diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,8 +1,8 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 5.0,
-  "global_step": 1115,
+  "epoch": 10.0,
+  "global_step": 2230,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -10,877 +10,877 @@
     {
       "epoch": 0.0,
       "learning_rate": 0.0,
-      "loss": 8.7788,
+      "loss": 8.7901,
       "step": 1
     },
     {
       "epoch": 0.01,
-      "learning_rate": 0.0,
-      "loss": 8.857,
+      "learning_rate": 6e-07,
+      "loss": 8.6749,
       "step": 2
     },
     {
       "epoch": 0.01,
-      "learning_rate": 6e-07,
-      "loss": 8.7704,
+      "learning_rate": 1.2e-06,
+      "loss": 8.6797,
       "step": 3
     },
     {
       "epoch": 0.02,
-      "learning_rate": 1.2e-06,
-      "loss": 8.6839,
+      "learning_rate": 1.8e-06,
+      "loss": 8.6105,
       "step": 4
     },
     {
       "epoch": 0.02,
-      "learning_rate": 1.8e-06,
-      "loss": 8.7517,
+      "learning_rate": 2.4e-06,
+      "loss": 8.4744,
       "step": 5
     },
     {
       "epoch": 0.03,
-      "learning_rate": 2.4e-06,
-      "loss": 8.3985,
+      "learning_rate": 2.9999999999999997e-06,
+      "loss": 8.3202,
       "step": 6
     },
     {
       "epoch": 0.03,
-      "learning_rate": 2.9999999999999997e-06,
-      "loss": 8.4278,
+      "learning_rate": 3.6e-06,
+      "loss": 8.2504,
       "step": 7
     },
     {
       "epoch": 0.04,
-      "learning_rate": 3.6e-06,
-      "loss": 8.0949,
+      "learning_rate": 4.2e-06,
+      "loss": 7.7477,
       "step": 8
     },
     {
       "epoch": 0.04,
-      "learning_rate": 4.2e-06,
-      "loss": 7.8106,
+      "learning_rate": 4.8e-06,
+      "loss": 7.6117,
       "step": 9
     },
     {
       "epoch": 0.04,
-      "learning_rate": 4.8e-06,
-      "loss": 7.4337,
+      "learning_rate": 5.399999999999999e-06,
+      "loss": 7.1625,
       "step": 10
     },
     {
       "epoch": 0.05,
-      "learning_rate": 5.399999999999999e-06,
-      "loss": 7.22,
+      "learning_rate": 5.999999999999999e-06,
+      "loss": 7.035,
       "step": 11
     },
     {
       "epoch": 0.05,
-      "learning_rate": 5.999999999999999e-06,
-      "loss": 6.9989,
+      "learning_rate": 6.599999999999999e-06,
+      "loss": 6.8002,
       "step": 12
     },
     {
       "epoch": 0.06,
-      "learning_rate": 6.599999999999999e-06,
-      "loss": 6.8282,
+      "learning_rate": 7.2e-06,
+      "loss": 6.5852,
       "step": 13
     },
     {
       "epoch": 0.06,
-      "learning_rate": 7.2e-06,
-      "loss": 6.6831,
+      "learning_rate": 7.799999999999998e-06,
+      "loss": 6.4332,
       "step": 14
     },
     {
       "epoch": 0.07,
-      "learning_rate": 7.799999999999998e-06,
-      "loss": 6.4666,
+      "learning_rate": 8.4e-06,
+      "loss": 6.2015,
       "step": 15
     },
     {
       "epoch": 0.07,
-      "learning_rate": 8.4e-06,
-      "loss": 6.2614,
+      "learning_rate": 8.999999999999999e-06,
+      "loss": 6.037,
       "step": 16
     },
     {
       "epoch": 0.08,
-      "learning_rate": 8.999999999999999e-06,
-      "loss": 6.1788,
+      "learning_rate": 9.6e-06,
+      "loss": 5.9392,
       "step": 17
     },
     {
       "epoch": 0.08,
-      "learning_rate": 9.6e-06,
-      "loss": 6.0435,
+      "learning_rate": 1.02e-05,
+      "loss": 5.7954,
       "step": 18
     },
     {
       "epoch": 0.09,
-      "learning_rate": 1.02e-05,
-      "loss": 5.8492,
+      "learning_rate": 1.0799999999999998e-05,
+      "loss": 5.6019,
       "step": 19
     },
     {
       "epoch": 0.09,
-      "learning_rate": 1.0799999999999998e-05,
-      "loss": 5.8448,
+      "learning_rate": 1.14e-05,
+      "loss": 5.6175,
       "step": 20
     },
     {
       "epoch": 0.09,
-      "learning_rate": 1.14e-05,
-      "loss": 5.6986,
+      "learning_rate": 1.1999999999999999e-05,
+      "loss": 5.481,
       "step": 21
     },
     {
       "epoch": 0.1,
-      "learning_rate": 1.1999999999999999e-05,
-      "loss": 5.673,
+      "learning_rate": 1.26e-05,
+      "loss": 5.4746,
       "step": 22
     },
     {
       "epoch": 0.1,
       "learning_rate": 1.26e-05,
-      "loss": 5.4552,
+      "loss": 5.3,
       "step": 23
     },
     {
       "epoch": 0.11,
       "learning_rate": 1.3199999999999997e-05,
-      "loss": 5.4442,
+      "loss": 5.3499,
       "step": 24
     },
     {
       "epoch": 0.11,
       "learning_rate": 1.3799999999999998e-05,
-      "loss": 5.2917,
+      "loss": 5.2045,
       "step": 25
     },
     {
       "epoch": 0.12,
       "learning_rate": 1.44e-05,
-      "loss": 5.29,
+      "loss": 5.2217,
       "step": 26
     },
     {
       "epoch": 0.12,
       "learning_rate": 1.4999999999999999e-05,
-      "loss": 5.1356,
+      "loss": 5.0872,
       "step": 27
     },
     {
       "epoch": 0.13,
       "learning_rate": 1.5599999999999996e-05,
-      "loss": 5.1709,
+      "loss": 5.1305,
       "step": 28
     },
     {
       "epoch": 0.13,
       "learning_rate": 1.6199999999999997e-05,
-      "loss": 4.9962,
+      "loss": 4.9678,
       "step": 29
     },
     {
       "epoch": 0.13,
       "learning_rate": 1.68e-05,
-      "loss": 5.1114,
+      "loss": 5.0943,
       "step": 30
     },
     {
       "epoch": 0.14,
       "learning_rate": 1.74e-05,
-      "loss": 5.1065,
+      "loss": 5.0962,
       "step": 31
     },
     {
       "epoch": 0.14,
       "learning_rate": 1.7999999999999997e-05,
-      "loss": 4.9779,
+      "loss": 4.9747,
       "step": 32
     },
     {
       "epoch": 0.15,
       "learning_rate": 1.8599999999999998e-05,
-      "loss": 4.9887,
+      "loss": 4.9798,
       "step": 33
     },
     {
       "epoch": 0.15,
       "learning_rate": 1.92e-05,
-      "loss": 4.9495,
+      "loss": 4.9415,
       "step": 34
     },
     {
       "epoch": 0.16,
       "learning_rate": 1.98e-05,
-      "loss": 4.8838,
+      "loss": 4.8916,
       "step": 35
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.04e-05,
-      "loss": 4.9061,
+      "loss": 4.9001,
       "step": 36
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.1e-05,
-      "loss": 4.9915,
+      "loss": 4.995,
       "step": 37
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.1599999999999996e-05,
-      "loss": 4.8265,
+      "loss": 4.8175,
       "step": 38
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.2199999999999998e-05,
-      "loss": 4.9207,
+      "loss": 4.9133,
       "step": 39
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.28e-05,
-      "loss": 4.9041,
+      "loss": 4.8894,
       "step": 40
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.34e-05,
-      "loss": 4.8838,
+      "loss": 4.8813,
       "step": 41
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.3999999999999997e-05,
-      "loss": 4.9159,
+      "loss": 4.9061,
       "step": 42
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.4599999999999998e-05,
-      "loss": 4.8392,
+      "loss": 4.8308,
       "step": 43
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.52e-05,
-      "loss": 4.9044,
+      "loss": 4.8939,
       "step": 44
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.5799999999999997e-05,
-      "loss": 4.8615,
+      "loss": 4.8437,
       "step": 45
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.6399999999999995e-05,
-      "loss": 4.9846,
+      "loss": 4.9756,
       "step": 46
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.6999999999999996e-05,
-      "loss": 4.8641,
+      "loss": 4.8558,
       "step": 47
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7599999999999997e-05,
-      "loss": 4.8026,
+      "loss": 4.7816,
       "step": 48
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.8199999999999998e-05,
-      "loss": 4.8971,
+      "loss": 4.8812,
       "step": 49
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.88e-05,
-      "loss": 4.6302,
+      "loss": 4.6317,
       "step": 50
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.94e-05,
-      "loss": 6.4289,
+      "loss": 6.1604,
       "step": 51
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.9999999999999997e-05,
-      "loss": 6.2769,
+      "loss": 6.0773,
       "step": 52
     },
     {
       "epoch": 0.24,
       "learning_rate": 3.06e-05,
-      "loss": 5.9356,
+      "loss": 5.8237,
       "step": 53
     },
     {
       "epoch": 0.24,
       "learning_rate": 3.119999999999999e-05,
-      "loss": 5.4623,
+      "loss": 5.4014,
       "step": 54
     },
     {
       "epoch": 0.25,
       "learning_rate": 3.1799999999999994e-05,
-      "loss": 5.1183,
+      "loss": 5.0832,
       "step": 55
     },
     {
       "epoch": 0.25,
       "learning_rate": 3.2399999999999995e-05,
-      "loss": 4.866,
+      "loss": 4.8432,
       "step": 56
     },
     {
       "epoch": 0.26,
       "learning_rate": 3.2999999999999996e-05,
-      "loss": 4.8544,
+      "loss": 4.8566,
       "step": 57
     },
     {
       "epoch": 0.26,
       "learning_rate": 3.36e-05,
-      "loss": 4.8141,
+      "loss": 4.8108,
       "step": 58
     },
     {
       "epoch": 0.26,
       "learning_rate": 3.42e-05,
-      "loss": 4.8047,
+      "loss": 4.7935,
       "step": 59
     },
     {
       "epoch": 0.27,
       "learning_rate": 3.48e-05,
-      "loss": 4.8444,
+      "loss": 4.8346,
       "step": 60
     },
     {
       "epoch": 0.27,
       "learning_rate": 3.539999999999999e-05,
-      "loss": 4.8777,
+      "loss": 4.8708,
       "step": 61
     },
     {
       "epoch": 0.28,
       "learning_rate": 3.5999999999999994e-05,
-      "loss": 4.7374,
+      "loss": 4.7303,
       "step": 62
     },
     {
       "epoch": 0.28,
       "learning_rate": 3.6599999999999995e-05,
-      "loss": 4.8447,
+      "loss": 4.8369,
       "step": 63
     },
     {
       "epoch": 0.29,
       "learning_rate": 3.7199999999999996e-05,
-      "loss": 4.7271,
+      "loss": 4.7123,
       "step": 64
     },
     {
       "epoch": 0.29,
       "learning_rate": 3.78e-05,
-      "loss": 4.691,
+      "loss": 4.6835,
       "step": 65
     },
     {
       "epoch": 0.3,
       "learning_rate": 3.84e-05,
-      "loss": 4.7024,
+      "loss": 4.7015,
       "step": 66
     },
     {
       "epoch": 0.3,
       "learning_rate": 3.9e-05,
-      "loss": 4.6623,
+      "loss": 4.6597,
       "step": 67
     },
     {
       "epoch": 0.3,
       "learning_rate": 3.96e-05,
-      "loss": 4.6917,
+      "loss": 4.6913,
       "step": 68
     },
     {
       "epoch": 0.31,
       "learning_rate": 4.02e-05,
-      "loss": 4.7466,
+      "loss": 4.7459,
       "step": 69
     },
     {
       "epoch": 0.31,
       "learning_rate": 4.08e-05,
-      "loss": 4.6977,
+      "loss": 4.6883,
       "step": 70
     },
     {
       "epoch": 0.32,
       "learning_rate": 4.14e-05,
-      "loss": 4.6366,
+      "loss": 4.6348,
       "step": 71
     },
     {
       "epoch": 0.32,
       "learning_rate": 4.2e-05,
-      "loss": 4.6913,
+      "loss": 4.6861,
       "step": 72
     },
     {
       "epoch": 0.33,
       "learning_rate": 4.259999999999999e-05,
-      "loss": 4.5726,
+      "loss": 4.5766,
       "step": 73
     },
     {
       "epoch": 0.33,
       "learning_rate": 4.319999999999999e-05,
-      "loss": 4.6529,
+      "loss": 4.6444,
       "step": 74
     },
     {
       "epoch": 0.34,
       "learning_rate": 4.3799999999999994e-05,
-      "loss": 4.7602,
+      "loss": 4.7433,
       "step": 75
     },
     {
       "epoch": 0.34,
       "learning_rate": 4.4399999999999995e-05,
-      "loss": 4.6657,
+      "loss": 4.6555,
       "step": 76
     },
     {
       "epoch": 0.35,
       "learning_rate": 4.4999999999999996e-05,
-      "loss": 4.5708,
+      "loss": 4.5683,
       "step": 77
     },
     {
       "epoch": 0.35,
       "learning_rate": 4.56e-05,
-      "loss": 4.6321,
+      "loss": 4.6298,
       "step": 78
     },
     {
       "epoch": 0.35,
       "learning_rate": 4.62e-05,
-      "loss": 4.5891,
+      "loss": 4.5865,
       "step": 79
     },
     {
       "epoch": 0.36,
       "learning_rate": 4.68e-05,
-      "loss": 4.695,
+      "loss": 4.6905,
       "step": 80
     },
     {
       "epoch": 0.36,
       "learning_rate": 4.7399999999999993e-05,
-      "loss": 4.5544,
+      "loss": 4.5504,
       "step": 81
     },
     {
       "epoch": 0.37,
       "learning_rate": 4.7999999999999994e-05,
-      "loss": 4.6184,
+      "loss": 4.6262,
       "step": 82
     },
     {
       "epoch": 0.37,
       "learning_rate": 4.8599999999999995e-05,
-      "loss": 4.6035,
+      "loss": 4.5915,
       "step": 83
     },
     {
       "epoch": 0.38,
       "learning_rate": 4.9199999999999997e-05,
-      "loss": 4.6037,
+      "loss": 4.6035,
       "step": 84
     },
     {
       "epoch": 0.38,
       "learning_rate": 4.98e-05,
-      "loss": 4.6785,
+      "loss": 4.6711,
       "step": 85
     },
     {
       "epoch": 0.39,
       "learning_rate": 5.04e-05,
-      "loss": 4.4987,
+      "loss": 4.493,
       "step": 86
     },
     {
       "epoch": 0.39,
       "learning_rate": 5.1e-05,
-      "loss": 4.6057,
+      "loss": 4.6,
       "step": 87
     },
     {
       "epoch": 0.39,
       "learning_rate": 5.1599999999999994e-05,
-      "loss": 4.5589,
+      "loss": 4.5418,
       "step": 88
     },
     {
       "epoch": 0.4,
       "learning_rate": 5.2199999999999995e-05,
-      "loss": 4.6228,
+      "loss": 4.6024,
       "step": 89
     },
     {
       "epoch": 0.4,
       "learning_rate": 5.279999999999999e-05,
-      "loss": 4.6534,
+      "loss": 4.6465,
       "step": 90
     },
     {
       "epoch": 0.41,
       "learning_rate": 5.339999999999999e-05,
-      "loss": 4.6014,
+      "loss": 4.5964,
       "step": 91
     },
     {
       "epoch": 0.41,
       "learning_rate": 5.399999999999999e-05,
-      "loss": 4.5941,
+      "loss": 4.5889,
       "step": 92
     },
     {
       "epoch": 0.42,
       "learning_rate": 5.459999999999999e-05,
-      "loss": 4.5625,
+      "loss": 4.5629,
       "step": 93
     },
     {
       "epoch": 0.42,
       "learning_rate": 5.519999999999999e-05,
-      "loss": 4.6271,
+      "loss": 4.6184,
       "step": 94
     },
     {
       "epoch": 0.43,
       "learning_rate": 5.5799999999999994e-05,
-      "loss": 4.6085,
+      "loss": 4.5852,
       "step": 95
     },
     {
       "epoch": 0.43,
       "learning_rate": 5.6399999999999995e-05,
