.7846,
+      "step": 960
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4790540569145878e-06,
+      "loss": 2.8359,
+      "step": 962
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.476908271758294e-06,
+      "loss": 2.607,
+      "step": 964
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.4747596404869791e-06,
+      "loss": 2.5559,
+      "step": 966
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.4726081759233438e-06,
+      "loss": 2.6534,
+      "step": 968
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.470453890906996e-06,
+      "loss": 2.8959,
+      "step": 970
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.4682967982943772e-06,
+      "loss": 2.8039,
+      "step": 972
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.4661369109586832e-06,
+      "loss": 3.3513,
+      "step": 974
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.4639742417897885e-06,
+      "loss": 2.993,
+      "step": 976
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.4618088036941693e-06,
+      "loss": 3.6406,
+      "step": 978
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.4596406095948261e-06,
+      "loss": 2.7913,
+      "step": 980
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.4574696724312068e-06,
+      "loss": 2.8517,
+      "step": 982
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.4552960051591296e-06,
+      "loss": 3.2277,
+      "step": 984
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.4531196207507064e-06,
+      "loss": 2.9813,
+      "step": 986
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.4509405321942626e-06,
+      "loss": 2.8984,
+      "step": 988
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.448758752494263e-06,
+      "loss": 2.8021,
+      "step": 990
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.446574294671233e-06,
+      "loss": 3.1709,
+      "step": 992
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.4443871717616794e-06,
+      "loss": 3.1339,
+      "step": 994
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.4421973968180142e-06,
+      "loss": 2.6732,
+      "step": 996
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.4400049829084772e-06,
+      "loss": 3.086,
+      "step": 998
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.4378099431170561e-06,
+      "loss": 2.8538,
+      "step": 1000
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.4356122905434106e-06,
+      "loss": 3.1565,
+      "step": 1002
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.4334120383027911e-06,
+      "loss": 2.7204,
+      "step": 1004
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.4312091995259647e-06,
+      "loss": 2.8988,
+      "step": 1006
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.4290037873591332e-06,
+      "loss": 2.7425,
+      "step": 1008
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.426795814963856e-06,
+      "loss": 2.9555,
+      "step": 1010
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.4245852955169726e-06,
+      "loss": 2.6559,
+      "step": 1012
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.4223722422105218e-06,
+      "loss": 2.7681,
+      "step": 1014
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.420156668251664e-06,
+      "loss": 2.9471,
+      "step": 1016
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.4179385868626035e-06,
+      "loss": 2.7926,
+      "step": 1018
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.4157180112805075e-06,
+      "loss": 2.8494,
+      "step": 1020
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.4134949547574292e-06,
+      "loss": 3.3081,
+      "step": 1022
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.4112694305602263e-06,
+      "loss": 2.8013,
+      "step": 1024
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.4090414519704851e-06,
+      "loss": 3.0064,
+      "step": 1026
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.4068110322844375e-06,
+      "loss": 2.7133,
+      "step": 1028
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.4045781848128845e-06,
+      "loss": 3.3862,
+      "step": 1030
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.4023429228811152e-06,
+      "loss": 2.7249,
+      "step": 1032
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.4001052598288282e-06,
+      "loss": 3.0071,
+      "step": 1034
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.3978652090100518e-06,
+      "loss": 3.2306,
+      "step": 1036
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.3956227837930636e-06,
+      "loss": 3.1295,
+      "step": 1038
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.3933779975603112e-06,
+      "loss": 3.0307,
+      "step": 1040
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.3911308637083334e-06,
+      "loss": 2.7784,
+      "step": 1042
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.388881395647678e-06,
+      "loss": 2.8326,
+      "step": 1044
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.3866296068028236e-06,
+      "loss": 3.0111,
+      "step": 1046
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.3843755106120985e-06,
+      "loss": 2.8593,
+      "step": 1048
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.3821191205276016e-06,
+      "loss": 2.6555,
+      "step": 1050
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.3798604500151206e-06,
+      "loss": 2.645,
+      "step": 1052
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.3775995125540532e-06,
+      "loss": 3.0032,
+      "step": 1054
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.375336321637325e-06,
+      "loss": 3.0761,
+      "step": 1056
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.3730708907713108e-06,
+      "loss": 2.8773,
+      "step": 1058
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.3708032334757524e-06,
+      "loss": 2.5142,
+      "step": 1060
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.368533363283679e-06,
+      "loss": 3.1279,
+      "step": 1062
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.366261293741326e-06,
+      "loss": 2.7788,
+      "step": 1064
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.3639870384080537e-06,
+      "loss": 3.0011,
+      "step": 1066
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.3617106108562673e-06,
+      "loss": 2.8576,
+      "step": 1068
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.3594320246713356e-06,
+      "loss": 2.6657,
+      "step": 1070
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.3571512934515091e-06,
+      "loss": 2.8379,
+      "step": 1072
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.3548684308078407e-06,
+      "loss": 2.7702,
+      "step": 1074
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.352583450364102e-06,
+      "loss": 2.5082,
+      "step": 1076
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.3502963657567041e-06,
+      "loss": 2.9535,
+      "step": 1078
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.3480071906346156e-06,
+      "loss": 3.1248,
+      "step": 1080
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.34571593865928e-06,
+      "loss": 2.939,
+      "step": 1082
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.3434226235045362e-06,
+      "loss": 2.8517,
+      "step": 1084
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.341127258856535e-06,
+      "loss": 2.7523,
+      "step": 1086
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.3388298584136592e-06,
+      "loss": 2.7239,
+      "step": 1088
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.3365304358864397e-06,
+      "loss": 2.8057,
+      "step": 1090
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.3342290049974762e-06,
+      "loss": 2.571,
+      "step": 1092
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.3319255794813528e-06,
+      "loss": 2.7409,
+      "step": 1094
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.3296201730845582e-06,
+      "loss": 2.7012,
+      "step": 1096
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.3273127995654014e-06,
+      "loss": 3.1777,
+      "step": 1098
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.325003472693933e-06,
+      "loss": 2.9287,
+      "step": 1100
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.322692206251859e-06,
+      "loss": 2.9347,
+      "step": 1102
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.320379014032461e-06,
+      "loss": 3.1072,
+      "step": 1104
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.3180639098405135e-06,
+      "loss": 3.2279,
+      "step": 1106
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3157469074922015e-06,
+      "loss": 2.9649,
+      "step": 1108
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3134280208150373e-06,
+      "loss": 2.9718,
+      "step": 1110
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3111072636477793e-06,
+      "loss": 2.5961,
+      "step": 1112
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3087846498403486e-06,
+      "loss": 3.1793,
+      "step": 1114
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3064601932537457e-06,
+      "loss": 2.8457,
+      "step": 1116
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.3041339077599696e-06,
+      "loss": 2.8363,
+      "step": 1118
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.3018058072419331e-06,
+      "loss": 2.5942,
+      "step": 1120
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2994759055933806e-06,
+      "loss": 2.8349,
+      "step": 1122
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2971442167188067e-06,
+      "loss": 2.8009,
+      "step": 1124
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.2948107545333704e-06,
+      "loss": 2.6612,
+      "step": 1126
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.2924755329628148e-06,
+      "loss": 3.0204,
+      "step": 1128
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.2901385659433816e-06,
+      "loss": 3.2938,
+      "step": 1130
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.2877998674217295e-06,
+      "loss": 2.6818,
+      "step": 1132
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.2854594513548507e-06,
+      "loss": 2.7761,
+      "step": 1134
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.283117331709988e-06,
+      "loss": 3.1311,
+      "step": 1136
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.2807735224645495e-06,
+      "loss": 3.0373,
+      "step": 1138
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.2784280376060274e-06,
+      "loss": 3.108,
+      "step": 1140
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.2760808911319141e-06,
+      "loss": 2.7939,
+      "step": 1142
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.2737320970496175e-06,
+      "loss": 2.8404,
+      "step": 1144
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.2713816693763786e-06,
+      "loss": 2.9958,
+      "step": 1146
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.2690296221391867e-06,
+      "loss": 2.832,
+      "step": 1148
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.2666759693746978e-06,
+      "loss": 3.1871,
+      "step": 1150
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.2643207251291484e-06,
+      "loss": 2.7197,
+      "step": 1152
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.261963903458273e-06,
+      "loss": 2.8825,
+      "step": 1154
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.2596055184272194e-06,
+      "loss": 2.6513,
+      "step": 1156
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.2572455841104664e-06,
+      "loss": 2.6503,
+      "step": 1158
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.2548841145917374e-06,
+      "loss": 2.9881,
+      "step": 1160
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.252521123963919e-06,
+      "loss": 2.6832,
+      "step": 1162
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.2501566263289744e-06,
+      "loss": 2.8204,
+      "step": 1164
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.2477906357978614e-06,
+      "loss": 2.727,
+      "step": 1166
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.2454231664904462e-06,
+      "loss": 3.0467,
+      "step": 1168
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.2430542325354205e-06,
+      "loss": 3.0145,
+      "step": 1170
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.2406838480702168e-06,
+      "loss": 3.2211,
+      "step": 1172
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.2383120272409243e-06,
+      "loss": 2.8369,