-      "loss": 4.5724,
+      "loss": 4.5761,
       "step": 96
     },
     {
       "epoch": 0.43,
       "learning_rate": 5.6999999999999996e-05,
-      "loss": 4.6734,
+      "loss": 4.6631,
       "step": 97
     },
     {
       "epoch": 0.44,
       "learning_rate": 5.76e-05,
-      "loss": 4.7297,
+      "loss": 4.7284,
       "step": 98
     },
     {
       "epoch": 0.44,
       "learning_rate": 5.82e-05,
-      "loss": 4.601,
+      "loss": 4.5872,
       "step": 99
     },
     {
       "epoch": 0.45,
       "learning_rate": 5.88e-05,
-      "loss": 4.5277,
+      "loss": 4.5276,
       "step": 100
     },
     {
       "epoch": 0.45,
       "learning_rate": 5.94e-05,
-      "loss": 5.9227,
+      "loss": 5.9137,
       "step": 101
     },
     {
       "epoch": 0.46,
       "learning_rate": 5.9999999999999995e-05,
-      "loss": 5.8449,
+      "loss": 5.8072,
       "step": 102
     },
     {
       "epoch": 0.46,
       "learning_rate": 6.0599999999999996e-05,
-      "loss": 5.5435,
+      "loss": 5.5062,
       "step": 103
     },
     {
       "epoch": 0.47,
       "learning_rate": 6.12e-05,
-      "loss": 5.1179,
+      "loss": 5.1048,
       "step": 104
     },
     {
       "epoch": 0.47,
       "learning_rate": 6.18e-05,
-      "loss": 4.9785,
+      "loss": 4.9539,
       "step": 105
     },
     {
       "epoch": 0.48,
       "learning_rate": 6.239999999999999e-05,
-      "loss": 4.8924,
+      "loss": 4.8815,
       "step": 106
     },
     {
       "epoch": 0.48,
       "learning_rate": 6.299999999999999e-05,
-      "loss": 4.7039,
+      "loss": 4.6952,
       "step": 107
     },
     {
       "epoch": 0.48,
       "learning_rate": 6.359999999999999e-05,
-      "loss": 4.7125,
+      "loss": 4.7131,
       "step": 108
     },
     {
       "epoch": 0.49,
       "learning_rate": 6.419999999999999e-05,
-      "loss": 4.6977,
+      "loss": 4.6939,
       "step": 109
     },
     {
       "epoch": 0.49,
       "learning_rate": 6.479999999999999e-05,
-      "loss": 4.6824,
+      "loss": 4.6896,
       "step": 110
     },
     {
       "epoch": 0.5,
       "learning_rate": 6.539999999999999e-05,
-      "loss": 4.6191,
+      "loss": 4.6205,
       "step": 111
     },
     {
       "epoch": 0.5,
       "learning_rate": 6.599999999999999e-05,
-      "loss": 4.6363,
+      "loss": 4.6257,
       "step": 112
     },
     {
       "epoch": 0.51,
       "learning_rate": 6.659999999999999e-05,
-      "loss": 4.6369,
+      "loss": 4.63,
       "step": 113
     },
     {
       "epoch": 0.51,
       "learning_rate": 6.72e-05,
-      "loss": 4.5246,
+      "loss": 4.5141,
       "step": 114
     },
     {
       "epoch": 0.52,
       "learning_rate": 6.78e-05,
-      "loss": 4.5943,
+      "loss": 4.5986,
       "step": 115
     },
     {
       "epoch": 0.52,
       "learning_rate": 6.84e-05,
-      "loss": 4.5545,
+      "loss": 4.55,
       "step": 116
     },
     {
       "epoch": 0.52,
       "learning_rate": 6.9e-05,
-      "loss": 4.5898,
+      "loss": 4.5866,
       "step": 117
     },
     {
       "epoch": 0.53,
       "learning_rate": 6.96e-05,
-      "loss": 4.5237,
+      "loss": 4.5122,
       "step": 118
     },
     {
       "epoch": 0.53,
       "learning_rate": 7.02e-05,
-      "loss": 4.5241,
+      "loss": 4.5169,
       "step": 119
     },
     {
       "epoch": 0.54,
       "learning_rate": 7.079999999999999e-05,
-      "loss": 4.5408,
+      "loss": 4.5464,
       "step": 120
     },
     {
       "epoch": 0.54,
       "learning_rate": 7.139999999999999e-05,
-      "loss": 4.4809,
+      "loss": 4.4779,
       "step": 121
     },
     {
       "epoch": 0.55,
       "learning_rate": 7.199999999999999e-05,
-      "loss": 4.5278,
+      "loss": 4.517,
       "step": 122
     },
     {
       "epoch": 0.55,
       "learning_rate": 7.259999999999999e-05,
-      "loss": 4.4764,
+      "loss": 4.4728,
       "step": 123
     },
     {
       "epoch": 0.56,
       "learning_rate": 7.319999999999999e-05,
-      "loss": 4.4472,
+      "loss": 4.4484,
       "step": 124
     },
     {
       "epoch": 0.56,
       "learning_rate": 7.379999999999999e-05,
-      "loss": 4.5971,
+      "loss": 4.6016,
       "step": 125
     },
     {
       "epoch": 0.57,
       "learning_rate": 7.439999999999999e-05,
-      "loss": 4.4813,
+      "loss": 4.4771,
       "step": 126
     },
     {
       "epoch": 0.57,
       "learning_rate": 7.5e-05,
-      "loss": 4.5371,
+      "loss": 4.5307,
       "step": 127
     },
     {
       "epoch": 0.57,
       "learning_rate": 7.56e-05,
-      "loss": 4.5538,
+      "loss": 4.5465,
       "step": 128
     },
     {
       "epoch": 0.58,
       "learning_rate": 7.62e-05,
-      "loss": 4.5222,
+      "loss": 4.5136,
       "step": 129
     },
     {
       "epoch": 0.58,
       "learning_rate": 7.68e-05,
-      "loss": 4.5567,
+      "loss": 4.5539,
       "step": 130
     },
     {
       "epoch": 0.59,
       "learning_rate": 7.74e-05,
-      "loss": 4.5519,
+      "loss": 4.549,
       "step": 131
     },
     {
       "epoch": 0.59,
       "learning_rate": 7.8e-05,
-      "loss": 4.3829,
+      "loss": 4.3814,
       "step": 132
     },
     {
       "epoch": 0.6,
       "learning_rate": 7.86e-05,
-      "loss": 4.3992,
+      "loss": 4.3906,
       "step": 133
     },
     {
       "epoch": 0.6,
       "learning_rate": 7.92e-05,
-      "loss": 4.4999,
+      "loss": 4.492,
       "step": 134
     },
     {
       "epoch": 0.61,
       "learning_rate": 7.98e-05,
-      "loss": 4.5518,
+      "loss": 4.5401,
       "step": 135
     },
     {
       "epoch": 0.61,
       "learning_rate": 8.04e-05,
-      "loss": 4.48,
+      "loss": 4.4804,
       "step": 136
     },
     {
       "epoch": 0.61,
       "learning_rate": 8.1e-05,
-      "loss": 4.5714,
+      "loss": 4.5568,
       "step": 137
     },
     {
       "epoch": 0.62,
       "learning_rate": 8.16e-05,
-      "loss": 4.4858,
+      "loss": 4.4821,
       "step": 138
     },
     {
       "epoch": 0.62,
       "learning_rate": 8.22e-05,
-      "loss": 4.4397,
+      "loss": 4.4264,
       "step": 139
     },
     {
       "epoch": 0.63,
       "learning_rate": 8.28e-05,
-      "loss": 4.5127,
+      "loss": 4.5104,
       "step": 140
     },
     {
       "epoch": 0.63,
       "learning_rate": 8.34e-05,
-      "loss": 4.5398,
+      "loss": 4.5322,
       "step": 141
     },
     {
       "epoch": 0.64,
       "learning_rate": 8.4e-05,
-      "loss": 4.5412,
+      "loss": 4.5441,
       "step": 142
     },
     {
       "epoch": 0.64,
       "learning_rate": 8.459999999999998e-05,
-      "loss": 4.6863,
+      "loss": 4.6673,
       "step": 143
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.519999999999998e-05,
-      "loss": 4.5778,
+      "loss": 4.5759,
       "step": 144
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.579999999999998e-05,
-      "loss": 4.4345,
+      "loss": 4.4321,
       "step": 145
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.639999999999999e-05,
-      "loss": 4.4066,
+      "loss": 4.4123,
       "step": 146
     },
     {
@@ -892,5841 +892,12549 @@
     {
       "epoch": 0.66,
       "learning_rate": 8.759999999999999e-05,
-      "loss": 4.6398,
+      "loss": 4.6419,
       "step": 148
     },
     {
       "epoch": 0.67,
       "learning_rate": 8.819999999999999e-05,
-      "loss": 4.4105,
+      "loss": 4.44,
       "step": 149
     },
     {
       "epoch": 0.67,
       "learning_rate": 8.879999999999999e-05,
-      "loss": 4.4866,
+      "loss": 4.4715,
       "step": 150
     },
     {
       "epoch": 0.68,
       "learning_rate": 8.939999999999999e-05,
-      "loss": 6.0924,
+      "loss": 6.287,
       "step": 151
     },
     {
       "epoch": 0.68,
       "learning_rate": 8.999999999999999e-05,
-      "loss": 5.8545,
+      "loss": 6.1091,
       "step": 152
     },
     {
       "epoch": 0.69,
       "learning_rate": 9.059999999999999e-05,
-      "loss": 5.3625,
+      "loss": 5.5967,
       "step": 153
     },
     {
       "epoch": 0.69,
       "learning_rate": 9.12e-05,
-      "loss": 4.9468,
+      "loss": 5.0873,
       "step": 154
     },
     {
       "epoch": 0.7,
       "learning_rate": 9.18e-05,
-      "loss": 4.8548,
+      "loss": 4.8993,
       "step": 155
     },
     {
       "epoch": 0.7,
       "learning_rate": 9.24e-05,
-      "loss": 4.7858,
+      "loss": 4.8036,
       "step": 156
     },
     {
       "epoch": 0.7,
       "learning_rate": 9.3e-05,
-      "loss": 4.6516,
+      "loss": 4.7089,
       "step": 157
     },
     {
       "epoch": 0.71,
       "learning_rate": 9.36e-05,
-      "loss": 4.6538,
+      "loss": 4.658,
       "step": 158
     },
     {
       "epoch": 0.71,
       "learning_rate": 9.419999999999999e-05,
-      "loss": 4.5987,
+      "loss": 4.5872,
       "step": 159
     },
     {
       "epoch": 0.72,
       "learning_rate": 9.479999999999999e-05,
-      "loss": 4.6631,
+      "loss": 4.6761,
       "step": 160
     },
     {
       "epoch": 0.72,
       "learning_rate": 9.539999999999999e-05,
-      "loss": 4.4868,
+      "loss": 4.5023,
       "step": 161
     },
     {
       "epoch": 0.73,
       "learning_rate": 9.599999999999999e-05,
-      "loss": 4.5287,
+      "loss": 4.5411,
       "step": 162
     },
     {
       "epoch": 0.73,
       "learning_rate": 9.659999999999999e-05,
-      "loss": 4.5528,
+      "loss": 4.5728,
       "step": 163
     },
     {
       "epoch": 0.74,
       "learning_rate": 9.719999999999999e-05,
-      "loss": 4.5033,
+      "loss": 4.5043,
       "step": 164
     },
     {
       "epoch": 0.74,
       "learning_rate": 9.779999999999999e-05,
-      "loss": 4.5908,
+      "loss": 4.5902,
       "step": 165
     },
     {
       "epoch": 0.74,
       "learning_rate": 9.839999999999999e-05,
-      "loss": 4.4931,
+      "loss": 4.4932,
       "step": 166
     },
     {
       "epoch": 0.75,
       "learning_rate": 9.9e-05,
-      "loss": 4.519,
+      "loss": 4.5148,
       "step": 167
     },
     {
       "epoch": 0.75,
       "learning_rate": 9.96e-05,
-      "loss": 4.5015,
+      "loss": 4.5009,
       "step": 168
     },
     {
       "epoch": 0.76,
       "learning_rate": 0.0001002,
-      "loss": 4.3689,
+      "loss": 4.3554,
       "step": 169
     },
     {
       "epoch": 0.76,
       "learning_rate": 0.0001008,
-      "loss": 4.5586,
+      "loss": 4.5545,
       "step": 170
     },
     {
       "epoch": 0.77,
       "learning_rate": 0.0001014,
-      "loss": 4.4859,
+      "loss": 4.4837,
       "step": 171
     },
     {
       "epoch": 0.77,
       "learning_rate": 0.000102,
-      "loss": 4.3494,
+      "loss": 4.3605,
       "step": 172
     },
     {
       "epoch": 0.78,
       "learning_rate": 0.0001026,
-      "loss": 4.4217,
+      "loss": 4.423,
       "step": 173
     },
     {
       "epoch": 0.78,
       "learning_rate": 0.00010319999999999999,
-      "loss": 4.3855,
+      "loss": 4.3804,
       "step": 174
     },
     {
       "epoch": 0.78,
       "learning_rate": 0.00010379999999999999,
-      "loss": 4.4697,
+      "loss": 4.4725,
       "step": 175
     },
     {
       "epoch": 0.79,
       "learning_rate": 0.00010439999999999999,
-      "loss": 4.4196,
+      "loss": 4.4212,
       "step": 176
     },
     {
       "epoch": 0.79,
       "learning_rate": 0.00010499999999999999,
-      "loss": 4.438,
+      "loss": 4.4313,
       "step": 177
     },
     {
       "epoch": 0.8,
       "learning_rate": 0.00010559999999999998,
-      "loss": 4.3658,
+      "loss": 4.3636,
       "step": 178
     },
     {
       "epoch": 0.8,
       "learning_rate": 0.00010619999999999998,
-      "loss": 4.4174,
+      "loss": 4.4336,
       "step": 179
     },
     {
       "epoch": 0.81,
       "learning_rate": 0.00010679999999999998,
-      "loss": 4.4003,
+      "loss": 4.3958,
       "step": 180
     },
     {
       "epoch": 0.81,
       "learning_rate": 0.00010739999999999998,
-      "loss": 4.5487,
+      "loss": 4.529,
       "step": 181
     },
     {
       "epoch": 0.82,
       "learning_rate": 0.00010799999999999998,
-      "loss": 4.35,
+      "loss": 4.3276,
       "step": 182
     },
     {
       "epoch": 0.82,
       "learning_rate": 0.00010859999999999998,
-      "loss": 4.5236,
+      "loss": 4.5321,
       "step": 183
     },
     {
       "epoch": 0.83,
       "learning_rate": 0.00010919999999999998,
-      "loss": 4.4523,
+      "loss": 4.4487,
       "step": 184
     },
     {
       "epoch": 0.83,
       "learning_rate": 0.00010979999999999999,
-      "loss": 4.4892,
+      "loss": 4.4949,
       "step": 185
     },
     {
       "epoch": 0.83,
       "learning_rate": 0.00011039999999999999,
-      "loss": 4.3919,
+      "loss": 4.4117,
       "step": 186
     },
     {
       "epoch": 0.84,
       "learning_rate": 0.00011099999999999999,
-      "loss": 4.4118,
+      "loss": 4.407,
       "step": 187
     },
     {
       "epoch": 0.84,
       "learning_rate": 0.00011159999999999999,
-      "loss": 4.5025,
+      "loss": 4.4996,
       "step": 188
     },
     {
       "epoch": 0.85,
       "learning_rate": 0.00011219999999999999,
-      "loss": 4.4032,
+      "loss": 4.3983,
       "step": 189
     },
     {
       "epoch": 0.85,
       "learning_rate": 0.00011279999999999999,
-      "loss": 4.4338,
+      "loss": 4.4501,
       "step": 190
     },
     {
       "epoch": 0.86,
       "learning_rate": 0.00011339999999999999,
-      "loss": 4.4694,
+      "loss": 4.4659,
       "step": 191
     },
     {
       "epoch": 0.86,
       "learning_rate": 0.00011399999999999999,
-      "loss": 4.3931,
+      "loss": 4.3934,
       "step": 192
     },
     {
       "epoch": 0.87,
       "learning_rate": 0.0001146,
-      "loss": 4.3904,
+      "loss": 4.3729,
       "step": 193
     },
     {
       "epoch": 0.87,
       "learning_rate": 0.0001152,
-      "loss": 4.4158,
+      "loss": 4.4194,
       "step": 194
     },
     {
       "epoch": 0.87,
       "learning_rate": 0.0001158,
-      "loss": 4.4491,
+      "loss": 4.4216,
       "step": 195
     },
     {
       "epoch": 0.88,
       "learning_rate": 0.0001164,
-      "loss": 4.4745,
+      "loss": 4.4952,
       "step": 196
     },
     {
       "epoch": 0.88,
       "learning_rate": 0.000117,
-      "loss": 4.4689,
+      "loss": 4.4412,
       "step": 197
     },
     {
       "epoch": 0.89,
       "learning_rate": 0.0001176,
-      "loss": 4.4095,
+      "loss": 4.4064,
       "step": 198
     },
     {
       "epoch": 0.89,
       "learning_rate": 0.0001182,
-      "loss": 4.461,
+      "loss": 4.4579,
       "step": 199
     },
     {
       "epoch": 0.9,
       "learning_rate": 0.0001188,
-      "loss": 4.2483,
+      "loss": 4.2313,
       "step": 200
     },
     {
       "epoch": 0.9,
       "learning_rate": 0.0001194,
-      "loss": 5.7607,
+      "loss": 5.4646,
       "step": 201
     },
     {
       "epoch": 0.91,
       "learning_rate": 0.00011999999999999999,
-      "loss": 5.7126,
+      "loss": 5.4181,
       "step": 202
     },
     {
       "epoch": 0.91,
       "learning_rate": 0.00012059999999999999,
-      "loss": 5.3124,
+      "loss": 5.0932,
       "step": 203
     },
     {
       "epoch": 0.91,
       "learning_rate": 0.00012119999999999999,
-      "loss": 4.9813,
+      "loss": 4.8642,
       "step": 204
     },
     {
       "epoch": 0.92,
       "learning_rate": 0.00012179999999999999,
-      "loss": 4.7145,
+      "loss": 4.6482,
       "step": 205
     },
     {
       "epoch": 0.92,
       "learning_rate": 0.0001224,
-      "loss": 4.7888,
+      "loss": 4.7276,
       "step": 206
     },
     {
       "epoch": 0.93,
       "learning_rate": 0.00012299999999999998,
-      "loss": 4.6485,
+      "loss": 4.6034,
       "step": 207
     },
     {
       "epoch": 0.93,
       "learning_rate": 0.0001236,
-      "loss": 4.6296,
+      "loss": 4.58,
       "step": 208
     },
     {
       "epoch": 0.94,
       "learning_rate": 0.00012419999999999998,
-      "loss": 4.4968,
+      "loss": 4.4605,
       "step": 209
     },
     {
       "epoch": 0.94,
       "learning_rate": 0.00012479999999999997,
-      "loss": 4.514,
+      "loss": 4.4878,
       "step": 210
     },
     {
       "epoch": 0.95,
       "learning_rate": 0.00012539999999999999,
-      "loss": 4.5182,
+      "loss": 4.502,
       "step": 211
     },
     {
       "epoch": 0.95,
       "learning_rate": 0.00012599999999999997,
-      "loss": 4.534,
+      "loss": 4.5168,
       "step": 212
     },
     {
       "epoch": 0.96,
       "learning_rate": 0.0001266,
-      "loss": 4.5318,
+      "loss": 4.5251,
       "step": 213
     },
     {
       "epoch": 0.96,
       "learning_rate": 0.00012719999999999997,
-      "loss": 4.4489,
+      "loss": 4.4259,
       "step": 214
     },
     {
       "epoch": 0.96,
       "learning_rate": 0.0001278,
-      "loss": 4.3867,
+      "loss": 4.3709,
       "step": 215
     },
     {
       "epoch": 0.97,
       "learning_rate": 0.00012839999999999998,
-      "loss": 4.5466,
+      "loss": 4.5284,
       "step": 216
     },
     {
       "epoch": 0.97,
       "learning_rate": 0.000129,
-      "loss": 4.3823,
+      "loss": 4.3736,
       "step": 217
     },
     {
       "epoch": 0.98,
       "learning_rate": 0.00012959999999999998,
-      "loss": 4.3786,
+      "loss": 4.3609,
       "step": 218
     },
     {
       "epoch": 0.98,
       "learning_rate": 0.0001302,
-      "loss": 4.4267,
+      "loss": 4.41,
       "step": 219
     },
     {
       "epoch": 0.99,
       "learning_rate": 0.00013079999999999998,
-      "loss": 4.433,
+      "loss": 4.4306,
       "step": 220
     },
     {
       "epoch": 0.99,
       "learning_rate": 0.0001314,
-      "loss": 4.4993,
+      "loss": 4.4937,
       "step": 221
     },
     {
       "epoch": 1.0,
       "learning_rate": 0.00013199999999999998,
-      "loss": 4.3747,
+      "loss": 4.3587,
       "step": 222
     },
     {
       "epoch": 1.0,
       "learning_rate": 0.0001326,
-      "loss": 4.3646,
+      "loss": 4.3467,
       "step": 223
     },
     {
       "epoch": 1.0,
       "learning_rate": 0.00013319999999999999,
-      "loss": 5.5865,
+      "loss": 5.8076,
       "step": 224
     },
     {
       "epoch": 1.01,
       "learning_rate": 0.0001338,
-      "loss": 5.3673,
+      "loss": 5.7454,
       "step": 225
     },
     {
       "epoch": 1.01,
       "learning_rate": 0.0001344,
-      "loss": 5.1522,
+      "loss": 5.6168,
       "step": 226
     },
     {
       "epoch": 1.02,
       "learning_rate": 0.000135,
-      "loss": 4.7286,
+      "loss": 5.1562,
       "step": 227
     },
     {
       "epoch": 1.02,
       "learning_rate": 0.0001356,
-      "loss": 4.6139,
+      "loss": 4.9135,
       "step": 228
     },
     {
       "epoch": 1.03,
       "learning_rate": 0.0001362,
-      "loss": 4.5005,
+      "loss": 4.6708,
       "step": 229
     },
     {
       "epoch": 1.03,
       "learning_rate": 0.0001368,
-      "loss": 4.4463,
+      "loss": 4.5292,
       "step": 230
     },
     {
       "epoch": 1.04,
       "learning_rate": 0.0001374,
-      "loss": 4.4191,
+      "loss": 4.4944,
       "step": 231
     },
     {
       "epoch": 1.04,
       "learning_rate": 0.000138,
-      "loss": 4.3203,
+      "loss": 4.3521,
       "step": 232
     },
     {
       "epoch": 1.04,
       "learning_rate": 0.0001386,
-      "loss": 4.4397,
+      "loss": 4.4591,
       "step": 233
     },
     {
       "epoch": 1.05,
       "learning_rate": 0.0001392,
-      "loss": 4.3437,
+      "loss": 4.3587,
       "step": 234
     },
     {
       "epoch": 1.05,
       "learning_rate": 0.00013979999999999998,
-      "loss": 4.3979,
+      "loss": 4.3937,
       "step": 235
     },
     {
       "epoch": 1.06,
       "learning_rate": 0.0001404,
-      "loss": 4.4071,
+      "loss": 4.3795,
       "step": 236
     },
     {
       "epoch": 1.06,
       "learning_rate": 0.00014099999999999998,
-      "loss": 4.3001,
+      "loss": 4.2489,
       "step": 237
     },
     {
       "epoch": 1.07,
       "learning_rate": 0.00014159999999999997,
-      "loss": 4.36,
+      "loss": 4.3007,
       "step": 238
     },
     {
       "epoch": 1.07,
       "learning_rate": 0.0001422,
-      "loss": 4.3493,
+      "loss": 4.2813,
       "step": 239
     },
     {
       "epoch": 1.08,
       "learning_rate": 0.00014279999999999997,
-      "loss": 4.4577,
+      "loss": 4.3177,
       "step": 240
     },
     {
       "epoch": 1.08,
       "learning_rate": 0.0001434,
-      "loss": 4.2136,
+      "loss": 4.0955,
       "step": 241
     },
     {
       "epoch": 1.09,
       "learning_rate": 0.00014399999999999998,
-      "loss": 4.2935,
+      "loss": 4.1885,
       "step": 242
     },
     {
       "epoch": 1.09,
       "learning_rate": 0.0001446,
-      "loss": 4.2193,
+      "loss": 4.1224,
       "step": 243
     },
     {
       "epoch": 1.09,
       "learning_rate": 0.00014519999999999998,
-      "loss": 4.3509,
+      "loss": 4.2549,
       "step": 244
     },
     {
       "epoch": 1.1,
       "learning_rate": 0.0001458,
-      "loss": 4.2148,
+      "loss": 4.1249,
       "step": 245
     },
     {
       "epoch": 1.1,
       "learning_rate": 0.00014639999999999998,
-      "loss": 4.2056,
+      "loss": 4.1269,
       "step": 246
     },
     {
       "epoch": 1.11,
       "learning_rate": 0.000147,
-      "loss": 4.2163,
+      "loss": 4.1725,
       "step": 247
     },
     {
       "epoch": 1.11,
       "learning_rate": 0.00014759999999999998,
-      "loss": 4.2646,
+      "loss": 4.2151,
       "step": 248
     },
     {
       "epoch": 1.12,
       "learning_rate": 0.0001482,
-      "loss": 4.1475,
+      "loss": 4.1098,
       "step": 249
     },
     {
       "epoch": 1.12,
       "learning_rate": 0.00014879999999999998,
-      "loss": 4.1661,
+      "loss": 4.1609,
       "step": 250
     },
     {
       "epoch": 1.13,
       "learning_rate": 0.0001494,
-      "loss": 4.0668,
+      "loss": 4.0873,
       "step": 251
     },
     {
       "epoch": 1.13,
       "learning_rate": 0.00015,
-      "loss": 4.0718,
+      "loss": 4.0377,
       "step": 252
     },
     {
       "epoch": 1.13,
       "learning_rate": 0.00015059999999999997,
-      "loss": 4.1521,
+      "loss": 4.1351,
       "step": 253
     },
     {
       "epoch": 1.14,
       "learning_rate": 0.0001512,
-      "loss": 4.0864,
+      "loss": 4.0665,
       "step": 254
     },
     {
       "epoch": 1.14,
       "learning_rate": 0.00015179999999999998,
-      "loss": 4.0792,
+      "loss": 4.0937,
       "step": 255
     },
     {
       "epoch": 1.15,
       "learning_rate": 0.0001524,
-      "loss": 4.0559,
+      "loss": 4.079,
       "step": 256
     },
     {
       "epoch": 1.15,
       "learning_rate": 0.00015299999999999998,
-      "loss": 4.0964,
+      "loss": 4.0957,
       "step": 257
     },
     {
       "epoch": 1.16,
       "learning_rate": 0.0001536,
-      "loss": 4.1454,
+      "loss": 4.1556,
       "step": 258
     },
     {
       "epoch": 1.16,
       "learning_rate": 0.00015419999999999998,
-      "loss": 4.0752,
+      "loss": 4.0854,
       "step": 259
     },
     {
       "epoch": 1.17,
       "learning_rate": 0.0001548,
-      "loss": 4.0471,
+      "loss": 4.0407,
       "step": 260
     },
     {
       "epoch": 1.17,
       "learning_rate": 0.00015539999999999998,
-      "loss": 4.1324,
+      "loss": 4.1202,
       "step": 261
     },
     {
       "epoch": 1.17,
       "learning_rate": 0.000156,
-      "loss": 4.0548,
+      "loss": 4.0466,
       "step": 262
     },
     {
       "epoch": 1.18,
       "learning_rate": 0.00015659999999999998,
-      "loss": 3.9025,
+      "loss": 3.8976,
       "step": 263
     },
     {
       "epoch": 1.18,
       "learning_rate": 0.0001572,
-      "loss": 4.1022,
+      "loss": 4.0776,
       "step": 264
     },
     {
       "epoch": 1.19,
       "learning_rate": 0.0001578,
-      "loss": 3.984,
+      "loss": 3.9628,
       "step": 265
     },
     {
       "epoch": 1.19,
       "learning_rate": 0.0001584,
-      "loss": 4.0924,
+      "loss": 4.0677,
       "step": 266
     },
     {
       "epoch": 1.2,
       "learning_rate": 0.000159,
-      "loss": 3.967,
+      "loss": 3.9569,
       "step": 267
     },
     {
       "epoch": 1.2,
       "learning_rate": 0.0001596,
-      "loss": 4.06,
+      "loss": 4.0117,
       "step": 268
     },
     {
       "epoch": 1.21,
       "learning_rate": 0.0001602,
-      "loss": 3.8391,
+      "loss": 3.835,
       "step": 269
     },
     {
       "epoch": 1.21,
       "learning_rate": 0.0001608,
-      "loss": 3.8725,
+      "loss": 3.8285,
       "step": 270
     },
     {
       "epoch": 1.22,
       "learning_rate": 0.0001614,
-      "loss": 3.947,
+      "loss": 3.9137,
       "step": 271
     },
     {
       "epoch": 1.22,
       "learning_rate": 0.000162,
-      "loss": 3.8932,
+      "loss": 3.9028,
       "step": 272
     },
     {
       "epoch": 1.22,
       "learning_rate": 0.0001626,
-      "loss": 3.6977,
+      "loss": 3.6852,
       "step": 273
     },
     {
       "epoch": 1.23,
       "learning_rate": 0.0001632,
-      "loss": 5.8216,
+      "loss": 6.3809,
       "step": 274
     },
     {
       "epoch": 1.23,
       "learning_rate": 0.0001638,
-      "loss": 5.3449,
+      "loss": 5.6719,
       "step": 275
     },
     {
       "epoch": 1.24,
       "learning_rate": 0.0001644,
-      "loss": 4.9906,
+      "loss": 5.153,
       "step": 276
     },
     {
       "epoch": 1.24,
       "learning_rate": 0.000165,
-      "loss": 4.7723,
+      "loss": 4.856,
       "step": 277
     },
     {
       "epoch": 1.25,
       "learning_rate": 0.0001656,
-      "loss": 4.5841,
+      "loss": 4.6105,
       "step": 278
     },
     {
       "epoch": 1.25,
       "learning_rate": 0.0001662,
-      "loss": 4.5655,
+      "loss": 4.6108,
       "step": 279
     },
     {
       "epoch": 1.26,
       "learning_rate": 0.0001668,
-      "loss": 4.5075,
+      "loss": 4.5733,
       "step": 280
     },
     {
       "epoch": 1.26,
       "learning_rate": 0.0001674,
-      "loss": 4.3671,
+      "loss": 4.4,
       "step": 281
     },
     {
       "epoch": 1.26,
       "learning_rate": 0.000168,
-      "loss": 4.4138,
+      "loss": 4.4329,
       "step": 282
     },
     {
       "epoch": 1.27,
       "learning_rate": 0.0001686,
-      "loss": 4.3305,
+      "loss": 4.3477,
       "step": 283
     },
     {
       "epoch": 1.27,
       "learning_rate": 0.00016919999999999997,
-      "loss": 4.2737,
+      "loss": 4.2902,
       "step": 284
     },
     {
       "epoch": 1.28,
       "learning_rate": 0.00016979999999999998,
-      "loss": 4.3387,
+      "loss": 4.3448,
       "step": 285
     },
     {
       "epoch": 1.28,
       "learning_rate": 0.00017039999999999997,
-      "loss": 4.218,
+      "loss": 4.2302,
       "step": 286
     },
     {
       "epoch": 1.29,
       "learning_rate": 0.00017099999999999998,
-      "loss": 4.2599,
+      "loss": 4.2629,
       "step": 287
     },
     {
       "epoch": 1.29,
       "learning_rate": 0.00017159999999999997,
-      "loss": 4.2436,
+      "loss": 4.2435,
       "step": 288
     },
     {
       "epoch": 1.3,
       "learning_rate": 0.00017219999999999998,
-      "loss": 4.1241,
+      "loss": 4.1274,
       "step": 289
     },
     {
       "epoch": 1.3,
       "learning_rate": 0.00017279999999999997,
-      "loss": 4.2546,
+      "loss": 4.28,
       "step": 290
     },
     {
       "epoch": 1.3,
       "learning_rate": 0.00017339999999999996,
-      "loss": 4.1435,
+      "loss": 4.1498,
       "step": 291
     },
     {
       "epoch": 1.31,
       "learning_rate": 0.00017399999999999997,
-      "loss": 4.0901,
+      "loss": 4.1554,
       "step": 292
     },
     {
       "epoch": 1.31,
       "learning_rate": 0.00017459999999999996,
-      "loss": 4.1774,
+      "loss": 4.1921,
       "step": 293
     },
     {
       "epoch": 1.32,
       "learning_rate": 0.00017519999999999998,
-      "loss": 4.1511,
+      "loss": 4.1944,
       "step": 294
     },
     {
       "epoch": 1.32,
       "learning_rate": 0.00017579999999999996,
-      "loss": 4.1222,
+      "loss": 4.1347,
       "step": 295
     },
     {
       "epoch": 1.33,
       "learning_rate": 0.00017639999999999998,
-      "loss": 4.0986,
+      "loss": 4.1154,
       "step": 296
     },
     {
       "epoch": 1.33,
       "learning_rate": 0.00017699999999999997,
-      "loss": 4.0735,
+      "loss": 4.0972,
       "step": 297
     },
     {
       "epoch": 1.34,
       "learning_rate": 0.00017759999999999998,
-      "loss": 4.0722,
+      "loss": 4.122,
       "step": 298
     },
     {
       "epoch": 1.34,
       "learning_rate": 0.00017819999999999997,
-      "loss": 4.162,
+      "loss": 4.1743,
       "step": 299
     },