+      "step": 1174
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.2359387842022035e-06,
+      "loss": 2.5051,
+      "step": 1176
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.2335641331172037e-06,
+      "loss": 3.3503,
+      "step": 1178
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.2311880881574754e-06,
+      "loss": 3.0488,
+      "step": 1180
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.2288106635028892e-06,
+      "loss": 2.9949,
+      "step": 1182
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.2264318733415486e-06,
+      "loss": 2.6827,
+      "step": 1184
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.2240517318697063e-06,
+      "loss": 2.903,
+      "step": 1186
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.22167025329168e-06,
+      "loss": 3.3562,
+      "step": 1188
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.2192874518197657e-06,
+      "loss": 2.7463,
+      "step": 1190
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.2169033416741562e-06,
+      "loss": 2.9223,
+      "step": 1192
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.2145179370828527e-06,
+      "loss": 2.8244,
+      "step": 1194
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.2121312522815816e-06,
+      "loss": 2.5914,
+      "step": 1196
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.20974330151371e-06,
+      "loss": 3.044,
+      "step": 1198
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.2073540990301602e-06,
+      "loss": 3.1083,
+      "step": 1200
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.204963659089324e-06,
+      "loss": 3.0048,
+      "step": 1202
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.2025719959569781e-06,
+      "loss": 3.0108,
+      "step": 1204
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.2001791239061994e-06,
+      "loss": 3.0631,
+      "step": 1206
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.197785057217279e-06,
+      "loss": 3.1584,
+      "step": 1208
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.1953898101776382e-06,
+      "loss": 3.051,
+      "step": 1210
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.1929933970817416e-06,
+      "loss": 3.1265,
+      "step": 1212
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.1905958322310132e-06,
+      "loss": 2.9612,
+      "step": 1214
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.1881971299337502e-06,
+      "loss": 2.9475,
+      "step": 1216
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.1857973045050382e-06,
+      "loss": 2.6938,
+      "step": 1218
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.183396370266665e-06,
+      "loss": 3.1884,
+      "step": 1220
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.1809943415470358e-06,
+      "loss": 2.7971,
+      "step": 1222
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.1785912326810877e-06,
+      "loss": 2.8811,
+      "step": 1224
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.1761870580102043e-06,
+      "loss": 3.1646,
+      "step": 1226
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.1737818318821286e-06,
+      "loss": 2.8446,
+      "step": 1228
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.1713755686508794e-06,
+      "loss": 3.2058,
+      "step": 1230
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.1689682826766648e-06,
+      "loss": 2.7236,
+      "step": 1232
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.1665599883257959e-06,
+      "loss": 2.9138,
+      "step": 1234
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.1641506999706026e-06,
+      "loss": 3.158,
+      "step": 1236
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.1617404319893458e-06,
+      "loss": 3.062,
+      "step": 1238
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.1593291987661331e-06,
+      "loss": 3.1808,
+      "step": 1240
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.1569170146908326e-06,
+      "loss": 2.7202,
+      "step": 1242
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.154503894158987e-06,
+      "loss": 3.3373,
+      "step": 1244
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1520898515717274e-06,
+      "loss": 2.9694,
+      "step": 1246
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1496749013356877e-06,
+      "loss": 3.053,
+      "step": 1248
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1472590578629185e-06,
+      "loss": 3.1397,
+      "step": 1250
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1448423355708013e-06,
+      "loss": 3.0751,
+      "step": 1252
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1424247488819617e-06,
+      "loss": 2.4685,
+      "step": 1254
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1400063122241843e-06,
+      "loss": 2.569,
+      "step": 1256
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1375870400303262e-06,
+      "loss": 2.7506,
+      "step": 1258
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1351669467382308e-06,
+      "loss": 2.9374,
+      "step": 1260
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1327460467906417e-06,
+      "loss": 3.1167,
+      "step": 1262
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1303243546351161e-06,
+      "loss": 2.9643,
+      "step": 1264
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.12790188472394e-06,
+      "loss": 2.6368,
+      "step": 1266
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.12547865151404e-06,
+      "loss": 2.8845,
+      "step": 1268
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1230546694668978e-06,
+      "loss": 2.8049,
+      "step": 1270
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.1206299530484647e-06,
+      "loss": 2.8742,
+      "step": 1272
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.1182045167290743e-06,
+      "loss": 3.016,
+      "step": 1274
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.1157783749833568e-06,
+      "loss": 2.8742,
+      "step": 1276
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.113351542290152e-06,
+      "loss": 2.8077,
+      "step": 1278
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.1109240331324229e-06,
+      "loss": 2.9582,
+      "step": 1280
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.1084958619971697e-06,
+      "loss": 2.6966,
+      "step": 1282
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.1060670433753435e-06,
+      "loss": 2.586,
+      "step": 1284
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.1036375917617594e-06,
+      "loss": 2.7536,
+      "step": 1286
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.10120752165501e-06,
+      "loss": 3.1543,
+      "step": 1288
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0987768475573781e-06,
+      "loss": 2.8458,
+      "step": 1290
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0963455839747533e-06,
+      "loss": 2.7247,
+      "step": 1292
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0939137454165405e-06,
+      "loss": 2.7273,
+      "step": 1294
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0914813463955779e-06,
+      "loss": 3.1296,
+      "step": 1296
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0890484014280475e-06,
+      "loss": 2.4097,
+      "step": 1298
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0866149250333895e-06,
+      "loss": 2.7304,
+      "step": 1300
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0841809317342163e-06,
+      "loss": 2.8627,
+      "step": 1302
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.0817464360562242e-06,
+      "loss": 2.6054,
+      "step": 1304
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.079311452528108e-06,
+      "loss": 2.7118,
+      "step": 1306
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0768759956814737e-06,
+      "loss": 2.9143,
+      "step": 1308
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0744400800507525e-06,
+      "loss": 3.0118,
+      "step": 1310
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.0720037201731132e-06,
+      "loss": 2.5814,
+      "step": 1312
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.0695669305883753e-06,
+      "loss": 3.0125,
+      "step": 1314
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.0671297258389242e-06,
+      "loss": 2.6754,
+      "step": 1316
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.0646921204696215e-06,
+      "loss": 3.3871,
+      "step": 1318
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.0622541290277201e-06,
+      "loss": 3.0433,
+      "step": 1320
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.0598157660627764e-06,
+      "loss": 2.8308,
+      "step": 1322
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.0573770461265658e-06,
+      "loss": 2.6814,
+      "step": 1324
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.0549379837729922e-06,
+      "loss": 3.0351,
+      "step": 1326
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.0524985935580032e-06,
+      "loss": 3.2508,
+      "step": 1328
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.050058890039504e-06,
+      "loss": 2.7,
+      "step": 1330
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.0476188877772687e-06,
+      "loss": 2.839,
+      "step": 1332
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.0451786013328542e-06,
+      "loss": 2.7308,
+      "step": 1334
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.0427380452695145e-06,
+      "loss": 2.9305,
+      "step": 1336
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.0402972341521112e-06,
+      "loss": 3.0336,
+      "step": 1338
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.0378561825470286e-06,
+      "loss": 2.7144,
+      "step": 1340
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.0354149050220866e-06,
+      "loss": 2.723,
+      "step": 1342
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.0329734161464529e-06,
+      "loss": 2.8139,
+      "step": 1344
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.0305317304905566e-06,
+      "loss": 3.3544,
+      "step": 1346
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.0280898626260006e-06,
+      "loss": 3.0795,
+      "step": 1348
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.0256478271254766e-06,
+      "loss": 3.2152,
+      "step": 1350
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.0232056385626756e-06,
+      "loss": 2.8333,
+      "step": 1352
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.0207633115122026e-06,
+      "loss": 3.3351,
+      "step": 1354
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0183208605494878e-06,
+      "loss": 3.2257,
+      "step": 1356
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0158783002507028e-06,
+      "loss": 2.6375,
+      "step": 1358
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0134356451926703e-06,
+      "loss": 2.9266,
+      "step": 1360
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0109929099527794e-06,
+      "loss": 2.9206,
+      "step": 1362
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.008550109108897e-06,
+      "loss": 2.9864,
+      "step": 1364
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0061072572392823e-06,
+      "loss": 3.0176,
+      "step": 1366
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0036643689224982e-06,
+      "loss": 2.9777,
+      "step": 1368
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0012214587373254e-06,
+      "loss": 3.1409,
+      "step": 1370
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 9.987785412626747e-07,
+      "loss": 3.1139,
+      "step": 1372
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.96335631077502e-07,
+      "loss": 2.5028,
+      "step": 1374
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.938927427607176e-07,
+      "loss": 3.0438,
+      "step": 1376
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.914498908911027e-07,
+      "loss": 2.6127,
+      "step": 1378
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.890070900472205e-07,
+      "loss": 2.7627,
+      "step": 1380
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.865643548073294e-07,
+      "loss": 3.1435,
+      "step": 1382
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 9.841216997492975e-07,
+      "loss": 2.7834,
+      "step": 1384
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 9.816791394505124e-07,
+      "loss": 2.8447,
+      "step": 1386
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 9.792366884877978e-07,
+      "loss": 3.0814,
+      "step": 1388