     {
       "epoch": 1.35,
       "learning_rate": 0.00017879999999999998,
-      "loss": 4.1414,
+      "loss": 4.1579,
       "step": 300
     },
     {
       "epoch": 1.35,
       "learning_rate": 0.00017939999999999997,
-      "loss": 4.1157,
+      "loss": 4.1053,
       "step": 301
     },
     {
       "epoch": 1.35,
       "learning_rate": 0.00017999999999999998,
-      "loss": 4.0798,
+      "loss": 4.0905,
       "step": 302
     },
     {
       "epoch": 1.36,
       "learning_rate": 0.00018059999999999997,
-      "loss": 4.1804,
+      "loss": 4.1271,
       "step": 303
     },
     {
       "epoch": 1.36,
       "learning_rate": 0.00018119999999999999,
-      "loss": 4.0958,
+      "loss": 4.0824,
       "step": 304
     },
     {
       "epoch": 1.37,
       "learning_rate": 0.00018179999999999997,
-      "loss": 4.0506,
+      "loss": 4.0832,
       "step": 305
     },
     {
       "epoch": 1.37,
       "learning_rate": 0.0001824,
-      "loss": 4.1091,
+      "loss": 4.0989,
       "step": 306
     },
     {
       "epoch": 1.38,
       "learning_rate": 0.00018299999999999998,
-      "loss": 4.1218,
+      "loss": 4.137,
       "step": 307
     },
     {
       "epoch": 1.38,
       "learning_rate": 0.0001836,
-      "loss": 4.1636,
+      "loss": 4.1543,
       "step": 308
     },
     {
       "epoch": 1.39,
       "learning_rate": 0.00018419999999999998,
-      "loss": 4.08,
+      "loss": 4.0891,
       "step": 309
     },
     {
       "epoch": 1.39,
       "learning_rate": 0.0001848,
-      "loss": 4.0701,
+      "loss": 4.0728,
       "step": 310
     },
     {
       "epoch": 1.39,
       "learning_rate": 0.00018539999999999998,
-      "loss": 3.9545,
+      "loss": 3.9376,
       "step": 311
     },
     {
       "epoch": 1.4,
       "learning_rate": 0.000186,
-      "loss": 3.8738,
+      "loss": 3.8833,
       "step": 312
     },
     {
       "epoch": 1.4,
       "learning_rate": 0.00018659999999999998,
-      "loss": 4.0669,
+      "loss": 4.0698,
       "step": 313
     },
     {
       "epoch": 1.41,
       "learning_rate": 0.0001872,
-      "loss": 4.0076,
+      "loss": 3.9994,
       "step": 314
     },
     {
       "epoch": 1.41,
       "learning_rate": 0.00018779999999999998,
-      "loss": 4.0221,
+      "loss": 4.0358,
       "step": 315
     },
     {
       "epoch": 1.42,
       "learning_rate": 0.00018839999999999997,
-      "loss": 3.9464,
+      "loss": 3.9731,
       "step": 316
     },
     {
       "epoch": 1.42,
       "learning_rate": 0.00018899999999999999,
-      "loss": 3.771,
+      "loss": 3.787,
       "step": 317
     },
     {
       "epoch": 1.43,
       "learning_rate": 0.00018959999999999997,
-      "loss": 3.8719,
+      "loss": 3.8781,
       "step": 318
     },
     {
       "epoch": 1.43,
       "learning_rate": 0.0001902,
-      "loss": 3.9347,
+      "loss": 3.9194,
       "step": 319
     },
     {
       "epoch": 1.43,
       "learning_rate": 0.00019079999999999998,
-      "loss": 3.8062,
+      "loss": 3.8176,
       "step": 320
     },
     {
       "epoch": 1.44,
       "learning_rate": 0.0001914,
-      "loss": 3.7106,
+      "loss": 3.7146,
       "step": 321
     },
     {
       "epoch": 1.44,
       "learning_rate": 0.00019199999999999998,
-      "loss": 3.8163,
+      "loss": 3.8007,
       "step": 322
     },
     {
       "epoch": 1.45,
       "learning_rate": 0.0001926,
-      "loss": 3.6449,
+      "loss": 3.8317,
       "step": 323
     },
     {
       "epoch": 1.45,
       "learning_rate": 0.00019319999999999998,
-      "loss": 5.7573,
+      "loss": 5.473,
       "step": 324
     },
     {
       "epoch": 1.46,
       "learning_rate": 0.0001938,
-      "loss": 5.6261,
+      "loss": 5.3192,
       "step": 325
     },
     {
       "epoch": 1.46,
       "learning_rate": 0.00019439999999999998,
-      "loss": 5.3271,
+      "loss": 5.0983,
       "step": 326
     },
     {
       "epoch": 1.47,
       "learning_rate": 0.000195,
-      "loss": 4.9718,
+      "loss": 4.7318,
       "step": 327
     },
     {
       "epoch": 1.47,
       "learning_rate": 0.00019559999999999998,
-      "loss": 4.8634,
+      "loss": 4.6187,
       "step": 328
     },
     {
       "epoch": 1.48,
-      "learning_rate": 0.00019559999999999998,
-      "loss": 5.2309,
+      "learning_rate": 0.0001962,
+      "loss": 4.4429,
       "step": 329
     },
     {
       "epoch": 1.48,
-      "learning_rate": 0.0001962,
-      "loss": 5.1864,
+      "learning_rate": 0.00019679999999999999,
+      "loss": 4.3152,
       "step": 330
     },
     {
       "epoch": 1.48,
-      "learning_rate": 0.00019679999999999999,
-      "loss": 7.7546,
+      "learning_rate": 0.0001974,
+      "loss": 4.4061,
       "step": 331
     },
     {
       "epoch": 1.49,
-      "learning_rate": 0.0001974,
-      "loss": 7.6626,
+      "learning_rate": 0.000198,
+      "loss": 4.3511,
       "step": 332
     },
     {
       "epoch": 1.49,
-      "learning_rate": 0.000198,
-      "loss": 7.7001,
+      "learning_rate": 0.0001986,
+      "loss": 4.3352,
       "step": 333
     },
     {
       "epoch": 1.5,
-      "learning_rate": 0.0001986,
-      "loss": 7.4637,
+      "learning_rate": 0.0001992,
+      "loss": 4.2468,
       "step": 334
     },
     {
       "epoch": 1.5,
-      "learning_rate": 0.0001992,
-      "loss": 7.272,
+      "learning_rate": 0.0001998,
+      "loss": 4.2837,
       "step": 335
     },
     {
       "epoch": 1.51,
-      "learning_rate": 0.0001998,
-      "loss": 7.0608,
+      "learning_rate": 0.0002004,
+      "loss": 4.1831,
       "step": 336
     },
     {
       "epoch": 1.51,
-      "learning_rate": 0.0002004,
-      "loss": 6.811,
+      "learning_rate": 0.000201,
+      "loss": 4.2178,
       "step": 337
     },
     {
       "epoch": 1.52,
-      "learning_rate": 0.000201,
-      "loss": 6.8756,
+      "learning_rate": 0.0002016,
+      "loss": 4.1241,
       "step": 338
     },
     {
       "epoch": 1.52,
-      "learning_rate": 0.0002016,
-      "loss": 6.0821,
+      "learning_rate": 0.0002022,
+      "loss": 4.1374,
       "step": 339
     },
     {
       "epoch": 1.52,
-      "learning_rate": 0.0002022,
-      "loss": 5.2635,
+      "learning_rate": 0.0002028,
+      "loss": 4.0581,
       "step": 340
     },
     {
       "epoch": 1.53,
-      "learning_rate": 0.0002028,
-      "loss": 5.1346,
+      "learning_rate": 0.00020339999999999998,
+      "loss": 4.1244,
       "step": 341
     },
     {
       "epoch": 1.53,
-      "learning_rate": 0.00020339999999999998,
-      "loss": 4.9469,
+      "learning_rate": 0.000204,
+      "loss": 4.101,
       "step": 342
     },
     {
       "epoch": 1.54,
-      "learning_rate": 0.000204,
-      "loss": 4.7854,
+      "learning_rate": 0.00020459999999999999,
+      "loss": 4.0908,
       "step": 343
     },
     {
       "epoch": 1.54,
-      "learning_rate": 0.00020459999999999999,
-      "loss": 4.4968,
+      "learning_rate": 0.0002052,
+      "loss": 3.9862,
       "step": 344
     },
     {
       "epoch": 1.55,
-      "learning_rate": 0.0002052,
-      "loss": 4.5434,
+      "learning_rate": 0.0002058,
+      "loss": 4.1008,
       "step": 345
     },
     {
       "epoch": 1.55,
-      "learning_rate": 0.0002058,
-      "loss": 4.4593,
+      "learning_rate": 0.00020639999999999998,
+      "loss": 4.0601,
       "step": 346
     },
     {
       "epoch": 1.56,
-      "learning_rate": 0.00020639999999999998,
-      "loss": 4.4982,
+      "learning_rate": 0.00020699999999999996,
+      "loss": 4.063,
       "step": 347
     },
     {
       "epoch": 1.56,
-      "learning_rate": 0.00020699999999999996,
-      "loss": 4.5314,
+      "learning_rate": 0.00020759999999999998,
+      "loss": 4.0309,
       "step": 348
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.00020759999999999998,
-      "loss": 4.4604,
+      "learning_rate": 0.00020819999999999996,
+      "loss": 4.0944,
       "step": 349
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.00020819999999999996,
-      "loss": 4.3669,
+      "learning_rate": 0.00020879999999999998,
+      "loss": 4.0536,
       "step": 350
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.00020879999999999998,
-      "loss": 4.2684,
+      "learning_rate": 0.00020939999999999997,
+      "loss": 4.0323,
       "step": 351
     },
     {
       "epoch": 1.58,
-      "learning_rate": 0.00020939999999999997,
-      "loss": 4.3108,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 4.041,
       "step": 352
     },
     {
       "epoch": 1.58,
-      "learning_rate": 0.00020999999999999998,
-      "loss": 4.3765,
+      "learning_rate": 0.00021059999999999997,
+      "loss": 4.0926,
       "step": 353
     },
     {
       "epoch": 1.59,
-      "learning_rate": 0.00021059999999999997,
-      "loss": 4.5628,
+      "learning_rate": 0.00021119999999999996,
+      "loss": 4.053,
       "step": 354
     },
     {
       "epoch": 1.59,
-      "learning_rate": 0.00021119999999999996,
-      "loss": 4.3952,
+      "learning_rate": 0.00021179999999999997,
+      "loss": 3.9695,
       "step": 355
     },
     {
       "epoch": 1.6,
-      "learning_rate": 0.00021179999999999997,
-      "loss": 4.2404,
+      "learning_rate": 0.00021239999999999996,
+      "loss": 3.9604,
       "step": 356
     },
     {
       "epoch": 1.6,
-      "learning_rate": 0.00021239999999999996,
-      "loss": 4.1993,
+      "learning_rate": 0.00021299999999999997,
+      "loss": 3.978,
       "step": 357
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.00021299999999999997,
-      "loss": 4.24,
+      "learning_rate": 0.00021359999999999996,
+      "loss": 3.9869,
       "step": 358
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.00021359999999999996,
-      "loss": 4.1699,
+      "learning_rate": 0.00021419999999999998,
+      "loss": 3.983,
       "step": 359
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.00021419999999999998,
-      "loss": 4.2247,
+      "learning_rate": 0.00021479999999999996,
+      "loss": 4.0209,
       "step": 360
     },
     {
       "epoch": 1.62,
-      "learning_rate": 0.00021479999999999996,
-      "loss": 4.2105,
+      "learning_rate": 0.00021539999999999998,
+      "loss": 4.0257,
       "step": 361
     },
     {
       "epoch": 1.62,
-      "learning_rate": 0.00021539999999999998,
-      "loss": 4.0637,
+      "learning_rate": 0.00021599999999999996,
+      "loss": 3.9414,
       "step": 362
     },
     {
       "epoch": 1.63,
-      "learning_rate": 0.00021599999999999996,
-      "loss": 4.1486,
+      "learning_rate": 0.00021659999999999998,
+      "loss": 4.0119,
       "step": 363
     },
     {
       "epoch": 1.63,
-      "learning_rate": 0.00021659999999999998,
-      "loss": 4.0632,
+      "learning_rate": 0.00021719999999999997,
+      "loss": 3.9501,
       "step": 364
     },
     {
       "epoch": 1.64,
-      "learning_rate": 0.00021719999999999997,
-      "loss": 4.0394,
+      "learning_rate": 0.00021779999999999998,
+      "loss": 3.915,
       "step": 365
     },
     {
       "epoch": 1.64,
-      "learning_rate": 0.00021779999999999998,
-      "loss": 4.1295,
+      "learning_rate": 0.00021839999999999997,
+      "loss": 4.018,
       "step": 366
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.00021839999999999997,
-      "loss": 4.1444,
+      "learning_rate": 0.00021899999999999998,
+      "loss": 3.9956,
       "step": 367
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.00021899999999999998,
-      "loss": 4.2003,
+      "learning_rate": 0.00021959999999999997,
+      "loss": 4.0166,
       "step": 368
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.00021959999999999997,
-      "loss": 4.0887,
+      "learning_rate": 0.00022019999999999999,
+      "loss": 3.98,
       "step": 369
     },
     {
       "epoch": 1.66,
-      "learning_rate": 0.00022019999999999999,
-      "loss": 3.9279,
+      "learning_rate": 0.00022079999999999997,
+      "loss": 3.7089,
       "step": 370
     },
     {
       "epoch": 1.66,
-      "learning_rate": 0.00022079999999999997,
-      "loss": 3.9932,
+      "learning_rate": 0.0002214,
+      "loss": 3.8226,
       "step": 371
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.0002214,
-      "loss": 3.7584,
+      "learning_rate": 0.00022199999999999998,
+      "loss": 3.6044,
       "step": 372
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.0002214,
-      "loss": 3.4602,
+      "learning_rate": 0.0002226,
+      "loss": 3.2785,
       "step": 373
     },
     {
       "epoch": 1.68,
-      "learning_rate": 0.00022199999999999998,
-      "loss": 5.1713,
+      "learning_rate": 0.00022319999999999998,
+      "loss": 5.0042,
       "step": 374
     },
     {
       "epoch": 1.68,
-      "learning_rate": 0.0002226,
-      "loss": 4.9749,
+      "learning_rate": 0.0002238,
+      "loss": 4.8161,
       "step": 375
     },
     {
       "epoch": 1.69,
-      "learning_rate": 0.00022319999999999998,
-      "loss": 4.77,
+      "learning_rate": 0.00022439999999999998,
+      "loss": 4.6441,
       "step": 376
     },
     {
       "epoch": 1.69,
-      "learning_rate": 0.0002238,
-      "loss": 4.525,
+      "learning_rate": 0.000225,
+      "loss": 4.3838,
       "step": 377
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.00022439999999999998,
-      "loss": 4.4688,
+      "learning_rate": 0.00022559999999999998,
+      "loss": 4.3542,
       "step": 378
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.000225,
-      "loss": 4.3824,
+      "learning_rate": 0.00022619999999999997,
+      "loss": 4.231,
       "step": 379
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.00022559999999999998,
-      "loss": 4.4258,
+      "learning_rate": 0.00022679999999999998,
+      "loss": 4.2993,
       "step": 380
     },
     {
       "epoch": 1.71,
-      "learning_rate": 0.00022619999999999997,
-      "loss": 4.4924,
+      "learning_rate": 0.00022739999999999997,
+      "loss": 4.2949,
       "step": 381
     },
     {
       "epoch": 1.71,
-      "learning_rate": 0.00022679999999999998,
-      "loss": 4.3423,
+      "learning_rate": 0.00022799999999999999,
+      "loss": 4.2101,
       "step": 382
     },
     {
       "epoch": 1.72,
-      "learning_rate": 0.00022739999999999997,
-      "loss": 4.439,
+      "learning_rate": 0.00022859999999999997,
+      "loss": 4.3218,
       "step": 383