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 9.767943614373243e-07,
+      "loss": 2.3932,
+      "step": 1390
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 9.74352172874523e-07,
+      "loss": 2.9254,
+      "step": 1392
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 9.719101373739993e-07,
+      "loss": 3.0355,
+      "step": 1394
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 9.694682695094435e-07,
+      "loss": 2.8689,
+      "step": 1396
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 9.670265838535473e-07,
+      "loss": 2.8432,
+      "step": 1398
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 9.645850949779133e-07,
+      "loss": 2.9826,
+      "step": 1400
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 9.621438174529715e-07,
+      "loss": 2.6587,
+      "step": 1402
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 9.597027658478887e-07,
+      "loss": 2.8838,
+      "step": 1404
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 9.572619547304854e-07,
+      "loss": 3.2349,
+      "step": 1406
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 9.548213986671457e-07,
+      "loss": 2.6617,
+      "step": 1408
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 9.523811122227314e-07,
+      "loss": 2.76,
+      "step": 1410
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 9.499411099604961e-07,
+      "loss": 3.2104,
+      "step": 1412
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 9.475014064419968e-07,
+      "loss": 2.5588,
+      "step": 1414
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 9.450620162270076e-07,
+      "loss": 2.6541,
+      "step": 1416
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 9.426229538734341e-07,
+      "loss": 2.9613,
+      "step": 1418
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 9.401842339372232e-07,
+      "loss": 2.8701,
+      "step": 1420
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 9.377458709722802e-07,
+      "loss": 2.8146,
+      "step": 1422
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 9.353078795303785e-07,
+      "loss": 2.8971,
+      "step": 1424
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 9.32870274161076e-07,
+      "loss": 2.826,
+      "step": 1426
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 9.304330694116246e-07,
+      "loss": 2.4784,
+      "step": 1428
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 9.279962798268869e-07,
+      "loss": 2.9121,
+      "step": 1430
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 9.255599199492477e-07,
+      "loss": 2.7017,
+      "step": 1432
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 9.231240043185263e-07,
+      "loss": 2.5173,
+      "step": 1434
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 9.206885474718924e-07,
+      "loss": 2.5665,
+      "step": 1436
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 9.182535639437757e-07,
+      "loss": 2.4979,
+      "step": 1438
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.158190682657834e-07,
+      "loss": 3.0036,
+      "step": 1440
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.133850749666104e-07,
+      "loss": 2.8472,
+      "step": 1442
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 9.109515985719524e-07,
+      "loss": 3.1484,
+      "step": 1444
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 9.085186536044221e-07,
+      "loss": 2.7544,
+      "step": 1446
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 9.060862545834594e-07,
+      "loss": 2.4876,
+      "step": 1448
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 9.03654416025247e-07,
+      "loss": 2.5342,
+      "step": 1450
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 9.012231524426218e-07,
+      "loss": 3.0598,
+      "step": 1452
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 8.987924783449902e-07,
+      "loss": 2.8366,
+      "step": 1454
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 8.963624082382406e-07,
+      "loss": 2.8295,
+      "step": 1456
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 8.939329566246562e-07,
+      "loss": 2.5202,
+      "step": 1458
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 8.915041380028305e-07,
+      "loss": 2.9637,
+      "step": 1460
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 8.890759668675773e-07,
+      "loss": 2.8707,
+      "step": 1462
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 8.866484577098482e-07,
+      "loss": 2.758,
+      "step": 1464
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 8.842216250166431e-07,
+      "loss": 3.0174,
+      "step": 1466
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 8.817954832709255e-07,
+      "loss": 2.8084,
+      "step": 1468
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 8.793700469515353e-07,
+      "loss": 2.8896,
+      "step": 1470
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 8.769453305331021e-07,
+      "loss": 2.7274,
+      "step": 1472
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 8.745213484859604e-07,
+      "loss": 2.8304,
+      "step": 1474
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 8.720981152760599e-07,
+      "loss": 2.9487,
+      "step": 1476
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 8.696756453648836e-07,
+      "loss": 2.6765,
+      "step": 1478
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 8.672539532093584e-07,
+      "loss": 2.9754,
+      "step": 1480
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 8.648330532617692e-07,
+      "loss": 3.1057,
+      "step": 1482
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 8.62412959969674e-07,
+      "loss": 3.1504,
+      "step": 1484
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 8.599936877758158e-07,
+      "loss": 3.1809,
+      "step": 1486
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 8.575752511180386e-07,
+      "loss": 2.8562,
+      "step": 1488
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 8.551576644291988e-07,
+      "loss": 2.9199,
+      "step": 1490
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 8.527409421370812e-07,
+      "loss": 3.0049,
+      "step": 1492
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 8.503250986643125e-07,
+      "loss": 2.7818,
+      "step": 1494
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 8.479101484282726e-07,
+      "loss": 3.0402,
+      "step": 1496
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 8.45496105841013e-07,
+      "loss": 2.9152,
+      "step": 1498
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 8.430829853091674e-07,
+      "loss": 3.2198,
+      "step": 1500
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 8.40670801233867e-07,
+      "loss": 2.9759,
+      "step": 1502
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 8.382595680106543e-07,
+      "loss": 2.7832,
+      "step": 1504
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 8.358493000293971e-07,
+      "loss": 2.8934,
+      "step": 1506
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 8.334400116742039e-07,
+      "loss": 3.1431,
+      "step": 1508
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 8.310317173233352e-07,
+      "loss": 2.6003,
+      "step": 1510
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 8.286244313491206e-07,
+      "loss": 2.5809,
+      "step": 1512
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 8.262181681178714e-07,
+      "loss": 2.5528,
+      "step": 1514
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 8.238129419897956e-07,
+      "loss": 3.105,
+      "step": 1516
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 8.214087673189123e-07,
+      "loss": 3.015,
+      "step": 1518
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 8.190056584529642e-07,
+      "loss": 2.6137,
+      "step": 1520
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 8.166036297333352e-07,
+      "loss": 2.8402,
+      "step": 1522
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 8.142026954949618e-07,
+      "loss": 2.863,
+      "step": 1524
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 8.118028700662499e-07,
+      "loss": 2.7368,
+      "step": 1526
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 8.094041677689869e-07,
+      "loss": 2.5069,
+      "step": 1528
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 8.070066029182582e-07,
+      "loss": 2.9194,
+      "step": 1530
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 8.046101898223618e-07,
+      "loss": 2.6058,
+      "step": 1532
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 8.022149427827209e-07,
+      "loss": 2.7355,
+      "step": 1534
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 7.998208760938008e-07,
+      "loss": 2.8185,
+      "step": 1536
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 7.974280040430219e-07,
+      "loss": 2.8695,
+      "step": 1538
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 7.950363409106758e-07,
+      "loss": 2.9614,
+      "step": 1540
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 7.926459009698397e-07,
+      "loss": 3.1771,
+      "step": 1542
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 7.902566984862897e-07,
+      "loss": 2.9998,
+      "step": 1544
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 7.878687477184184e-07,
+      "loss": 2.5025,
+      "step": 1546
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 7.854820629171475e-07,
+      "loss": 2.7393,
+      "step": 1548
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 7.830966583258441e-07,
+      "loss": 2.8933,
+      "step": 1550
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 7.807125481802342e-07,
+      "loss": 2.7926,
+      "step": 1552
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 7.783297467083201e-07,
+      "loss": 2.8088,
+      "step": 1554
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 7.759482681302937e-07,
+      "loss": 2.8894,
+      "step": 1556
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 7.735681266584513e-07,
+      "loss": 2.9384,
+      "step": 1558
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 7.711893364971109e-07,
+      "loss": 2.7174,
+      "step": 1560
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 7.688119118425246e-07,
+      "loss": 3.0065,
+      "step": 1562
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 7.664358668827966e-07,
+      "loss": 2.7175,
+      "step": 1564
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 7.640612157977963e-07,
+      "loss": 2.9579,
+      "step": 1566
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 7.616879727590757e-07,
+      "loss": 2.8526,
+      "step": 1568
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 7.593161519297833e-07,
+      "loss": 2.9314,
+      "step": 1570
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 7.569457674645795e-07,
+      "loss": 3.2361,
+      "step": 1572
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 7.54576833509554e-07,
+      "loss": 2.8873,
+      "step": 1574
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 7.522093642021385e-07,
+      "loss": 2.7272,
+      "step": 1576
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 7.498433736710252e-07,
+      "loss": 2.9086,
+      "step": 1578
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 7.474788760360808e-07,
+      "loss": 2.7368,
+      "step": 1580
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 7.451158854082625e-07,
+      "loss": 2.5693,
+      "step": 1582
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 7.427544158895339e-07,
+      "loss": 3.0066,
+      "step": 1584
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 7.403944815727808e-07,
+      "loss": 2.8588,
+      "step": 1586
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 7.380360965417274e-07,
+      "loss": 2.5618,
+      "step": 1588
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 7.356792748708516e-07,
+      "loss": 3.1683,
+      "step": 1590
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 7.333240306253019e-07,
+      "loss": 2.4461,
+      "step": 1592
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 7.309703778608134e-07,
+      "loss": 2.6854,
+      "step": 1594
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 7.286183306236214e-07,
+      "loss": 2.6439,
+      "step": 1596
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 7.262679029503827e-07,
+      "loss": 2.9089,
+      "step": 1598
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 7.239191088680859e-07,
+      "loss": 2.9493,
+      "step": 1600