     },
     {
       "epoch": 1.72,
-      "learning_rate": 0.00022799999999999999,
-      "loss": 4.2055,
+      "learning_rate": 0.0002292,
+      "loss": 4.1069,
       "step": 384
     },
     {
       "epoch": 1.73,
-      "learning_rate": 0.00022859999999999997,
-      "loss": 4.1395,
+      "learning_rate": 0.00022979999999999997,
+      "loss": 4.0308,
       "step": 385
     },
     {
       "epoch": 1.73,
-      "learning_rate": 0.0002292,
-      "loss": 4.1919,
+      "learning_rate": 0.0002304,
+      "loss": 4.0998,
       "step": 386
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.00022979999999999997,
-      "loss": 4.1744,
+      "learning_rate": 0.00023099999999999998,
+      "loss": 4.0487,
       "step": 387
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.0002304,
-      "loss": 4.1852,
+      "learning_rate": 0.0002316,
+      "loss": 4.0931,
       "step": 388
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.00023099999999999998,
-      "loss": 4.1599,
+      "learning_rate": 0.00023219999999999998,
+      "loss": 4.0471,
       "step": 389
     },
     {
       "epoch": 1.75,
-      "learning_rate": 0.0002316,
-      "loss": 4.1255,
+      "learning_rate": 0.0002328,
+      "loss": 4.0103,
       "step": 390
     },
     {
       "epoch": 1.75,
-      "learning_rate": 0.00023219999999999998,
-      "loss": 4.1317,
+      "learning_rate": 0.00023339999999999998,
+      "loss": 4.0283,
       "step": 391
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.0002328,
-      "loss": 4.1679,
+      "learning_rate": 0.000234,
+      "loss": 4.0728,
       "step": 392
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.00023339999999999998,
-      "loss": 4.1555,
+      "learning_rate": 0.00023459999999999998,
+      "loss": 4.0685,
       "step": 393
     },
     {
       "epoch": 1.77,
-      "learning_rate": 0.000234,
-      "loss": 4.1475,
+      "learning_rate": 0.0002352,
+      "loss": 4.0599,
       "step": 394
     },
     {
       "epoch": 1.77,
-      "learning_rate": 0.00023459999999999998,
-      "loss": 4.1145,
+      "learning_rate": 0.00023579999999999999,
+      "loss": 4.0421,
       "step": 395
     },
     {
       "epoch": 1.78,
-      "learning_rate": 0.0002352,
-      "loss": 4.0143,
+      "learning_rate": 0.0002364,
+      "loss": 3.9362,
       "step": 396
     },
     {
       "epoch": 1.78,
-      "learning_rate": 0.00023579999999999999,
-      "loss": 4.1586,
+      "learning_rate": 0.000237,
+      "loss": 4.08,
       "step": 397
     },
     {
       "epoch": 1.78,
-      "learning_rate": 0.0002364,
-      "loss": 4.0736,
+      "learning_rate": 0.0002376,
+      "loss": 4.0144,
       "step": 398
     },
     {
       "epoch": 1.79,
-      "learning_rate": 0.000237,
-      "loss": 4.1268,
+      "learning_rate": 0.0002382,
+      "loss": 4.0673,
       "step": 399
     },
     {
       "epoch": 1.79,
-      "learning_rate": 0.0002376,
-      "loss": 4.1166,
+      "learning_rate": 0.0002388,
+      "loss": 4.0356,
       "step": 400
     },
     {
       "epoch": 1.8,
-      "learning_rate": 0.0002382,
-      "loss": 4.0384,
+      "learning_rate": 0.0002394,
+      "loss": 3.9756,
       "step": 401
     },
     {
       "epoch": 1.8,
-      "learning_rate": 0.0002388,
-      "loss": 3.9957,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 3.9296,
       "step": 402
     },
     {
       "epoch": 1.81,
-      "learning_rate": 0.0002394,
-      "loss": 4.0793,
+      "learning_rate": 0.0002406,
+      "loss": 3.9861,
       "step": 403
     },
     {
       "epoch": 1.81,
-      "learning_rate": 0.00023999999999999998,
-      "loss": 4.1288,
+      "learning_rate": 0.00024119999999999998,
+      "loss": 4.0462,
       "step": 404
     },
     {
       "epoch": 1.82,
-      "learning_rate": 0.0002406,
-      "loss": 3.9214,
+      "learning_rate": 0.0002418,
+      "loss": 3.82,
       "step": 405
     },
     {
       "epoch": 1.82,
-      "learning_rate": 0.00024119999999999998,
-      "loss": 4.0041,
+      "learning_rate": 0.00024239999999999998,
+      "loss": 3.9398,
       "step": 406
     },
     {
       "epoch": 1.83,
-      "learning_rate": 0.0002418,
-      "loss": 4.0388,
+      "learning_rate": 0.000243,
+      "loss": 3.9655,
       "step": 407
     },
     {
       "epoch": 1.83,
-      "learning_rate": 0.00024239999999999998,
-      "loss": 3.989,
+      "learning_rate": 0.00024359999999999999,
+      "loss": 3.8895,
       "step": 408
     },
     {
       "epoch": 1.83,
-      "learning_rate": 0.000243,
-      "loss": 4.0942,
+      "learning_rate": 0.00024419999999999997,
+      "loss": 4.0266,
       "step": 409
     },
     {
       "epoch": 1.84,
-      "learning_rate": 0.00024359999999999999,
-      "loss": 3.9554,
+      "learning_rate": 0.0002448,
+      "loss": 3.8918,
       "step": 410
     },
     {
       "epoch": 1.84,
-      "learning_rate": 0.00024419999999999997,
-      "loss": 4.0312,
+      "learning_rate": 0.00024539999999999995,
+      "loss": 3.9576,
       "step": 411
     },
     {
       "epoch": 1.85,
-      "learning_rate": 0.0002448,
-      "loss": 3.9571,
+      "learning_rate": 0.00024599999999999996,
+      "loss": 3.8606,
       "step": 412
     },
     {
       "epoch": 1.85,
-      "learning_rate": 0.00024539999999999995,
-      "loss": 4.0576,
+      "learning_rate": 0.0002466,
+      "loss": 3.9656,
       "step": 413
     },
     {
       "epoch": 1.86,
-      "learning_rate": 0.00024599999999999996,
-      "loss": 3.9382,
+      "learning_rate": 0.0002472,
+      "loss": 3.841,
       "step": 414
     },
     {
       "epoch": 1.86,
-      "learning_rate": 0.0002466,
-      "loss": 3.8844,
+      "learning_rate": 0.00024779999999999995,
+      "loss": 3.7759,
       "step": 415
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.0002472,
-      "loss": 3.8587,
+      "learning_rate": 0.00024839999999999997,
+      "loss": 3.756,
       "step": 416
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.00024779999999999995,
-      "loss": 3.8616,
+      "learning_rate": 0.000249,
+      "loss": 3.7692,
       "step": 417
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.00024839999999999997,
-      "loss": 3.8594,
+      "learning_rate": 0.00024959999999999994,
+      "loss": 3.727,
       "step": 418
     },
     {
       "epoch": 1.88,
-      "learning_rate": 0.000249,
-      "loss": 3.698,
+      "learning_rate": 0.00025019999999999996,
+      "loss": 3.5521,
       "step": 419
     },
     {
       "epoch": 1.88,
-      "learning_rate": 0.00024959999999999994,
-      "loss": 3.7176,
+      "learning_rate": 0.00025079999999999997,
+      "loss": 3.5871,
       "step": 420
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.00025019999999999996,
-      "loss": 3.5745,
+      "learning_rate": 0.0002514,
+      "loss": 3.4838,
       "step": 421
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.00025079999999999997,
-      "loss": 3.5771,
+      "learning_rate": 0.00025199999999999995,
+      "loss": 3.456,
       "step": 422
     },
     {
       "epoch": 1.9,
-      "learning_rate": 0.0002514,
-      "loss": 3.1711,
+      "learning_rate": 0.00025259999999999996,
+      "loss": 2.9289,
       "step": 423
     },
     {
       "epoch": 1.9,
-      "learning_rate": 0.00025199999999999995,
-      "loss": 5.6223,
+      "learning_rate": 0.0002532,
+      "loss": 6.1709,
       "step": 424
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.00025259999999999996,
-      "loss": 5.5827,
+      "learning_rate": 0.0002538,
+      "loss": 5.5822,
       "step": 425
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.0002532,
-      "loss": 5.1421,
+      "learning_rate": 0.00025439999999999995,
+      "loss": 4.9478,
       "step": 426
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.0002538,
-      "loss": 4.695,
+      "learning_rate": 0.00025499999999999996,
+      "loss": 4.5524,
       "step": 427
     },
     {
       "epoch": 1.92,
-      "learning_rate": 0.00025439999999999995,
-      "loss": 4.5779,
+      "learning_rate": 0.0002556,
+      "loss": 4.4286,
       "step": 428
     },
     {
       "epoch": 1.92,
-      "learning_rate": 0.00025499999999999996,
-      "loss": 4.5219,
+      "learning_rate": 0.0002562,
+      "loss": 4.3649,
       "step": 429
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.0002556,
-      "loss": 4.2696,
+      "learning_rate": 0.00025679999999999995,
+      "loss": 4.143,
       "step": 430
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.0002562,
-      "loss": 4.1901,
+      "learning_rate": 0.00025739999999999997,
+      "loss": 4.1339,
       "step": 431
     },
     {
       "epoch": 1.94,
-      "learning_rate": 0.00025679999999999995,
-      "loss": 4.1369,
+      "learning_rate": 0.000258,
+      "loss": 4.0862,
       "step": 432
     },
     {
       "epoch": 1.94,
-      "learning_rate": 0.00025739999999999997,
-      "loss": 4.0599,
+      "learning_rate": 0.0002586,
+      "loss": 3.9971,
       "step": 433
     },
     {
       "epoch": 1.95,
-      "learning_rate": 0.000258,
-      "loss": 4.1342,
+      "learning_rate": 0.00025919999999999996,
+      "loss": 4.0285,
       "step": 434
     },
     {
       "epoch": 1.95,
-      "learning_rate": 0.0002586,
-      "loss": 4.1208,
+      "learning_rate": 0.00025979999999999997,
+      "loss": 4.0671,
       "step": 435
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.00025919999999999996,
-      "loss": 4.0045,
+      "learning_rate": 0.0002604,
+      "loss": 3.9465,
       "step": 436
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.00025979999999999997,
-      "loss": 4.0492,
+      "learning_rate": 0.000261,
+      "loss": 3.9439,
       "step": 437
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.0002604,
-      "loss": 4.0202,
+      "learning_rate": 0.00026159999999999996,
+      "loss": 3.91,
       "step": 438
     },
     {
       "epoch": 1.97,
-      "learning_rate": 0.000261,
-      "loss": 3.9223,
+      "learning_rate": 0.0002622,
+      "loss": 3.819,
       "step": 439
     },
     {
       "epoch": 1.97,
-      "learning_rate": 0.00026159999999999996,
-      "loss": 3.9732,
+      "learning_rate": 0.0002628,
+      "loss": 3.8625,
       "step": 440
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.0002622,
-      "loss": 3.9046,
+      "learning_rate": 0.00026339999999999995,
+      "loss": 3.7986,
       "step": 441
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.0002628,
-      "loss": 3.9394,
+      "learning_rate": 0.00026399999999999997,
+      "loss": 3.8111,
       "step": 442
     },
     {
       "epoch": 1.99,
-      "learning_rate": 0.00026339999999999995,
-      "loss": 3.8511,
+      "learning_rate": 0.0002646,
+      "loss": 3.739,
       "step": 443
     },
     {
       "epoch": 1.99,
-      "learning_rate": 0.00026399999999999997,
-      "loss": 3.7997,
+      "learning_rate": 0.0002652,
+      "loss": 3.6672,
       "step": 444
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.0002646,
-      "loss": 3.6102,
+      "learning_rate": 0.00026579999999999996,
+      "loss": 3.5265,
       "step": 445
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.0002652,
-      "loss": 3.2133,
+      "learning_rate": 0.00026639999999999997,
+      "loss": 3.0454,
       "step": 446
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00026579999999999996,
-      "loss": 5.3176,
+      "learning_rate": 0.000267,
+      "loss": 5.4695,
       "step": 447
     },
     {
       "epoch": 2.01,
-      "learning_rate": 0.00026639999999999997,
-      "loss": 5.2469,
+      "learning_rate": 0.0002676,
+      "loss": 5.3123,
       "step": 448
     },
     {
       "epoch": 2.01,
-      "learning_rate": 0.000267,
-      "loss": 4.9693,
+      "learning_rate": 0.00026819999999999996,
+      "loss": 4.9094,
       "step": 449
     },
     {
       "epoch": 2.02,
-      "learning_rate": 0.0002676,
-      "loss": 4.5121,
+      "learning_rate": 0.0002688,
+      "loss": 4.4246,
       "step": 450
     },
     {
       "epoch": 2.02,
-      "learning_rate": 0.00026819999999999996,
-      "loss": 4.5022,
+      "learning_rate": 0.0002694,
+      "loss": 4.3338,
       "step": 451
     },
     {
       "epoch": 2.03,
-      "learning_rate": 0.0002688,
-      "loss": 4.2283,
+      "learning_rate": 0.00027,
+      "loss": 4.0653,
       "step": 452
     },
     {
       "epoch": 2.03,
-      "learning_rate": 0.0002694,
-      "loss": 4.0816,
+      "learning_rate": 0.00027059999999999996,
+      "loss": 3.9293,
       "step": 453
     },
     {
       "epoch": 2.04,
-      "learning_rate": 0.00027,
-      "loss": 4.091,
+      "learning_rate": 0.0002712,
+      "loss": 3.924,
       "step": 454
     },
     {
       "epoch": 2.04,
-      "learning_rate": 0.00027059999999999996,
-      "loss": 4.0024,
+      "learning_rate": 0.0002718,
+      "loss": 3.8679,
       "step": 455
     },
     {
       "epoch": 2.04,
-      "learning_rate": 0.0002712,
-      "loss": 3.9794,
+      "learning_rate": 0.0002724,
+      "loss": 3.8316,
       "step": 456
     },
     {
       "epoch": 2.05,
-      "learning_rate": 0.0002718,
-      "loss": 3.8077,
+      "learning_rate": 0.00027299999999999997,
+      "loss": 3.6439,
       "step": 457
     },
     {
       "epoch": 2.05,
-      "learning_rate": 0.0002724,
-      "loss": 3.86,
+      "learning_rate": 0.0002736,
+      "loss": 3.6644,
       "step": 458
     },
     {
       "epoch": 2.06,
-      "learning_rate": 0.00027299999999999997,
-      "loss": 3.6727,
+      "learning_rate": 0.0002742,
+      "loss": 3.4917,
       "step": 459
     },
     {
       "epoch": 2.06,
-      "learning_rate": 0.0002736,
-      "loss": 3.7923,
+      "learning_rate": 0.0002748,
+      "loss": 3.5827,
       "step": 460
     },
     {
       "epoch": 2.07,
-      "learning_rate": 0.0002742,
-      "loss": 3.7753,
+      "learning_rate": 0.00027539999999999997,
+      "loss": 3.5428,
       "step": 461
     },
     {
       "epoch": 2.07,
-      "learning_rate": 0.0002748,
-      "loss": 3.7465,
+      "learning_rate": 0.000276,
+      "loss": 3.5134,
       "step": 462
     },
     {
       "epoch": 2.08,
-      "learning_rate": 0.00027539999999999997,
-      "loss": 3.712,
+      "learning_rate": 0.0002766,
+      "loss": 3.5094,
       "step": 463
     },
     {