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 7.215719623939724e-07,
+      "loss": 2.6733,
+      "step": 1602
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 7.192264775354507e-07,
+      "loss": 3.1419,
+      "step": 1604
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 7.16882668290012e-07,
+      "loss": 2.8254,
+      "step": 1606
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 7.145405486451492e-07,
+      "loss": 2.821,
+      "step": 1608
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 7.122001325782707e-07,
+      "loss": 2.646,
+      "step": 1610
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 7.098614340566187e-07,
+      "loss": 2.6237,
+      "step": 1612
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 7.075244670371853e-07,
+      "loss": 2.8235,
+      "step": 1614
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 7.051892454666293e-07,
+      "loss": 2.4762,
+      "step": 1616
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 7.028557832811934e-07,
+      "loss": 2.7275,
+      "step": 1618
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 7.005240944066193e-07,
+      "loss": 2.9431,
+      "step": 1620
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 6.981941927580672e-07,
+      "loss": 2.8502,
+      "step": 1622
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 6.958660922400304e-07,
+      "loss": 2.8384,
+      "step": 1624
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 6.935398067462544e-07,
+      "loss": 2.8361,
+      "step": 1626
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 6.912153501596516e-07,
+      "loss": 2.5719,
+      "step": 1628
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 6.888927363522203e-07,
+      "loss": 2.5443,
+      "step": 1630
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 6.865719791849627e-07,
+      "loss": 2.7307,
+      "step": 1632
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 6.842530925077985e-07,
+      "loss": 2.9377,
+      "step": 1634
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 6.819360901594866e-07,
+      "loss": 3.1469,
+      "step": 1636
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 6.796209859675391e-07,
+      "loss": 3.0058,
+      "step": 1638
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 6.773077937481409e-07,
+      "loss": 2.7583,
+      "step": 1640
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 6.749965273060669e-07,
+      "loss": 3.0511,
+      "step": 1642
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 6.726872004345983e-07,
+      "loss": 2.712,
+      "step": 1644
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 6.70379826915442e-07,
+      "loss": 2.7067,
+      "step": 1646
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 6.680744205186471e-07,
+      "loss": 2.664,
+      "step": 1648
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 6.657709950025241e-07,
+      "loss": 3.0128,
+      "step": 1650
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 6.634695641135603e-07,
+      "loss": 2.8639,
+      "step": 1652
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 6.611701415863409e-07,
+      "loss": 2.5107,
+      "step": 1654
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 6.588727411434649e-07,
+      "loss": 2.9211,
+      "step": 1656
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 6.565773764954639e-07,
+      "loss": 2.6104,
+      "step": 1658
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 6.542840613407202e-07,
+      "loss": 2.9756,
+      "step": 1660
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 6.519928093653843e-07,
+      "loss": 2.8468,
+      "step": 1662
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 6.49703634243296e-07,
+      "loss": 2.8689,
+      "step": 1664
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 6.474165496358981e-07,
+      "loss": 2.8522,
+      "step": 1666
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 6.451315691921593e-07,
+      "loss": 2.5796,
+      "step": 1668
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 6.428487065484908e-07,
+      "loss": 2.9733,
+      "step": 1670
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 6.405679753286644e-07,
+      "loss": 3.1356,
+      "step": 1672
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 6.382893891437328e-07,
+      "loss": 2.8461,
+      "step": 1674
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 6.360129615919465e-07,
+      "loss": 2.919,
+      "step": 1676
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 6.33738706258674e-07,
+      "loss": 2.6785,
+      "step": 1678
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 6.314666367163209e-07,
+      "loss": 3.3216,
+      "step": 1680
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 6.291967665242475e-07,
+      "loss": 2.9245,
+      "step": 1682
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 6.269291092286895e-07,
+      "loss": 2.5513,
+      "step": 1684
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 6.24663678362675e-07,
+      "loss": 2.8984,
+      "step": 1686
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 6.224004874459471e-07,
+      "loss": 2.9254,
+      "step": 1688
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 6.201395499848794e-07,
+      "loss": 2.6489,
+      "step": 1690
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 6.178808794723985e-07,
+      "loss": 2.8017,
+      "step": 1692
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 6.156244893879016e-07,
+      "loss": 2.8137,
+      "step": 1694
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 6.133703931971764e-07,
+      "loss": 2.6998,
+      "step": 1696
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 6.111186043523222e-07,
+      "loss": 2.6918,
+      "step": 1698
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 6.088691362916667e-07,
+      "loss": 3.1488,
+      "step": 1700
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 6.066220024396886e-07,
+      "loss": 2.7749,
+      "step": 1702
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 6.043772162069365e-07,
+      "loss": 2.7309,
+      "step": 1704
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 6.021347909899482e-07,
+      "loss": 2.8535,
+      "step": 1706
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 5.998947401711719e-07,
+      "loss": 2.7431,
+      "step": 1708
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5.976570771188849e-07,
+      "loss": 2.9236,
+      "step": 1710
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5.954218151871156e-07,
+      "loss": 2.7266,
+      "step": 1712
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5.931889677155624e-07,
+      "loss": 2.6797,
+      "step": 1714
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5.909585480295146e-07,
+      "loss": 2.841,
+      "step": 1716
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 5.887305694397735e-07,
+      "loss": 2.8735,
+      "step": 1718
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5.865050452425709e-07,
+      "loss": 2.7291,
+      "step": 1720
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 5.842819887194925e-07,
+      "loss": 3.0589,
+      "step": 1722
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5.820614131373967e-07,
+      "loss": 2.8343,
+      "step": 1724
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 5.798433317483361e-07,
+      "loss": 2.6256,
+      "step": 1726
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5.776277577894785e-07,
+      "loss": 2.83,
+      "step": 1728
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 5.754147044830271e-07,
+      "loss": 2.8623,
+      "step": 1730
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5.732041850361435e-07,
+      "loss": 2.9518,
+      "step": 1732
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 5.709962126408668e-07,
+      "loss": 2.6937,
+      "step": 1734
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5.687908004740353e-07,
+      "loss": 2.6544,
+      "step": 1736
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5.665879616972089e-07,
+      "loss": 2.5396,
+      "step": 1738
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 5.643877094565894e-07,
+      "loss": 2.9101,
+      "step": 1740
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5.621900568829437e-07,
+      "loss": 3.1239,
+      "step": 1742
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 5.599950170915228e-07,
+      "loss": 2.9285,
+      "step": 1744
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5.578026031819858e-07,
+      "loss": 2.872,
+      "step": 1746
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 5.556128282383211e-07,
+      "loss": 2.8598,
+      "step": 1748
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5.534257053287673e-07,
+      "loss": 2.815,
+      "step": 1750
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5.512412475057367e-07,
+      "loss": 2.9105,
+      "step": 1752
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5.490594678057374e-07,
+      "loss": 3.0166,
+      "step": 1754
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 5.468803792492937e-07,
+      "loss": 2.9043,
+      "step": 1756
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5.447039948408699e-07,
+      "loss": 3.0569,
+      "step": 1758
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5.425303275687933e-07,
+      "loss": 2.7908,
+      "step": 1760
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 5.403593904051739e-07,
+      "loss": 2.7985,
+      "step": 1762
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 5.381911963058308e-07,
+      "loss": 2.8058,
+      "step": 1764
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 5.360257582102114e-07,
+      "loss": 2.4029,
+      "step": 1766
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 5.338630890413165e-07,
+      "loss": 2.5988,
+      "step": 1768
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 5.317032017056229e-07,
+      "loss": 3.0387,
+      "step": 1770
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 5.295461090930038e-07,
+      "loss": 2.9505,
+      "step": 1772
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 5.273918240766564e-07,
+      "loss": 3.269,
+      "step": 1774
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.252403595130208e-07,
+      "loss": 2.4454,
+      "step": 1776
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.230917282417058e-07,
+      "loss": 2.8803,
+      "step": 1778
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 5.209459430854124e-07,
+      "loss": 2.9931,
+      "step": 1780
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.18803016849854e-07,
+      "loss": 2.8475,
+      "step": 1782
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 5.16662962323685e-07,
+      "loss": 2.9663,
+      "step": 1784
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.145257922784205e-07,
+      "loss": 2.4356,
+      "step": 1786
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 5.123915194683617e-07,
+      "loss": 2.7208,
+      "step": 1788
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5.102601566305192e-07,
+      "loss": 2.8387,
+      "step": 1790
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 5.081317164845384e-07,
+      "loss": 2.6345,
+      "step": 1792
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 5.060062117326219e-07,
+      "loss": 2.923,
+      "step": 1794
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 5.038836550594538e-07,
+      "loss": 3.0051,
+      "step": 1796
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 5.017640591321252e-07,
+      "loss": 2.6653,
+      "step": 1798
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.996474366000574e-07,
+      "loss": 2.8347,
+      "step": 1800
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 4.975338000949278e-07,
+      "loss": 2.3805,
+      "step": 1802
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 4.95423162230593e-07,
+      "loss": 2.8202,
+      "step": 1804
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 4.933155356030142e-07,
+      "loss": 2.874,
+      "step": 1806
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 4.912109327901821e-07,
+      "loss": 2.5629,
+      "step": 1808
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 4.891093663520413e-07,
+      "loss": 3.2899,
+      "step": 1810
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 4.870108488304174e-07,
+      "loss": 3.0521,
+      "step": 1812