       "epoch": 2.08,
-      "learning_rate": 0.000276,
-      "loss": 3.5746,
+      "learning_rate": 0.0002772,
+      "loss": 3.4225,
       "step": 464
     },
     {
       "epoch": 2.09,
-      "learning_rate": 0.0002766,
-      "loss": 3.6256,
+      "learning_rate": 0.0002778,
+      "loss": 3.4751,
       "step": 465
     },
     {
       "epoch": 2.09,
-      "learning_rate": 0.0002772,
-      "loss": 3.5238,
+      "learning_rate": 0.0002784,
+      "loss": 3.3967,
       "step": 466
     },
     {
       "epoch": 2.09,
-      "learning_rate": 0.0002778,
-      "loss": 3.4411,
+      "learning_rate": 0.000279,
+      "loss": 3.3292,
       "step": 467
     },
     {
       "epoch": 2.1,
-      "learning_rate": 0.0002784,
-      "loss": 3.5065,
+      "learning_rate": 0.00027959999999999997,
+      "loss": 3.3373,
       "step": 468
     },
     {
       "epoch": 2.1,
-      "learning_rate": 0.000279,
-      "loss": 3.5197,
+      "learning_rate": 0.0002802,
+      "loss": 3.3579,
       "step": 469
     },
     {
       "epoch": 2.11,
-      "learning_rate": 0.00027959999999999997,
-      "loss": 3.4927,
+      "learning_rate": 0.0002808,
+      "loss": 3.3337,
       "step": 470
     },
     {
       "epoch": 2.11,
-      "learning_rate": 0.0002802,
-      "loss": 3.5846,
+      "learning_rate": 0.00028139999999999996,
+      "loss": 3.3649,
       "step": 471
     },
     {
       "epoch": 2.12,
-      "learning_rate": 0.0002808,
-      "loss": 3.4224,
+      "learning_rate": 0.00028199999999999997,
+      "loss": 3.2899,
       "step": 472
     },
     {
       "epoch": 2.12,
-      "learning_rate": 0.00028139999999999996,
-      "loss": 3.4171,
+      "learning_rate": 0.0002826,
+      "loss": 3.2992,
       "step": 473
     },
     {
       "epoch": 2.13,
-      "learning_rate": 0.00028199999999999997,
-      "loss": 3.3167,
+      "learning_rate": 0.00028319999999999994,
+      "loss": 3.197,
       "step": 474
     },
     {
       "epoch": 2.13,
-      "learning_rate": 0.0002826,
-      "loss": 3.4456,
+      "learning_rate": 0.00028379999999999996,
+      "loss": 3.2116,
       "step": 475
     },
     {
       "epoch": 2.13,
-      "learning_rate": 0.00028319999999999994,
-      "loss": 3.4501,
+      "learning_rate": 0.0002844,
+      "loss": 3.251,
       "step": 476
     },
     {
       "epoch": 2.14,
-      "learning_rate": 0.00028379999999999996,
-      "loss": 3.3677,
+      "learning_rate": 0.000285,
+      "loss": 3.2578,
       "step": 477
     },
     {
       "epoch": 2.14,
-      "learning_rate": 0.0002844,
-      "loss": 3.2927,
+      "learning_rate": 0.00028559999999999995,
+      "loss": 3.2191,
       "step": 478
     },
     {
       "epoch": 2.15,
-      "learning_rate": 0.000285,
-      "loss": 3.3387,
+      "learning_rate": 0.00028619999999999996,
+      "loss": 3.1944,
       "step": 479
     },
     {
       "epoch": 2.15,
-      "learning_rate": 0.00028559999999999995,
-      "loss": 3.3384,
+      "learning_rate": 0.0002868,
+      "loss": 3.2094,
       "step": 480
     },
     {
       "epoch": 2.16,
-      "learning_rate": 0.00028619999999999996,
-      "loss": 3.2779,
+      "learning_rate": 0.00028739999999999994,
+      "loss": 3.1629,
       "step": 481
     },
     {
       "epoch": 2.16,
-      "learning_rate": 0.0002868,
-      "loss": 3.2192,
+      "learning_rate": 0.00028799999999999995,
+      "loss": 3.114,
       "step": 482
     },
     {
       "epoch": 2.17,
-      "learning_rate": 0.00028739999999999994,
-      "loss": 3.2844,
+      "learning_rate": 0.00028859999999999997,
+      "loss": 3.1973,
       "step": 483
     },
     {
       "epoch": 2.17,
-      "learning_rate": 0.00028799999999999995,
-      "loss": 3.178,
+      "learning_rate": 0.0002892,
+      "loss": 3.0682,
       "step": 484
     },
     {
       "epoch": 2.17,
-      "learning_rate": 0.00028859999999999997,
-      "loss": 3.179,
+      "learning_rate": 0.00028979999999999994,
+      "loss": 3.1276,
       "step": 485
     },
     {
       "epoch": 2.18,
-      "learning_rate": 0.0002892,
-      "loss": 3.1362,
+      "learning_rate": 0.00029039999999999996,
+      "loss": 3.0697,
       "step": 486
     },
     {
       "epoch": 2.18,
-      "learning_rate": 0.00028979999999999994,
-      "loss": 3.1162,
+      "learning_rate": 0.00029099999999999997,
+      "loss": 3.0525,
       "step": 487
     },
     {
       "epoch": 2.19,
-      "learning_rate": 0.00029039999999999996,
-      "loss": 3.0807,
+      "learning_rate": 0.0002916,
+      "loss": 3.0135,
       "step": 488
     },
     {
       "epoch": 2.19,
-      "learning_rate": 0.00029099999999999997,
-      "loss": 3.0098,
+      "learning_rate": 0.00029219999999999995,
+      "loss": 2.8437,
       "step": 489
     },
     {
       "epoch": 2.2,
-      "learning_rate": 0.0002916,
-      "loss": 2.8276,
+      "learning_rate": 0.00029279999999999996,
+      "loss": 2.7921,
       "step": 490
     },
     {
       "epoch": 2.2,
-      "learning_rate": 0.00029219999999999995,
-      "loss": 2.8082,
+      "learning_rate": 0.0002934,
+      "loss": 2.6695,
       "step": 491
     },
     {
       "epoch": 2.21,
-      "learning_rate": 0.00029279999999999996,
-      "loss": 2.7419,
+      "learning_rate": 0.000294,
+      "loss": 2.7093,
       "step": 492
     },
     {
       "epoch": 2.21,
-      "learning_rate": 0.0002934,
-      "loss": 2.6181,
+      "learning_rate": 0.00029459999999999995,
+      "loss": 2.5491,
       "step": 493
     },
     {
       "epoch": 2.22,
-      "learning_rate": 0.000294,
-      "loss": 2.5094,
+      "learning_rate": 0.00029519999999999997,
+      "loss": 2.4168,
       "step": 494
     },
     {
       "epoch": 2.22,
-      "learning_rate": 0.00029459999999999995,
-      "loss": 2.455,
+      "learning_rate": 0.0002958,
+      "loss": 2.3929,
       "step": 495
     },
     {
       "epoch": 2.22,
-      "learning_rate": 0.00029519999999999997,
-      "loss": 2.1926,
+      "learning_rate": 0.0002964,
+      "loss": 2.0101,
       "step": 496
     },
     {
       "epoch": 2.23,
-      "learning_rate": 0.0002958,
-      "loss": 6.3152,
+      "learning_rate": 0.00029699999999999996,
+      "loss": 6.3452,
       "step": 497
     },
     {
       "epoch": 2.23,
-      "learning_rate": 0.0002964,
-      "loss": 5.1198,
+      "learning_rate": 0.00029759999999999997,
+      "loss": 5.654,
       "step": 498
     },
     {
       "epoch": 2.24,
-      "learning_rate": 0.00029699999999999996,
-      "loss": 4.7265,
+      "learning_rate": 0.0002982,
+      "loss": 4.9005,
       "step": 499
     },
     {
       "epoch": 2.24,
-      "learning_rate": 0.00029759999999999997,
-      "loss": 4.3539,
+      "learning_rate": 0.0002988,
+      "loss": 4.4437,
       "step": 500
     },
     {
       "epoch": 2.24,
-      "eval_loss": 4.050220966339111,
-      "eval_runtime": 594.0081,
-      "eval_samples_per_second": 4.448,
-      "eval_steps_per_second": 0.557,
-      "eval_wer": 1.7867314557715193,
+      "eval_loss": 3.9967808723449707,
+      "eval_runtime": 586.6262,
+      "eval_samples_per_second": 4.504,
+      "eval_steps_per_second": 0.564,
+      "eval_wer": 1.6151527171757238,
       "step": 500
     },
     {
       "epoch": 2.25,
-      "learning_rate": 0.0002982,
-      "loss": 4.1344,
+      "learning_rate": 0.00029939999999999996,
+      "loss": 4.117,
       "step": 501
     },
     {
       "epoch": 2.25,
-      "learning_rate": 0.0002988,
-      "loss": 4.2398,
+      "learning_rate": 0.0003,
+      "loss": 3.9755,
       "step": 502
     },
     {
       "epoch": 2.26,
-      "learning_rate": 0.00029939999999999996,
-      "loss": 4.0204,
+      "learning_rate": 0.00029982658959537567,
+      "loss": 3.8336,
       "step": 503
     },
     {
       "epoch": 2.26,
-      "learning_rate": 0.0003,
-      "loss": 4.032,
+      "learning_rate": 0.0002996531791907514,
+      "loss": 3.8994,
       "step": 504
     },
     {
       "epoch": 2.26,
-      "learning_rate": 0.0002995121951219512,
-      "loss": 3.9013,
+      "learning_rate": 0.00029947976878612716,
+      "loss": 3.765,
       "step": 505
     },
     {
       "epoch": 2.27,
-      "learning_rate": 0.0002990243902439024,
-      "loss": 3.7927,
+      "learning_rate": 0.00029930635838150286,
+      "loss": 3.677,
       "step": 506
     },
     {
       "epoch": 2.27,
-      "learning_rate": 0.0002985365853658536,
-      "loss": 3.7671,
+      "learning_rate": 0.0002991329479768786,
+      "loss": 3.6203,
       "step": 507
     },
     {
       "epoch": 2.28,
-      "learning_rate": 0.00029804878048780484,
-      "loss": 3.7971,
+      "learning_rate": 0.0002989595375722543,
+      "loss": 3.6362,
       "step": 508
     },
     {
       "epoch": 2.28,
-      "learning_rate": 0.00029756097560975606,
-      "loss": 3.7401,
+      "learning_rate": 0.00029878612716763005,
+      "loss": 3.5662,
       "step": 509
     },
     {
       "epoch": 2.29,
-      "learning_rate": 0.0002970731707317073,
-      "loss": 3.5998,
+      "learning_rate": 0.00029861271676300574,
+      "loss": 3.4386,
       "step": 510
     },
     {
       "epoch": 2.29,
-      "learning_rate": 0.0002965853658536585,
-      "loss": 3.6632,
+      "learning_rate": 0.0002984393063583815,
+      "loss": 3.4955,
       "step": 511
     },
     {
       "epoch": 2.3,
-      "learning_rate": 0.0002960975609756097,
-      "loss": 3.6384,
+      "learning_rate": 0.0002982658959537572,
+      "loss": 3.463,
       "step": 512
     },
     {
       "epoch": 2.3,
-      "learning_rate": 0.0002956097560975609,
-      "loss": 3.6145,
+      "learning_rate": 0.00029809248554913293,
+      "loss": 3.4607,
       "step": 513
     },
     {
       "epoch": 2.3,
-      "learning_rate": 0.0002951219512195122,
-      "loss": 3.5137,
+      "learning_rate": 0.0002979190751445086,
+      "loss": 3.4455,
       "step": 514
     },
     {
       "epoch": 2.31,
-      "learning_rate": 0.00029463414634146336,
-      "loss": 3.5113,
+      "learning_rate": 0.00029774566473988437,
+      "loss": 3.4025,
       "step": 515
     },
     {
       "epoch": 2.31,
-      "learning_rate": 0.0002941463414634146,
-      "loss": 3.4298,
+      "learning_rate": 0.00029757225433526006,
+      "loss": 3.2991,
       "step": 516
     },
     {
       "epoch": 2.32,
-      "learning_rate": 0.00029365853658536585,
-      "loss": 3.3688,
+      "learning_rate": 0.0002973988439306358,
+      "loss": 3.2948,
       "step": 517
     },
     {
       "epoch": 2.32,
-      "learning_rate": 0.00029317073170731706,
-      "loss": 3.3792,
+      "learning_rate": 0.00029722543352601156,
+      "loss": 3.3225,
       "step": 518
     },
     {
       "epoch": 2.33,
-      "learning_rate": 0.0002926829268292683,
-      "loss": 3.4692,
+      "learning_rate": 0.00029705202312138725,
+      "loss": 3.1961,
       "step": 519
     },
     {
       "epoch": 2.33,
-      "learning_rate": 0.0002921951219512195,
-      "loss": 3.3661,
+      "learning_rate": 0.00029687861271676295,
+      "loss": 3.2374,
       "step": 520
     },
     {
       "epoch": 2.34,
-      "learning_rate": 0.0002917073170731707,
-      "loss": 3.4411,
+      "learning_rate": 0.0002967052023121387,
+      "loss": 3.2768,
       "step": 521
     },
     {
       "epoch": 2.34,
-      "learning_rate": 0.00029121951219512193,
-      "loss": 3.509,
+      "learning_rate": 0.00029653179190751444,
+      "loss": 3.3283,
       "step": 522
     },
     {
       "epoch": 2.35,
-      "learning_rate": 0.00029073170731707315,
-      "loss": 3.3465,
+      "learning_rate": 0.00029635838150289014,
+      "loss": 3.2414,
       "step": 523
     },
     {
       "epoch": 2.35,
-      "learning_rate": 0.00029024390243902437,
-      "loss": 3.2509,
+      "learning_rate": 0.00029618497109826583,
+      "loss": 3.1435,
       "step": 524
     },
     {
       "epoch": 2.35,
-      "learning_rate": 0.0002897560975609756,
-      "loss": 3.1902,
+      "learning_rate": 0.0002960115606936416,
+      "loss": 3.1363,
       "step": 525
     },
     {
       "epoch": 2.36,
-      "learning_rate": 0.0002892682926829268,
-      "loss": 3.2894,
+      "learning_rate": 0.0002958381502890173,
+      "loss": 3.0984,
       "step": 526
     },
     {
       "epoch": 2.36,
-      "learning_rate": 0.000288780487804878,
-      "loss": 3.2102,
+      "learning_rate": 0.000295664739884393,
+      "loss": 3.1239,
       "step": 527
     },
     {
       "epoch": 2.37,
-      "learning_rate": 0.00028829268292682923,
-      "loss": 2.9926,
+      "learning_rate": 0.00029549132947976877,
+      "loss": 2.9329,
       "step": 528
     },
     {
       "epoch": 2.37,
-      "learning_rate": 0.00028780487804878045,
-      "loss": 3.1141,
+      "learning_rate": 0.0002953179190751445,
+      "loss": 3.0781,
       "step": 529
     },
     {
       "epoch": 2.38,
-      "learning_rate": 0.00028731707317073167,
-      "loss": 2.9693,
+      "learning_rate": 0.0002951445086705202,
+      "loss": 2.9579,
       "step": 530
     },
     {
       "epoch": 2.38,
-      "learning_rate": 0.0002868292682926829,
-      "loss": 3.0834,
+      "learning_rate": 0.0002949710982658959,
+      "loss": 3.0964,
       "step": 531
     },
     {
       "epoch": 2.39,
-      "learning_rate": 0.0002863414634146341,
-      "loss": 3.0293,
+      "learning_rate": 0.00029479768786127165,
+      "loss": 3.0365,
       "step": 532
     },
     {
       "epoch": 2.39,
-      "learning_rate": 0.00028585365853658537,
-      "loss": 2.9632,
+      "learning_rate": 0.0002946242774566474,
+      "loss": 2.962,
       "step": 533
     },
     {
       "epoch": 2.39,
-      "learning_rate": 0.00028536585365853654,
-      "loss": 2.8581,
+      "learning_rate": 0.0002944508670520231,
+      "loss": 2.882,
       "step": 534
     },
     {
       "epoch": 2.4,
-      "learning_rate": 0.00028487804878048775,
-      "loss": 2.6841,
+      "learning_rate": 0.00029427745664739884,
+      "loss": 2.7589,
       "step": 535
     },
     {
       "epoch": 2.4,
-      "learning_rate": 0.000284390243902439,
-      "loss": 2.7088,
+      "learning_rate": 0.00029410404624277453,
+      "loss": 2.8493,
       "step": 536
     },
     {
       "epoch": 2.41,
-      "learning_rate": 0.0002839024390243902,
-      "loss": 2.8209,
+      "learning_rate": 0.0002939306358381503,
+      "loss": 2.9264,