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 4.849153927489377e-07,
+      "loss": 2.5282,
+      "step": 1814
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 4.828230106129623e-07,
+      "loss": 2.7628,
+      "step": 1816
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 4.807337149095045e-07,
+      "loss": 3.2389,
+      "step": 1818
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.786475181071587e-07,
+      "loss": 3.0311,
+      "step": 1820
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.7656443265602687e-07,
+      "loss": 2.9986,
+      "step": 1822
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 4.744844709876401e-07,
+      "loss": 2.8551,
+      "step": 1824
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 4.7240764551488985e-07,
+      "loss": 2.4769,
+      "step": 1826
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 4.703339686319503e-07,
+      "loss": 2.8349,
+      "step": 1828
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 4.6826345271420454e-07,
+      "loss": 2.6749,
+      "step": 1830
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 4.661961101181733e-07,
+      "loss": 2.7625,
+      "step": 1832
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 4.6413195318143774e-07,
+      "loss": 3.217,
+      "step": 1834
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 4.6207099422256835e-07,
+      "loss": 2.6652,
+      "step": 1836
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 4.6001324554105026e-07,
+      "loss": 2.7919,
+      "step": 1838
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 4.5795871941720996e-07,
+      "loss": 3.0125,
+      "step": 1840
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 4.5590742811214335e-07,
+      "loss": 2.8058,
+      "step": 1842
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 4.538593838676401e-07,
+      "loss": 2.6426,
+      "step": 1844
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 4.518145989061124e-07,
+      "loss": 2.7242,
+      "step": 1846
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 4.497730854305217e-07,
+      "loss": 2.7069,
+      "step": 1848
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 4.4773485562430546e-07,
+      "loss": 2.7588,
+      "step": 1850
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 4.4569992165130466e-07,
+      "loss": 2.8127,
+      "step": 1852
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 4.43668295655692e-07,
+      "loss": 3.2283,
+      "step": 1854
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 4.4163998976189776e-07,
+      "loss": 2.7166,
+      "step": 1856
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 4.396150160745384e-07,
+      "loss": 2.9462,
+      "step": 1858
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 4.375933866783458e-07,
+      "loss": 2.7383,
+      "step": 1860
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 4.355751136380909e-07,
+      "loss": 2.7109,
+      "step": 1862
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 4.335602089985171e-07,
+      "loss": 3.0711,
+      "step": 1864
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 4.31548684784264e-07,
+      "loss": 2.7965,
+      "step": 1866
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 4.2954055299979784e-07,
+      "loss": 2.9035,
+      "step": 1868
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 4.275358256293403e-07,
+      "loss": 2.644,
+      "step": 1870
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 4.255345146367938e-07,
+      "loss": 3.3082,
+      "step": 1872
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 4.235366319656749e-07,
+      "loss": 2.8945,
+      "step": 1874
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 4.2154218953903875e-07,
+      "loss": 2.9888,
+      "step": 1876
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 4.1955119925940996e-07,
+      "loss": 2.8864,
+      "step": 1878
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 4.1756367300871255e-07,
+      "loss": 2.535,
+      "step": 1880
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 4.155796226481951e-07,
+      "loss": 2.8278,
+      "step": 1882
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 4.135990600183654e-07,
+      "loss": 2.6605,
+      "step": 1884
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 4.1162199693891555e-07,
+      "loss": 2.5885,
+      "step": 1886
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 4.0964844520865284e-07,
+      "loss": 2.8822,
+      "step": 1888
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 4.0767841660542943e-07,
+      "loss": 2.6706,
+      "step": 1890
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 4.0571192288607305e-07,
+      "loss": 2.6113,
+      "step": 1892
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 4.037489757863145e-07,
+      "loss": 2.7017,
+      "step": 1894
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 4.017895870207193e-07,
+      "loss": 2.6754,
+      "step": 1896
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 3.9983376828261764e-07,
+      "loss": 2.9501,
+      "step": 1898
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 3.9788153124403343e-07,
+      "loss": 2.6901,
+      "step": 1900
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 3.9593288755561737e-07,
+      "loss": 2.7851,
+      "step": 1902
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 3.939878488465739e-07,
+      "loss": 2.9811,
+      "step": 1904
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 3.920464267245941e-07,
+      "loss": 2.5464,
+      "step": 1906
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 3.901086327757859e-07,
+      "loss": 2.7844,
+      "step": 1908
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 3.8817447856460427e-07,
+      "loss": 2.944,
+      "step": 1910
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 3.8624397563378453e-07,
+      "loss": 2.8095,
+      "step": 1912
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 3.843171355042688e-07,
+      "loss": 2.9404,
+      "step": 1914
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 3.823939696751427e-07,
+      "loss": 2.5454,
+      "step": 1916
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 3.804744896235641e-07,
+      "loss": 2.7126,
+      "step": 1918
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 3.7855870680469225e-07,
+      "loss": 2.9984,
+      "step": 1920
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 3.7664663265162487e-07,
+      "loss": 2.7908,
+      "step": 1922
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 3.7473827857532535e-07,
+      "loss": 2.5333,
+      "step": 1924
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 3.728336559645563e-07,
+      "loss": 3.2184,
+      "step": 1926
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 3.709327761858118e-07,
+      "loss": 2.422,
+      "step": 1928
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 3.69035650583249e-07,
+      "loss": 3.0345,
+      "step": 1930
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 3.6714229047862156e-07,
+      "loss": 2.8207,
+      "step": 1932
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 3.652527071712101e-07,
+      "loss": 2.6455,
+      "step": 1934
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 3.633669119377568e-07,
+      "loss": 2.9046,
+      "step": 1936
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 3.614849160323965e-07,
+      "loss": 2.9388,
+      "step": 1938
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 3.596067306865902e-07,
+      "loss": 2.7176,
+      "step": 1940
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 3.5773236710905917e-07,
+      "loss": 3.2415,
+      "step": 1942
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 3.558618364857158e-07,
+      "loss": 2.7344,
+      "step": 1944
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 3.5399514997959847e-07,
+      "loss": 3.1536,
+      "step": 1946
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 3.521323187308043e-07,
+      "loss": 2.9705,
+      "step": 1948
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 3.5027335385642285e-07,
+      "loss": 2.8474,
+      "step": 1950
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 3.484182664504697e-07,
+      "loss": 2.572,
+      "step": 1952
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.465670675838209e-07,
+      "loss": 3.1516,
+      "step": 1954
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.4471976830414564e-07,
+      "loss": 3.0114,
+      "step": 1956
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.428763796358406e-07,
+      "loss": 2.7337,
+      "step": 1958
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.410369125799661e-07,
+      "loss": 2.9596,
+      "step": 1960
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.392013781141767e-07,
+      "loss": 2.9242,
+      "step": 1962
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.3736978719265995e-07,
+      "loss": 2.5161,
+      "step": 1964
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.3554215074606775e-07,
+      "loss": 2.5524,
+      "step": 1966
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.3371847968145197e-07,
+      "loss": 2.929,
+      "step": 1968
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.318987848822016e-07,
+      "loss": 2.6728,
+      "step": 1970
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.300830772079731e-07,
+      "loss": 3.0084,
+      "step": 1972
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.282713674946309e-07,
+      "loss": 2.7006,
+      "step": 1974
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.264636665541789e-07,
+      "loss": 2.726,
+      "step": 1976
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.2465998517469705e-07,
+      "loss": 2.6391,
+      "step": 1978
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.228603341202788e-07,
+      "loss": 2.7296,
+      "step": 1980
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.2106472413096286e-07,
+      "loss": 2.7943,
+      "step": 1982
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.192731659226735e-07,
+      "loss": 2.7679,
+      "step": 1984
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.1748567018715366e-07,
+      "loss": 2.8648,
+      "step": 1986
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.1570224759190234e-07,
+      "loss": 2.8623,
+      "step": 1988
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.1392290878010994e-07,
+      "loss": 2.87,
+      "step": 1990
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.1214766437059703e-07,
+      "loss": 2.4638,
+      "step": 1992
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 3.1037652495774804e-07,
+      "loss": 3.205,
+      "step": 1994
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 3.086095011114498e-07,
+      "loss": 2.5909,
+      "step": 1996
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 3.0684660337702817e-07,
+      "loss": 2.9936,
+      "step": 1998
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 3.050878422751847e-07,
+      "loss": 2.8594,
+      "step": 2000
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 3.0333322830193497e-07,
+      "loss": 2.891,
+      "step": 2002
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 3.0158277192854433e-07,
+      "loss": 2.8304,
+      "step": 2004
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.998364836014665e-07,
+      "loss": 2.9801,
+      "step": 2006
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.98094373742281e-07,
+      "loss": 3.0548,
+      "step": 2008
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.963564527476302e-07,
+      "loss": 3.0822,
+      "step": 2010
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.946227309891598e-07,
+      "loss": 3.1002,
+      "step": 2012
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.9289321881345254e-07,
+      "loss": 2.5651,
+      "step": 2014
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.91167926541971e-07,
+      "loss": 2.7928,
+      "step": 2016
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 2.894468644709941e-07,
+      "loss": 2.6256,
+      "step": 2018
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.8773004287155343e-07,
+      "loss": 2.5386,
+      "step": 2020
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.860174719893768e-07,
+      "loss": 3.026,
+      "step": 2022
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.8430916204482246e-07,
+      "loss": 2.6773,
+      "step": 2024
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 2.8260512323282093e-07,
+      "loss": 2.7538,
+      "step": 2026
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.809053657228131e-07,
+      "loss": 2.8801,
+      "step": 2028
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.792098996586891e-07,
+      "loss": 2.6075,
+      "step": 2030
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 2.7751873515872993e-07,
+      "loss": 2.4356,
+      "step": 2032
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.7583188231554386e-07,
+      "loss": 3.1992,
+      "step": 2034
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.741493511960088e-07,
+      "loss": 2.7167,
+      "step": 2036
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.724711518412108e-07,
+      "loss": 3.2261,
+      "step": 2038
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 2.7079729426638466e-07,
+      "loss": 2.4948,
+      "step": 2040
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.691277884608547e-07,
+      "loss": 2.6697,
+      "step": 2042
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.6746264438797396e-07,
+      "loss": 2.9029,
+      "step": 2044
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.658018719850654e-07,
+      "loss": 2.9799,
+      "step": 2046
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.6414548116336244e-07,
+      "loss": 2.7396,
+      "step": 2048
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.6249348180795073e-07,
+      "loss": 2.9322,
+      "step": 2050
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.6084588377770664e-07,
+      "loss": 2.906,
+      "step": 2052
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.592026969052421e-07,
+      "loss": 2.7256,
+      "step": 2054
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.575639309968425e-07,
+      "loss": 3.139,
+      "step": 2056
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.559295958324097e-07,
+      "loss": 2.8881,
+      "step": 2058
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.542997011654051e-07,
+      "loss": 2.8097,
+      "step": 2060
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.5267425672278733e-07,
+      "loss": 2.6047,
+      "step": 2062
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.510532722049591e-07,
+      "loss": 2.5878,
+      "step": 2064
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.4943675728570567e-07,
+      "loss": 2.9128,
+      "step": 2066
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.478247216121386e-07,
+      "loss": 2.8503,
+      "step": 2068
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.462171748046392e-07,
+      "loss": 3.01,
+      "step": 2070
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.4461412645679746e-07,
+      "loss": 3.1334,
+      "step": 2072
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.4301558613535967e-07,
+      "loss": 2.6883,
+      "step": 2074
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.414215633801678e-07,
+      "loss": 2.5622,
+      "step": 2076
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.398320677041035e-07,
+      "loss": 2.9577,
+      "step": 2078
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.3824710859303222e-07,
+      "loss": 3.0543,
+      "step": 2080
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.3666669550574547e-07,
+      "loss": 2.4341,
+      "step": 2082
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.350908378739047e-07,
+      "loss": 2.8255,
+      "step": 2084
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 2.3351954510198534e-07,
+      "loss": 2.8486,
+      "step": 2086
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 2.3195282656722026e-07,
+      "loss": 2.7911,
+      "step": 2088
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.3039069161954383e-07,
+      "loss": 2.7081,
+      "step": 2090
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.288331495815371e-07,
+      "loss": 2.803,
+      "step": 2092
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.2728020974837036e-07,
+      "loss": 2.9866,
+      "step": 2094
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.2573188138774923e-07,
+      "loss": 2.8164,
+      "step": 2096
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.2418817373985854e-07,
+      "loss": 2.3785,
+      "step": 2098
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.2264909601730708e-07,
+      "loss": 2.8835,
+      "step": 2100
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.2111465740507395e-07,
+      "loss": 2.6537,
+      "step": 2102
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.1958486706045198e-07,
+      "loss": 2.8952,
+      "step": 2104
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.1805973411299406e-07,
+      "loss": 2.8878,
+      "step": 2106
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.165392676644582e-07,
+      "loss": 2.8798,
+      "step": 2108
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.1502347678875365e-07,
+      "loss": 2.8305,
+      "step": 2110
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 2.135123705318873e-07,
+      "loss": 2.7106,
+      "step": 2112
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 2.1200595791190723e-07,
+      "loss": 2.7477,
+      "step": 2114
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 2.1050424791885212e-07,
+      "loss": 2.839,
+      "step": 2116
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.0900724951469617e-07,
+      "loss": 3.1958,
+      "step": 2118
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.0751497163329413e-07,
+      "loss": 2.6208,
+      "step": 2120
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 2.0602742318033072e-07,
+      "loss": 3.0725,
+      "step": 2122
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 2.0454461303326575e-07,
+      "loss": 2.8657,
+      "step": 2124
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 2.0306655004128138e-07,
+      "loss": 2.8774,
+      "step": 2126
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 2.0159324302522962e-07,
+      "loss": 2.8357,
+      "step": 2128
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 2.0012470077757926e-07,
+      "loss": 2.469,
+      "step": 2130
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.9866093206236457e-07,
+      "loss": 2.7227,
+      "step": 2132
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.9720194561513116e-07,
+      "loss": 3.198,
+      "step": 2134
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.9574775014288515e-07,
+      "loss": 2.5335,
+      "step": 2136
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.9429835432404095e-07,
+      "loss": 2.5348,
+      "step": 2138
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.9285376680836895e-07,
+      "loss": 2.9377,
+      "step": 2140
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.914139962169452e-07,
+      "loss": 2.6431,
+      "step": 2142
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1.8997905114209834e-07,
+      "loss": 2.9976,
+      "step": 2144
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1.8854894014735932e-07,
+      "loss": 2.9322,
+      "step": 2146
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 1.8712367176740973e-07,
+      "loss": 2.6202,
+      "step": 2148
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 1.8570325450803215e-07,
+      "loss": 2.5657,
+      "step": 2150
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.8428769684605682e-07,
+      "loss": 2.7867,
+      "step": 2152
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.8287700722931432e-07,
+      "loss": 2.853,
+      "step": 2154
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.814711940765824e-07,
+      "loss": 2.8105,
+      "step": 2156
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.8007026577753702e-07,
+      "loss": 2.6706,
+      "step": 2158
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.7867423069270293e-07,
+      "loss": 2.5742,
+      "step": 2160
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.772830971534013e-07,
+      "loss": 2.667,
+      "step": 2162
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.7589687346170312e-07,
+      "loss": 2.6658,
+      "step": 2164
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.7451556789037757e-07,
+      "loss": 3.429,
+      "step": 2166
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.7313918868284294e-07,
+      "loss": 2.7417,
+      "step": 2168
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1.717677440531189e-07,
+      "loss": 2.9242,
+      "step": 2170
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1.7040124218577433e-07,
+      "loss": 2.7466,
+      "step": 2172
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1.690396912358827e-07,
+      "loss": 2.7001,
+      "step": 2174
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1.676830993289695e-07,
+      "loss": 2.9698,
+      "step": 2176
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1.6633147456096608e-07,
+      "loss": 2.9615,
+      "step": 2178
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.649848249981609e-07,
+      "loss": 2.512,
+      "step": 2180
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.6364315867715085e-07,
+      "loss": 2.8938,
+      "step": 2182
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1.623064836047935e-07,
+      "loss": 2.6545,
+      "step": 2184
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1.6097480775815954e-07,
+      "loss": 2.8343,
+      "step": 2186
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1.5964813908448505e-07,
+      "loss": 2.6101,
+      "step": 2188
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1.5832648550112393e-07,
+      "loss": 2.6439,
+      "step": 2190
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.5700985489550134e-07,
+      "loss": 2.6966,
+      "step": 2192
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.5569825512506552e-07,
+      "loss": 2.8579,
+      "step": 2194
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.5439169401724151e-07,
+      "loss": 2.6781,
+      "step": 2196
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.5309017936938473e-07,
+      "loss": 2.7252,
+      "step": 2198
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.517937189487335e-07,
+      "loss": 2.494,
+      "step": 2200
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1.5050232049236421e-07,
+      "loss": 3.0062,
+      "step": 2202
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1.492159917071435e-07,
+      "loss": 2.6262,
+      "step": 2204
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 1.479347402696828e-07,
+      "loss": 2.8857,
+      "step": 2206
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 1.4665857382629398e-07,
+      "loss": 2.9944,
+      "step": 2208
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1.4538749999294053e-07,
+      "loss": 2.9556,
+      "step": 2210
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1.441215263551958e-07,
+      "loss": 2.8379,
+      "step": 2212
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 1.4286066046819511e-07,
+      "loss": 2.7581,
+      "step": 2214
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 1.4160490985659123e-07,
+      "loss": 2.761,
+      "step": 2216
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1.4035428201451095e-07,
+      "loss": 2.5619,
+      "step": 2218
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1.3910878440550732e-07,
+      "loss": 2.8362,
+      "step": 2220
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1.378684244625189e-07,
+      "loss": 2.7421,
+      "step": 2222
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 1.366332095878221e-07,
+      "loss": 2.8827,
+      "step": 2224
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 1.3540314715298907e-07,
+      "loss": 3.023,
+      "step": 2226
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1.341782444988424e-07,
+      "loss": 2.5154,
+      "step": 2228
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1.3295850893541237e-07,
+      "loss": 2.7627,
+      "step": 2230
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 1.317439477418929e-07,
+      "loss": 2.8211,
+      "step": 2232
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 1.305345681665978e-07,
+      "loss": 2.4411,
+      "step": 2234
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 1.2933037742691776e-07,
+      "loss": 2.781,
+      "step": 2236
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 1.281313827092776e-07,