       "step": 537
     },
     {
       "epoch": 2.41,
-      "learning_rate": 0.00028341463414634146,
-      "loss": 2.5995,
+      "learning_rate": 0.00029375722543352597,
+      "loss": 2.6882,
       "step": 538
     },
     {
       "epoch": 2.42,
-      "learning_rate": 0.0002829268292682927,
-      "loss": 2.7186,
+      "learning_rate": 0.0002935838150289017,
+      "loss": 2.6279,
       "step": 539
     },
     {
       "epoch": 2.42,
-      "learning_rate": 0.00028243902439024384,
-      "loss": 2.3942,
+      "learning_rate": 0.0002934104046242774,
+      "loss": 2.6705,
       "step": 540
     },
     {
       "epoch": 2.43,
-      "learning_rate": 0.0002819512195121951,
-      "loss": 2.4166,
+      "learning_rate": 0.00029323699421965316,
+      "loss": 2.3889,
       "step": 541
     },
     {
       "epoch": 2.43,
-      "learning_rate": 0.0002814634146341463,
-      "loss": 2.1934,
+      "learning_rate": 0.00029306358381502886,
+      "loss": 2.2856,
       "step": 542
     },
     {
       "epoch": 2.43,
-      "learning_rate": 0.00028097560975609754,
-      "loss": 2.0537,
+      "learning_rate": 0.0002928901734104046,
+      "loss": 2.1915,
       "step": 543
     },
     {
       "epoch": 2.44,
-      "learning_rate": 0.00028048780487804876,
-      "loss": 1.8673,
+      "learning_rate": 0.0002927167630057803,
+      "loss": 2.028,
       "step": 544
     },
     {
       "epoch": 2.44,
-      "learning_rate": 0.00028,
-      "loss": 1.7229,
+      "learning_rate": 0.00029254335260115604,
+      "loss": 1.7781,
       "step": 545
     },
     {
       "epoch": 2.45,
-      "learning_rate": 0.0002795121951219512,
-      "loss": 1.6225,
+      "learning_rate": 0.0002923699421965318,
+      "loss": 1.6539,
       "step": 546
     },
     {
       "epoch": 2.45,
-      "learning_rate": 0.0002790243902439024,
-      "loss": 6.0962,
+      "learning_rate": 0.0002921965317919075,
+      "loss": 7.0245,
       "step": 547
     },
     {
       "epoch": 2.46,
-      "learning_rate": 0.0002785365853658536,
-      "loss": 5.1113,
+      "learning_rate": 0.0002920231213872832,
+      "loss": 6.3081,
       "step": 548
     },
     {
       "epoch": 2.46,
-      "learning_rate": 0.00027804878048780484,
-      "loss": 4.4441,
+      "learning_rate": 0.0002918497109826589,
+      "loss": 5.4681,
       "step": 549
     },
     {
       "epoch": 2.47,
-      "learning_rate": 0.00027756097560975606,
-      "loss": 3.93,
+      "learning_rate": 0.0002916763005780347,
+      "loss": 4.8535,
       "step": 550
     },
     {
       "epoch": 2.47,
-      "learning_rate": 0.0002770731707317073,
-      "loss": 3.6004,
+      "learning_rate": 0.00029150289017341037,
+      "loss": 4.2709,
       "step": 551
     },
     {
       "epoch": 2.48,
-      "learning_rate": 0.00027658536585365855,
-      "loss": 3.4879,
+      "learning_rate": 0.0002913294797687861,
+      "loss": 3.9271,
       "step": 552
     },
     {
       "epoch": 2.48,
-      "learning_rate": 0.0002760975609756097,
-      "loss": 3.3519,
+      "learning_rate": 0.00029115606936416186,
+      "loss": 3.7921,
       "step": 553
     },
     {
       "epoch": 2.48,
-      "learning_rate": 0.00027560975609756093,
-      "loss": 3.2981,
+      "learning_rate": 0.00029098265895953756,
+      "loss": 3.7069,
       "step": 554
     },
     {
       "epoch": 2.49,
-      "learning_rate": 0.0002751219512195122,
-      "loss": 3.1892,
+      "learning_rate": 0.00029080924855491325,
+      "loss": 3.6785,
       "step": 555
     },
     {
       "epoch": 2.49,
-      "learning_rate": 0.00027463414634146336,
-      "loss": 3.1885,
+      "learning_rate": 0.000290635838150289,
+      "loss": 3.6915,
       "step": 556
     },
     {
       "epoch": 2.5,
-      "learning_rate": 0.00027414634146341463,
-      "loss": 2.9167,
+      "learning_rate": 0.00029046242774566475,
+      "loss": 3.4802,
       "step": 557
     },
     {
       "epoch": 2.5,
-      "learning_rate": 0.00027365853658536585,
-      "loss": 2.8781,
+      "learning_rate": 0.00029028901734104044,
+      "loss": 3.393,
       "step": 558
     },
     {
       "epoch": 2.51,
-      "learning_rate": 0.000273170731707317,
-      "loss": 2.8923,
+      "learning_rate": 0.00029011560693641613,
+      "loss": 3.4273,
       "step": 559
     },
     {
       "epoch": 2.51,
-      "learning_rate": 0.0002726829268292683,
-      "loss": 2.6902,
+      "learning_rate": 0.0002899421965317919,
+      "loss": 3.2599,
       "step": 560
     },
     {
       "epoch": 2.52,
-      "learning_rate": 0.0002721951219512195,
-      "loss": 2.6606,
+      "learning_rate": 0.00028976878612716763,
+      "loss": 3.2463,
       "step": 561
     },
     {
       "epoch": 2.52,
-      "learning_rate": 0.0002717073170731707,
-      "loss": 2.5854,
+      "learning_rate": 0.0002895953757225433,
+      "loss": 3.1323,
       "step": 562
     },
     {
       "epoch": 2.52,
-      "learning_rate": 0.00027121951219512193,
-      "loss": 2.4889,
+      "learning_rate": 0.00028942196531791907,
+      "loss": 3.1298,
       "step": 563
     },
     {
       "epoch": 2.53,
-      "learning_rate": 0.00027073170731707315,
-      "loss": 2.3526,
+      "learning_rate": 0.00028924855491329476,
+      "loss": 3.0224,
       "step": 564
     },
     {
       "epoch": 2.53,
-      "learning_rate": 0.00027024390243902437,
-      "loss": 2.3056,
+      "learning_rate": 0.0002890751445086705,
+      "loss": 3.0445,
       "step": 565
     },
     {
       "epoch": 2.54,
-      "learning_rate": 0.0002697560975609756,
-      "loss": 2.2212,
+      "learning_rate": 0.0002889017341040462,
+      "loss": 2.9297,
       "step": 566
     },
     {
       "epoch": 2.54,
-      "learning_rate": 0.0002692682926829268,
-      "loss": 2.1994,
+      "learning_rate": 0.00028872832369942195,
+      "loss": 2.8817,
       "step": 567
     },
     {
       "epoch": 2.55,
-      "learning_rate": 0.000268780487804878,
-      "loss": 2.1812,
+      "learning_rate": 0.00028855491329479765,
+      "loss": 2.8181,
       "step": 568
     },
     {
       "epoch": 2.55,
-      "learning_rate": 0.00026829268292682924,
-      "loss": 2.058,
+      "learning_rate": 0.0002883815028901734,
+      "loss": 2.6842,
       "step": 569
     },
     {
       "epoch": 2.56,
-      "learning_rate": 0.00026780487804878045,
-      "loss": 1.9894,
+      "learning_rate": 0.0002882080924855491,
+      "loss": 2.6766,
       "step": 570
     },
     {
       "epoch": 2.56,
-      "learning_rate": 0.0002673170731707317,
-      "loss": 2.0365,
+      "learning_rate": 0.00028803468208092484,
+      "loss": 2.6896,
       "step": 571
     },
     {
       "epoch": 2.57,
-      "learning_rate": 0.0002668292682926829,
-      "loss": 1.9143,
+      "learning_rate": 0.00028786127167630053,
+      "loss": 2.6096,
       "step": 572
     },
     {
       "epoch": 2.57,
-      "learning_rate": 0.0002663414634146341,
-      "loss": 1.7815,
+      "learning_rate": 0.0002876878612716763,
+      "loss": 2.5341,
       "step": 573
     },
     {
       "epoch": 2.57,
-      "learning_rate": 0.0002658536585365854,
-      "loss": 1.728,
+      "learning_rate": 0.000287514450867052,
+      "loss": 2.4034,
       "step": 574
     },
     {
       "epoch": 2.58,
-      "learning_rate": 0.00026536585365853654,
-      "loss": 1.7936,
+      "learning_rate": 0.0002873410404624277,
+      "loss": 2.4183,
       "step": 575
     },
     {
       "epoch": 2.58,
-      "learning_rate": 0.0002648780487804878,
-      "loss": 1.7649,
+      "learning_rate": 0.0002871676300578034,
+      "loss": 2.3373,
       "step": 576
     },
     {
       "epoch": 2.59,
-      "learning_rate": 0.000264390243902439,
-      "loss": 1.6668,
+      "learning_rate": 0.00028699421965317916,
+      "loss": 2.3058,
       "step": 577
     },
     {
       "epoch": 2.59,
-      "learning_rate": 0.0002639024390243902,
-      "loss": 1.5596,
+      "learning_rate": 0.0002868208092485549,
+      "loss": 2.5114,
       "step": 578
     },
     {
       "epoch": 2.6,
-      "learning_rate": 0.00026341463414634146,
-      "loss": 1.5518,
+      "learning_rate": 0.0002866473988439306,
+      "loss": 2.2027,
       "step": 579
     },
     {
       "epoch": 2.6,
-      "learning_rate": 0.0002629268292682927,
-      "loss": 1.5524,
+      "learning_rate": 0.00028647398843930635,
+      "loss": 2.546,
       "step": 580
     },
     {
       "epoch": 2.61,
-      "learning_rate": 0.0002624390243902439,
-      "loss": 1.426,
+      "learning_rate": 0.0002863005780346821,
+      "loss": 2.2915,
       "step": 581
     },
     {
       "epoch": 2.61,
-      "learning_rate": 0.0002619512195121951,
-      "loss": 1.4211,
+      "learning_rate": 0.0002861271676300578,
+      "loss": 2.4333,
       "step": 582
     },
     {
       "epoch": 2.61,
-      "learning_rate": 0.0002614634146341463,
-      "loss": 1.4552,
+      "learning_rate": 0.0002859537572254335,
+      "loss": 2.3528,
       "step": 583
     },
     {
       "epoch": 2.62,
-      "learning_rate": 0.00026097560975609754,
-      "loss": 1.3657,
+      "learning_rate": 0.00028578034682080923,
+      "loss": 2.0364,
       "step": 584
     },
     {
       "epoch": 2.62,
-      "learning_rate": 0.00026048780487804876,
-      "loss": 1.3157,
+      "learning_rate": 0.000285606936416185,
+      "loss": 1.9785,
       "step": 585
     },
     {
       "epoch": 2.63,
-      "learning_rate": 0.00026,
-      "loss": 1.2056,
+      "learning_rate": 0.0002854335260115607,
+      "loss": 1.8719,
       "step": 586
     },
     {
       "epoch": 2.63,
-      "learning_rate": 0.0002595121951219512,
-      "loss": 1.1727,
+      "learning_rate": 0.00028526011560693637,
+      "loss": 1.7591,
       "step": 587
     },
     {
       "epoch": 2.64,
-      "learning_rate": 0.0002590243902439024,
-      "loss": 1.1081,
+      "learning_rate": 0.0002850867052023121,
+      "loss": 1.6455,
       "step": 588
     },
     {
       "epoch": 2.64,
-      "learning_rate": 0.00025853658536585363,
-      "loss": 1.1073,
+      "learning_rate": 0.00028491329479768786,
+      "loss": 1.6273,
       "step": 589
     },
     {
       "epoch": 2.65,
-      "learning_rate": 0.00025804878048780484,
-      "loss": 1.0629,
+      "learning_rate": 0.00028473988439306356,
+      "loss": 1.5585,
       "step": 590
     },
     {
       "epoch": 2.65,
-      "learning_rate": 0.00025756097560975606,
-      "loss": 1.0085,
+      "learning_rate": 0.0002845664739884393,
+      "loss": 1.3973,
       "step": 591
     },
     {
       "epoch": 2.65,
-      "learning_rate": 0.0002570731707317073,
-      "loss": 0.9956,
+      "learning_rate": 0.000284393063583815,
+      "loss": 1.3006,
       "step": 592
     },
     {
       "epoch": 2.66,
-      "learning_rate": 0.0002565853658536585,
-      "loss": 0.9413,
+      "learning_rate": 0.00028421965317919075,
+      "loss": 1.239,
       "step": 593
     },
     {
       "epoch": 2.66,
-      "learning_rate": 0.0002560975609756097,
-      "loss": 0.9303,
+      "learning_rate": 0.00028404624277456644,
+      "loss": 1.1947,
       "step": 594
     },
     {
       "epoch": 2.67,
-      "learning_rate": 0.000255609756097561,
-      "loss": 0.917,
+      "learning_rate": 0.0002838728323699422,
+      "loss": 1.1989,
       "step": 595
     },
     {
       "epoch": 2.67,
-      "learning_rate": 0.0002551219512195122,
-      "loss": 1.081,
+      "learning_rate": 0.0002836994219653179,
+      "loss": 1.1465,
       "step": 596
     },
     {
       "epoch": 2.68,
-      "learning_rate": 0.00025463414634146336,
-      "loss": 9.2655,
+      "learning_rate": 0.00028352601156069363,
+      "loss": 7.6295,
       "step": 597
     },
     {
       "epoch": 2.68,
-      "learning_rate": 0.00025414634146341463,
-      "loss": 6.3665,
+      "learning_rate": 0.0002833526011560693,
+      "loss": 5.1178,
       "step": 598
     },
     {
       "epoch": 2.69,
-      "learning_rate": 0.00025365853658536585,
-      "loss": 4.6114,
+      "learning_rate": 0.00028317919075144507,
+      "loss": 4.1009,
       "step": 599
     },
     {
       "epoch": 2.69,
-      "learning_rate": 0.00025317073170731707,
-      "loss": 3.6587,
+      "learning_rate": 0.00028300578034682076,
+      "loss": 3.297,
       "step": 600
     },
     {
       "epoch": 2.7,
-      "learning_rate": 0.0002526829268292683,
-      "loss": 3.0331,
+      "learning_rate": 0.0002828323699421965,
+      "loss": 3.0486,
       "step": 601
     },
     {
       "epoch": 2.7,
-      "learning_rate": 0.0002521951219512195,
-      "loss": 2.6576,
+      "learning_rate": 0.00028265895953757226,
+      "loss": 2.8205,
       "step": 602
     },
     {
       "epoch": 2.7,
-      "learning_rate": 0.0002517073170731707,
-      "loss": 2.4141,
+      "learning_rate": 0.00028248554913294795,
+      "loss": 2.6073,
       "step": 603
     },
     {
       "epoch": 2.71,
-      "learning_rate": 0.00025121951219512194,
-      "loss": 2.2096,
+      "learning_rate": 0.00028231213872832365,
+      "loss": 2.476,
       "step": 604
     },
     {
       "epoch": 2.71,
-      "learning_rate": 0.00025073170731707315,
-      "loss": 2.0265,
+      "learning_rate": 0.0002821387283236994,
+      "loss": 2.3003,
       "step": 605
     },
     {
       "epoch": 2.72,
-      "learning_rate": 0.00025024390243902437,
-      "loss": 1.9003,
+      "learning_rate": 0.00028196531791907514,
+      "loss": 2.2104,
       "step": 606
     },
     {
       "epoch": 2.72,
-      "learning_rate": 0.0002497560975609756,
-      "loss": 1.7156,
+      "learning_rate": 0.00028179190751445083,
+      "loss": 2.0377,
       "step": 607
     },
     {
       "epoch": 2.73,
-      "learning_rate": 0.0002492682926829268,
-      "loss": 1.6201,
+      "learning_rate": 0.0002816184971098266,
+      "loss": 1.9675,
       "step": 608
     },
     {
       "epoch": 2.73,
-      "learning_rate": 0.000248780487804878,
-      "loss": 1.4503,
+      "learning_rate": 0.0002814450867052023,
+      "loss": 1.7629,
       "step": 609
     },
     {
       "epoch": 2.74,
-      "learning_rate": 0.00024829268292682924,
-      "loss": 1.4161,
+      "learning_rate": 0.000281271676300578,
+      "loss": 1.7966,
       "step": 610
     },
     {
       "epoch": 2.74,
-      "learning_rate": 0.00024780487804878045,
-      "loss": 1.2728,