+      "loss": 3.0131,
+      "step": 2238
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 1.269375911690924e-07,
+      "loss": 2.7534,
+      "step": 2240
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 1.257490099307268e-07,
+      "loss": 2.7227,
+      "step": 2242
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 1.2456564608744968e-07,
+      "loss": 2.9133,
+      "step": 2244
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 1.2338750670139431e-07,
+      "loss": 2.7381,
+      "step": 2246
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 1.2221459880351447e-07,
+      "loss": 2.7767,
+      "step": 2248
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 1.2104692939354444e-07,
+      "loss": 2.831,
+      "step": 2250
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 1.198845054399541e-07,
+      "loss": 3.0102,
+      "step": 2252
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 1.1872733387991108e-07,
+      "loss": 2.7581,
+      "step": 2254
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 1.1757542161923662e-07,
+      "loss": 2.8965,
+      "step": 2256
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 1.164287755323653e-07,
+      "loss": 3.1573,
+      "step": 2258
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 1.1528740246230462e-07,
+      "loss": 2.8608,
+      "step": 2260
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 1.1415130922059224e-07,
+      "loss": 2.9055,
+      "step": 2262
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 1.1302050258725814e-07,
+      "loss": 2.9471,
+      "step": 2264
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 1.1189498931078168e-07,
+      "loss": 2.8742,
+      "step": 2266
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 1.107747761080523e-07,
+      "loss": 2.8975,
+      "step": 2268
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 1.0965986966433038e-07,
+      "loss": 2.4924,
+      "step": 2270
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 1.085502766332046e-07,
+      "loss": 2.5379,
+      "step": 2272
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 1.0744600363655587e-07,
+      "loss": 3.0433,
+      "step": 2274
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 1.0634705726451476e-07,
+      "loss": 2.8508,
+      "step": 2276
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 1.052534440754237e-07,
+      "loss": 2.7688,
+      "step": 2278
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 1.0416517059579788e-07,
+      "loss": 2.8261,
+      "step": 2280
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 1.0308224332028548e-07,
+      "loss": 2.4478,
+      "step": 2282
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 1.0200466871162938e-07,
+      "loss": 3.2098,
+      "step": 2284
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 1.0093245320062871e-07,
+      "loss": 2.7841,
+      "step": 2286
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 9.986560318610026e-08,
+      "loss": 2.7897,
+      "step": 2288
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 9.880412503484026e-08,
+      "loss": 3.0868,
+      "step": 2290
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 9.774802508158698e-08,
+      "loss": 2.8863,
+      "step": 2292
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 9.669730962898182e-08,
+      "loss": 2.9924,
+      "step": 2294
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 9.56519849475328e-08,
+      "loss": 2.3201,
+      "step": 2296
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 9.461205727557609e-08,
+      "loss": 3.0813,
+      "step": 2298
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 9.357753281923975e-08,
+      "loss": 2.7553,
+      "step": 2300
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 9.25484177524064e-08,
+      "loss": 2.6823,
+      "step": 2302
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 9.152471821667595e-08,
+      "loss": 2.7055,
+      "step": 2304
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 9.050644032132915e-08,
+      "loss": 2.7976,
+      "step": 2306
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 8.949359014329205e-08,
+      "loss": 3.0599,
+      "step": 2308
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 8.848617372709755e-08,
+      "loss": 2.7271,
+      "step": 2310
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 8.748419708485199e-08,
+      "loss": 3.149,
+      "step": 2312
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 8.64876661961974e-08,
+      "loss": 2.569,
+      "step": 2314
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 8.5496587008276e-08,
+      "loss": 2.8628,
+      "step": 2316
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 8.451096543569636e-08,
+      "loss": 2.5182,
+      "step": 2318
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 8.353080736049522e-08,
+      "loss": 2.8243,
+      "step": 2320
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 8.255611863210521e-08,
+      "loss": 2.7879,
+      "step": 2322
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 8.158690506731835e-08,
+      "loss": 2.8732,
+      "step": 2324
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 8.062317245025174e-08,
+      "loss": 3.134,
+      "step": 2326
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 7.966492653231293e-08,
+      "loss": 2.7235,
+      "step": 2328
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 7.871217303216571e-08,
+      "loss": 2.9314,
+      "step": 2330
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 7.776491763569625e-08,
+      "loss": 2.4789,
+      "step": 2332
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 7.68231659959786e-08,
+      "loss": 3.1469,
+      "step": 2334
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 7.588692373324124e-08,
+      "loss": 2.7273,
+      "step": 2336
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 7.495619643483364e-08,
+      "loss": 2.7551,
+      "step": 2338
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 7.403098965519283e-08,
+      "loss": 2.7504,
+      "step": 2340
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 7.311130891581029e-08,
+      "loss": 3.006,
+      "step": 2342
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 7.219715970519869e-08,
+      "loss": 2.8902,
+      "step": 2344
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 7.128854747885959e-08,
+      "loss": 2.552,
+      "step": 2346
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 7.038547765925051e-08,
+      "loss": 2.7034,
+      "step": 2348
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 6.948795563575327e-08,
+      "loss": 2.9586,
+      "step": 2350
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 6.85959867646403e-08,
+      "loss": 2.9514,
+      "step": 2352
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 6.770957636904484e-08,
+      "loss": 2.9875,
+      "step": 2354
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 6.682872973892717e-08,
+      "loss": 2.4188,
+      "step": 2356
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 6.595345213104397e-08,
+      "loss": 2.9447,
+      "step": 2358
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 6.508374876891753e-08,
+      "loss": 2.9847,
+      "step": 2360
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 6.421962484280275e-08,
+      "loss": 2.4446,
+      "step": 2362
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 6.336108550965857e-08,
+      "loss": 2.738,
+      "step": 2364
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 6.250813589311499e-08,
+      "loss": 3.2748,
+      "step": 2366
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 6.166078108344364e-08,
+      "loss": 2.7968,
+      "step": 2368
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 6.081902613752787e-08,
+      "loss": 2.825,
+      "step": 2370
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 5.998287607883102e-08,
+      "loss": 2.7658,
+      "step": 2372
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 5.915233589736768e-08,
+      "loss": 2.9618,
+      "step": 2374
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 5.8327410549673715e-08,
+      "loss": 2.6213,
+      "step": 2376
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 5.750810495877623e-08,
+      "loss": 2.9399,
+      "step": 2378
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 5.669442401416491e-08,
+      "loss": 2.5547,
+      "step": 2380
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 5.588637257176188e-08,
+      "loss": 2.8335,
+      "step": 2382
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 5.508395545389377e-08,
+      "loss": 2.726,
+      "step": 2384
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 5.4287177449262035e-08,
+      "loss": 2.9261,
+      "step": 2386
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 5.349604331291491e-08,
+      "loss": 2.5964,
+      "step": 2388
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 5.271055776621891e-08,
+      "loss": 2.7978,
+      "step": 2390
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 5.1930725496830754e-08,
+      "loss": 2.9067,
+      "step": 2392
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 5.1156551158669234e-08,
+      "loss": 2.8638,
+      "step": 2394
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 5.038803937188729e-08,
+      "loss": 3.2469,
+      "step": 2396
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 4.9625194722844767e-08,
+      "loss": 2.9303,
+      "step": 2398
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 4.886802176408067e-08,
+      "loss": 3.0958,
+      "step": 2400
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 4.811652501428687e-08,
+      "loss": 2.8197,
+      "step": 2402
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 4.737070895827988e-08,
+      "loss": 3.0567,
+      "step": 2404
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 4.6630578046974787e-08,
+      "loss": 2.8699,
+      "step": 2406
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 4.589613669735914e-08,
+      "loss": 2.7792,
+      "step": 2408
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 4.516738929246511e-08,
+      "loss": 2.7646,
+      "step": 2410
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 4.444434018134524e-08,
+      "loss": 2.9106,
+      "step": 2412
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 4.3726993679044866e-08,
+      "loss": 2.5622,
+      "step": 2414
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 4.301535406657742e-08,
+      "loss": 2.6251,
+      "step": 2416
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 4.2309425590898454e-08,
+      "loss": 2.4432,
+      "step": 2418
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 4.160921246488003e-08,
+      "loss": 2.7231,
+      "step": 2420
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 4.0914718867286255e-08,
+      "loss": 3.0152,
+      "step": 2422
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 4.022594894274778e-08,
+      "loss": 3.1005,
+      "step": 2424
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 3.954290680173722e-08,
+      "loss": 2.8143,
+      "step": 2426
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 3.886559652054489e-08,
+      "loss": 2.9262,
+      "step": 2428
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 3.819402214125389e-08,
+      "loss": 2.4468,
+      "step": 2430
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 3.752818767171684e-08,
+      "loss": 2.7465,
+      "step": 2432
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 3.6868097085530956e-08,
+      "loss": 2.8041,
+      "step": 2434
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 3.621375432201512e-08,
+      "loss": 2.8799,
+      "step": 2436
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 3.556516328618597e-08,
+      "loss": 2.7,
+      "step": 2438
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 3.492232784873494e-08,
+      "loss": 2.8016,
+      "step": 2440
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 3.4285251846004725e-08,
+      "loss": 3.0438,
+      "step": 2442
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 3.365393907996661e-08,
+      "loss": 2.7742,
+      "step": 2444
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 3.3028393318197845e-08,
+      "loss": 2.8359,
+      "step": 2446
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 3.240861829385899e-08,
+      "loss": 2.8595,
+      "step": 2448
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 3.1794617705672156e-08,
+      "loss": 2.6778,
+      "step": 2450
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 3.118639521789768e-08,
+      "loss": 2.7007,
+      "step": 2452
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 3.0583954460313924e-08,
+      "loss": 3.1771,
+      "step": 2454
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 2.99872990281943e-08,
+      "loss": 2.4846,
+      "step": 2456
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 2.9396432482286405e-08,
+      "loss": 3.0357,
+      "step": 2458
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 2.881135834879078e-08,
+      "loss": 2.8547,
+      "step": 2460
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 2.8232080119339418e-08,
+      "loss": 2.8275,
+      "step": 2462
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 2.7658601250975632e-08,
+      "loss": 3.0354,
+      "step": 2464
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 2.7090925166132873e-08,
+      "loss": 2.5147,
+      "step": 2466
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 2.6529055252614285e-08,
+      "loss": 2.6707,
+      "step": 2468
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 2.5972994863573183e-08,
+      "loss": 2.5903,
+      "step": 2470
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 2.5422747317491944e-08,
+      "loss": 2.7884,
+      "step": 2472
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 2.487831589816314e-08,
+      "loss": 2.8384,
+      "step": 2474
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 2.4339703854669657e-08,
+      "loss": 2.6199,
+      "step": 2476
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 2.380691440136484e-08,
+      "loss": 2.6724,
+      "step": 2478
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 2.3279950717854157e-08,
+      "loss": 3.0304,
+      "step": 2480
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 2.275881594897566e-08,
+      "loss": 2.823,
+      "step": 2482
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 2.224351320478124e-08,
+      "loss": 2.9788,
+      "step": 2484
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 2.1734045560518166e-08,
+      "loss": 2.9393,
+      "step": 2486
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 2.12304160566108e-08,
+      "loss": 2.8637,
+      "step": 2488
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 2.073262769864226e-08,
+      "loss": 2.6603,
+      "step": 2490
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 2.024068345733676e-08,
+      "loss": 2.7437,
+      "step": 2492
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 1.9754586268541874e-08,
+      "loss": 2.748,
+      "step": 2494
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 1.9274339033210296e-08,
+      "loss": 2.8572,
+      "step": 2496
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 1.8799944617383877e-08,
+      "loss": 2.8819,
+      "step": 2498
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 1.833140585217485e-08,
+      "loss": 2.8412,
+      "step": 2500
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 1.786872553375074e-08,
+      "loss": 2.6918,
+      "step": 2502
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 1.7411906423316138e-08,
+      "loss": 2.7927,
+      "step": 2504
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 1.6960951247096954e-08,
+      "loss": 3.0818,
+      "step": 2506
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 1.6515862696324547e-08,
+      "loss": 2.8682,
+      "step": 2508
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 1.607664342721815e-08,
+      "loss": 2.7692,
+      "step": 2510
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 1.5643296060971144e-08,
+      "loss": 2.8995,
+      "step": 2512
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 1.52158231837336e-08,
+      "loss": 2.6709,
+      "step": 2514
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 1.479422734659752e-08,
+      "loss": 2.6749,
+      "step": 2516
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 1.4378511065582299e-08,
+      "loss": 2.7863,
+      "step": 2518
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 1.3968676821618065e-08,
+      "loss": 2.8782,
+      "step": 2520
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 1.35647270605328e-08,
+      "loss": 2.6534,
+      "step": 2522
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 1.3166664193036248e-08,
+      "loss": 2.7203,
+      "step": 2524
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 1.2774490594706032e-08,
+      "loss": 2.7822,
+      "step": 2526
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 1.2388208605974115e-08,
+      "loss": 2.5871,
+      "step": 2528
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 1.2007820532111467e-08,
+      "loss": 2.6158,
+      "step": 2530
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 1.1633328643215534e-08,
+      "loss": 3.0137,
+      "step": 2532
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 1.126473517419635e-08,
+      "loss": 2.7719,
+      "step": 2534
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 1.0902042324762662e-08,
+      "loss": 2.9144,
+      "step": 2536
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 1.054525225940961e-08,
+      "loss": 2.5536,
+      "step": 2538
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 1.0194367107405177e-08,
+      "loss": 3.0706,
+      "step": 2540
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 9.849388962777982e-09,
+      "loss": 2.903,
+      "step": 2542
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 9.510319884304175e-09,
+      "loss": 2.9641,
+      "step": 2544
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 9.177161895496e-09,
+      "loss": 3.0056,
+      "step": 2546
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 8.849916984588702e-09,
+      "loss": 2.5194,
+      "step": 2548
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 8.528587104529638e-09,
+      "loss": 2.7959,
+      "step": 2550
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 8.213174172965742e-09,
+      "loss": 2.4067,
+      "step": 2552
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 7.903680072233077e-09,
+      "loss": 2.6204,
+      "step": 2554
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 7.600106649344518e-09,
+      "loss": 2.9532,
+      "step": 2556
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 7.3024557159793166e-09,
+      "loss": 2.633,
+      "step": 2558
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 7.0107290484726635e-09,
+      "loss": 2.4816,
+      "step": 2560
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 6.7249283878039185e-09,
+      "loss": 2.8236,
+      "step": 2562
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 6.445055439587732e-09,
+      "loss": 3.11,
+      "step": 2564
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 6.171111874062496e-09,
+      "loss": 2.9049,
+      "step": 2566
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 5.90309932608124e-09,
+      "loss": 2.9968,
+      "step": 2568
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 5.641019395101421e-09,
+      "loss": 2.6047,
+      "step": 2570
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 5.384873645175813e-09,
+      "loss": 3.1338,
+      "step": 2572
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 5.134663604942635e-09,
+      "loss": 2.5713,
+      "step": 2574
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 4.89039076761677e-09,
+      "loss": 2.7235,
+      "step": 2576
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 4.652056590980891e-09,
+      "loss": 3.1282,
+      "step": 2578
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 4.419662497376797e-09,
+      "loss": 2.8595,
+      "step": 2580
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 4.193209873696313e-09,
+      "loss": 2.8237,
+      "step": 2582
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 3.972700071373958e-09,
+      "loss": 2.9459,
+      "step": 2584
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 3.7581344063780666e-09,
+      "loss": 2.6471,
+      "step": 2586
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 3.5495141592035702e-09,
+      "loss": 2.6666,
+      "step": 2588
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 3.346840574864007e-09,
+      "loss": 2.7841,
+      "step": 2590
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 3.150114862883857e-09,
+      "loss": 2.9788,
+      "step": 2592
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.959338197291994e-09,
+      "loss": 2.5035,
+      "step": 2594
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.7745117166140255e-09,
+      "loss": 2.7372,
+      "step": 2596
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.5956365238660736e-09,
+      "loss": 2.7331,
+      "step": 2598
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 2.4227136865474506e-09,
+      "loss": 2.745,
+      "step": 2600
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 2.2557442366351043e-09,
+      "loss": 2.6551,
+      "step": 2602
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 2.09472917057707e-09,
+      "loss": 2.6462,
+      "step": 2604
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 1.939669449286252e-09,
+      "loss": 2.6305,
+      "step": 2606
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 1.7905659981353183e-09,
+      "loss": 2.857,
+      "step": 2608
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 1.6474197069504815e-09,
+      "loss": 2.8444,
+      "step": 2610
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 1.5102314300068365e-09,
+      "loss": 2.5423,
+      "step": 2612
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 1.3790019860229206e-09,
+      "loss": 3.0096,
+      "step": 2614
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 1.2537321581557182e-09,
+      "loss": 2.4271,
+      "step": 2616
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 1.1344226939964397e-09,
+      "loss": 2.6693,
+      "step": 2618
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 1.0210743055651949e-09,
+      "loss": 2.638,
+      "step": 2620
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 9.136876693081052e-10,
+      "loss": 2.7145,
+      "step": 2622
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 8.122634260917527e-10,
+      "loss": 2.7503,
+      "step": 2624
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 7.168021812005154e-10,
+      "loss": 3.0314,
+      "step": 2626
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 6.273045043322378e-10,
+      "loss": 2.923,
+      "step": 2628
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 5.437709295955662e-10,
+      "loss": 2.6863,
+      "step": 2630
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 4.662019555056185e-10,
+      "loss": 2.7078,
+      "step": 2632
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 3.945980449823194e-10,
+      "loss": 3.1157,
+      "step": 2634
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 3.289596253467364e-10,
+      "loss": 2.7896,
+      "step": 2636
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 2.6928708831874814e-10,
+      "loss": 2.8468,
+      "step": 2638
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 2.1558079001493534e-10,
+      "loss": 2.5971,
+      "step": 2640
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 1.6784105094602706e-10,
+      "loss": 2.3665,
+      "step": 2642
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 1.2606815601567955e-10,
+      "loss": 2.8667,
+      "step": 2644
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 9.026235451803365e-11,
+      "loss": 2.8586,
+      "step": 2646
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 6.042386013638267e-11,
+      "loss": 2.7275,
+      "step": 2648
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 3.6552850942506156e-11,
+      "loss": 2.9348,
+      "step": 2650
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 1.8649469394782512e-11,
+      "loss": 2.5662,
+      "step": 2652
+    },
+    {
+      "epoch": 6.0,
+      "step": 2652,
+      "total_flos": 5.825747791275622e+16,
+      "train_loss": 3.2421822742099677,
+      "train_runtime": 19016.0105,
+      "train_samples_per_second": 8.929,
+      "train_steps_per_second": 0.139
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 2652,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 4000,
+  "total_flos": 5.825747791275622e+16,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}