+      "learning_rate": 0.0002810982658959537,
+      "loss": 1.6484,
       "step": 611
     },
     {
       "epoch": 2.74,
-      "learning_rate": 0.00024731707317073167,
-      "loss": 1.2627,
+      "learning_rate": 0.00028092485549132947,
+      "loss": 1.5786,
       "step": 612
     },
     {
       "epoch": 2.75,
-      "learning_rate": 0.0002468292682926829,
-      "loss": 1.301,
+      "learning_rate": 0.00028075144508670516,
+      "loss": 1.5999,
       "step": 613
     },
     {
       "epoch": 2.75,
-      "learning_rate": 0.00024634146341463416,
-      "loss": 1.1987,
+      "learning_rate": 0.0002805780346820809,
+      "loss": 1.5781,
       "step": 614
     },
     {
       "epoch": 2.76,
-      "learning_rate": 0.0002458536585365853,
-      "loss": 1.0851,
+      "learning_rate": 0.0002804046242774566,
+      "loss": 1.4,
       "step": 615
     },
     {
       "epoch": 2.76,
-      "learning_rate": 0.00024536585365853654,
-      "loss": 1.207,
+      "learning_rate": 0.00028023121387283235,
+      "loss": 1.4785,
       "step": 616
     },
     {
       "epoch": 2.77,
-      "learning_rate": 0.0002448780487804878,
-      "loss": 1.0918,
+      "learning_rate": 0.00028005780346820804,
+      "loss": 1.3169,
       "step": 617
     },
     {
       "epoch": 2.77,
-      "learning_rate": 0.00024439024390243897,
-      "loss": 1.1469,
+      "learning_rate": 0.0002798843930635838,
+      "loss": 1.3737,
       "step": 618
     },
     {
       "epoch": 2.78,
-      "learning_rate": 0.00024390243902439022,
-      "loss": 0.9882,
+      "learning_rate": 0.00027971098265895954,
+      "loss": 1.3111,
       "step": 619
     },
     {
       "epoch": 2.78,
-      "learning_rate": 0.00024341463414634146,
-      "loss": 1.0057,
+      "learning_rate": 0.00027953757225433523,
+      "loss": 1.2543,
       "step": 620
     },
     {
       "epoch": 2.78,
-      "learning_rate": 0.00024292682926829268,
-      "loss": 0.9869,
+      "learning_rate": 0.0002793641618497109,
+      "loss": 1.1839,
       "step": 621
     },
     {
       "epoch": 2.79,
-      "learning_rate": 0.00024243902439024387,
-      "loss": 0.9118,
+      "learning_rate": 0.00027919075144508667,
+      "loss": 1.1236,
       "step": 622
     },
     {
       "epoch": 2.79,
-      "learning_rate": 0.0002419512195121951,
-      "loss": 0.8895,
+      "learning_rate": 0.0002790173410404624,
+      "loss": 1.1345,
       "step": 623
     },
     {
       "epoch": 2.8,
-      "learning_rate": 0.00024146341463414633,
-      "loss": 0.7842,
+      "learning_rate": 0.0002788439306358381,
+      "loss": 0.9673,
       "step": 624
     },
     {
       "epoch": 2.8,
-      "learning_rate": 0.00024097560975609755,
-      "loss": 0.8535,
+      "learning_rate": 0.00027867052023121386,
+      "loss": 1.0724,
       "step": 625
     },
     {
       "epoch": 2.81,
-      "learning_rate": 0.00024048780487804876,
-      "loss": 0.9101,
+      "learning_rate": 0.00027849710982658955,
+      "loss": 1.0625,
       "step": 626
     },
     {
       "epoch": 2.81,
-      "learning_rate": 0.00023999999999999998,
-      "loss": 0.9084,
+      "learning_rate": 0.0002783236994219653,
+      "loss": 1.1293,
       "step": 627
     },
     {
       "epoch": 2.82,
-      "learning_rate": 0.0002395121951219512,
-      "loss": 0.8409,
+      "learning_rate": 0.000278150289017341,
+      "loss": 0.9543,
       "step": 628
     },
     {
       "epoch": 2.82,
-      "learning_rate": 0.0002390243902439024,
-      "loss": 0.8037,
+      "learning_rate": 0.00027797687861271674,
+      "loss": 0.9536,
       "step": 629
     },
     {
       "epoch": 2.83,
-      "learning_rate": 0.00023853658536585366,
-      "loss": 0.7994,
+      "learning_rate": 0.0002778034682080925,
+      "loss": 0.9573,
       "step": 630
     },
     {
       "epoch": 2.83,
-      "learning_rate": 0.00023804878048780485,
-      "loss": 0.8201,
+      "learning_rate": 0.0002776300578034682,
+      "loss": 0.9805,
       "step": 631
     },
     {
       "epoch": 2.83,
-      "learning_rate": 0.00023756097560975606,
-      "loss": 0.7229,
+      "learning_rate": 0.0002774566473988439,
+      "loss": 0.8767,
       "step": 632
     },
     {
       "epoch": 2.84,
-      "learning_rate": 0.0002370731707317073,
-      "loss": 0.7606,
+      "learning_rate": 0.0002772832369942196,
+      "loss": 0.8657,
       "step": 633
     },
     {
       "epoch": 2.84,
-      "learning_rate": 0.0002365853658536585,
-      "loss": 0.7244,
+      "learning_rate": 0.0002771098265895954,
+      "loss": 0.8442,
       "step": 634
     },
     {
       "epoch": 2.85,
-      "learning_rate": 0.00023609756097560974,
-      "loss": 0.7032,
+      "learning_rate": 0.00027693641618497107,
+      "loss": 0.8549,
       "step": 635
     },
     {
       "epoch": 2.85,
-      "learning_rate": 0.00023560975609756096,
-      "loss": 0.6894,
+      "learning_rate": 0.0002767630057803468,
+      "loss": 0.8211,
       "step": 636
     },
     {
       "epoch": 2.86,
-      "learning_rate": 0.00023512195121951215,
-      "loss": 0.7164,
+      "learning_rate": 0.0002765895953757225,
+      "loss": 0.8208,
       "step": 637
     },
     {
       "epoch": 2.86,
-      "learning_rate": 0.0002346341463414634,
-      "loss": 0.6915,
+      "learning_rate": 0.00027641618497109826,
+      "loss": 0.8521,
       "step": 638
     },
     {
       "epoch": 2.87,
-      "learning_rate": 0.0002341463414634146,
-      "loss": 0.7157,
+      "learning_rate": 0.00027624277456647395,
+      "loss": 0.8427,
       "step": 639
     },
     {
       "epoch": 2.87,
-      "learning_rate": 0.00023365853658536583,
-      "loss": 0.7308,
+      "learning_rate": 0.0002760693641618497,
+      "loss": 0.8618,
       "step": 640
     },
     {
       "epoch": 2.87,
-      "learning_rate": 0.00023317073170731704,
-      "loss": 0.6428,
+      "learning_rate": 0.0002758959537572254,
+      "loss": 0.7377,
       "step": 641
     },
     {
       "epoch": 2.88,
-      "learning_rate": 0.0002326829268292683,
-      "loss": 0.6652,
+      "learning_rate": 0.00027572254335260114,
+      "loss": 0.7555,
       "step": 642
     },
     {
       "epoch": 2.88,
-      "learning_rate": 0.00023219512195121948,
-      "loss": 0.6669,
+      "learning_rate": 0.00027554913294797683,
+      "loss": 0.7199,
       "step": 643
     },
     {
       "epoch": 2.89,
-      "learning_rate": 0.0002317073170731707,
-      "loss": 0.6834,
+      "learning_rate": 0.0002753757225433526,
+      "loss": 0.8234,
       "step": 644
     },
     {
       "epoch": 2.89,
-      "learning_rate": 0.00023121951219512194,
-      "loss": 0.703,
+      "learning_rate": 0.0002752023121387283,
+      "loss": 0.8251,
       "step": 645
     },
     {
       "epoch": 2.9,
-      "learning_rate": 0.00023073170731707315,
-      "loss": 0.792,
+      "learning_rate": 0.000275028901734104,
+      "loss": 0.8766,
       "step": 646
     },
     {
       "epoch": 2.9,
-      "learning_rate": 0.00023024390243902437,
-      "loss": 7.3397,
+      "learning_rate": 0.00027485549132947977,
+      "loss": 9.3322,
       "step": 647
     },
     {
       "epoch": 2.91,
-      "learning_rate": 0.0002297560975609756,
-      "loss": 6.4753,
+      "learning_rate": 0.00027468208092485546,
+      "loss": 7.5506,
       "step": 648
     },
     {
       "epoch": 2.91,
-      "learning_rate": 0.00022926829268292683,
-      "loss": 4.5254,
+      "learning_rate": 0.00027450867052023116,
+      "loss": 5.1447,
       "step": 649
     },
     {
       "epoch": 2.91,
-      "learning_rate": 0.00022878048780487802,
-      "loss": 3.1011,
+      "learning_rate": 0.0002743352601156069,
+      "loss": 3.5468,
       "step": 650
     },
     {
       "epoch": 2.92,
-      "learning_rate": 0.00022829268292682924,
-      "loss": 2.1706,
+      "learning_rate": 0.00027416184971098265,
+      "loss": 2.5995,
       "step": 651
     },
     {
       "epoch": 2.92,
-      "learning_rate": 0.00022780487804878048,
-      "loss": 1.8403,
+      "learning_rate": 0.00027398843930635835,
+      "loss": 2.1182,
       "step": 652
     },
     {
       "epoch": 2.93,
-      "learning_rate": 0.00022731707317073167,
-      "loss": 1.7287,
+      "learning_rate": 0.0002738150289017341,
+      "loss": 1.9193,
       "step": 653
     },
     {
       "epoch": 2.93,
-      "learning_rate": 0.00022682926829268292,
-      "loss": 1.5158,
+      "learning_rate": 0.00027364161849710984,
+      "loss": 1.7254,
       "step": 654
     },
     {
       "epoch": 2.94,
-      "learning_rate": 0.00022634146341463413,
-      "loss": 1.2994,
+      "learning_rate": 0.00027346820809248554,
+      "loss": 1.513,
       "step": 655
     },
     {
       "epoch": 2.94,
-      "learning_rate": 0.00022585365853658532,
-      "loss": 1.099,
+      "learning_rate": 0.00027329479768786123,
+      "loss": 1.3363,
       "step": 656
     },
     {
       "epoch": 2.95,
-      "learning_rate": 0.00022536585365853657,
-      "loss": 0.9544,
+      "learning_rate": 0.000273121387283237,
+      "loss": 1.1537,
       "step": 657
     },
     {
       "epoch": 2.95,
-      "learning_rate": 0.00022487804878048778,
-      "loss": 0.97,
+      "learning_rate": 0.0002729479768786127,
+      "loss": 1.1289,
       "step": 658
     },
     {
       "epoch": 2.96,
-      "learning_rate": 0.000224390243902439,
-      "loss": 0.9019,
+      "learning_rate": 0.0002727745664739884,
+      "loss": 1.0504,
       "step": 659
     },
     {
       "epoch": 2.96,
-      "learning_rate": 0.00022390243902439022,
-      "loss": 0.873,
+      "learning_rate": 0.0002726011560693641,
+      "loss": 1.0136,
       "step": 660
     },
     {
       "epoch": 2.96,
-      "learning_rate": 0.00022341463414634146,
-      "loss": 0.8081,
+      "learning_rate": 0.00027242774566473986,
+      "loss": 0.9461,
       "step": 661
     },
     {
       "epoch": 2.97,
-      "learning_rate": 0.00022292682926829265,
-      "loss": 0.8081,
+      "learning_rate": 0.0002722543352601156,
+      "loss": 0.9157,
       "step": 662
     },
     {
       "epoch": 2.97,
-      "learning_rate": 0.00022243902439024387,
-      "loss": 0.697,
+      "learning_rate": 0.0002720809248554913,
+      "loss": 0.7936,
       "step": 663
     },
     {
       "epoch": 2.98,
-      "learning_rate": 0.0002219512195121951,
-      "loss": 0.663,
+      "learning_rate": 0.00027190751445086705,
+      "loss": 0.827,
       "step": 664
     },
     {
       "epoch": 2.98,
-      "learning_rate": 0.0002214634146341463,
-      "loss": 0.6758,
+      "learning_rate": 0.00027173410404624274,
+      "loss": 0.7706,
       "step": 665
     },
     {
       "epoch": 2.99,
-      "learning_rate": 0.00022097560975609755,
-      "loss": 0.6657,
+      "learning_rate": 0.0002715606936416185,
+      "loss": 0.7488,
       "step": 666
     },
     {
       "epoch": 2.99,
-      "learning_rate": 0.00022048780487804876,
-      "loss": 0.5622,
+      "learning_rate": 0.0002713872832369942,
+      "loss": 0.6611,
       "step": 667
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00021999999999999995,
-      "loss": 0.607,
+      "learning_rate": 0.00027121387283236993,
+      "loss": 0.6923,
       "step": 668
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.0002195121951219512,
-      "loss": 0.6879,
+      "learning_rate": 0.0002710404624277456,
+      "loss": 0.7828,
       "step": 669
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00021902439024390241,
-      "loss": 3.5432,
+      "learning_rate": 0.0002708670520231214,
+      "loss": 3.6436,
       "step": 670
     },
     {
       "epoch": 3.01,
-      "learning_rate": 0.00021853658536585366,
-      "loss": 3.2184,
+      "learning_rate": 0.00027069364161849707,
+      "loss": 3.0477,
       "step": 671
     },
     {
       "epoch": 3.01,
-      "learning_rate": 0.00021804878048780485,
-      "loss": 1.9385,
+      "learning_rate": 0.0002705202312138728,
+      "loss": 1.9186,
       "step": 672
     },
     {
       "epoch": 3.02,
-      "learning_rate": 0.0002175609756097561,
-      "loss": 1.2234,
+      "learning_rate": 0.0002703468208092485,
+      "loss": 1.7018,
       "step": 673
     },
     {
       "epoch": 3.02,
-      "learning_rate": 0.0002170731707317073,
-      "loss": 1.2986,
+      "learning_rate": 0.00027017341040462426,
+      "loss": 1.4485,
       "step": 674
     },
     {
       "epoch": 3.03,
-      "learning_rate": 0.0002165853658536585,
-      "loss": 1.1385,
+      "learning_rate": 0.00027,
+      "loss": 1.1206,
       "step": 675
     },
     {
       "epoch": 3.03,
-      "learning_rate": 0.00021609756097560974,
-      "loss": 1.0251,
+      "learning_rate": 0.0002698265895953757,
+      "loss": 1.1319,
       "step": 676
     },
     {
       "epoch": 3.04,
-      "learning_rate": 0.00021560975609756096,
-      "loss": 0.8948,
+      "learning_rate": 0.0002696531791907514,
+      "loss": 1.0149,
       "step": 677
     },
     {
       "epoch": 3.04,
-      "learning_rate": 0.00021512195121951218,
-      "loss": 0.6973,
+      "learning_rate": 0.00026947976878612714,
+      "loss": 0.809,
       "step": 678
     },
     {
       "epoch": 3.04,
-      "learning_rate": 0.0002146341463414634,
-      "loss": 0.7551,
+      "learning_rate": 0.0002693063583815029,
+      "loss": 0.8025,
       "step": 679
     },
     {
       "epoch": 3.05,
-      "learning_rate": 0.00021414634146341464,
-      "loss": 0.6294,
+      "learning_rate": 0.0002691329479768786,
+      "loss": 0.8262,
       "step": 680
     },
     {
       "epoch": 3.05,
-      "learning_rate": 0.00021365853658536583,
-      "loss": 0.6347,
+      "learning_rate": 0.00026895953757225433,
+      "loss": 0.6764,
       "step": 681
     },
     {
       "epoch": 3.06,
-      "learning_rate": 0.00021317073170731704,
-      "loss": 0.6635,
+      "learning_rate": 0.0002687861271676301,
+      "loss": 0.7046,
       "step": 682
     },
     {
       "epoch": 3.06,
-      "learning_rate": 0.0002126829268292683,
-      "loss": 0.5707,
+      "learning_rate": 0.00026861271676300577,
+      "loss": 0.6042,
       "step": 683
     },
     {
       "epoch": 3.07,
-      "learning_rate": 0.00021219512195121948,
-      "loss": 0.5126,
+      "learning_rate": 0.00026843930635838146,
+      "loss": 0.553,
       "step": 684
     },
     {
       "epoch": 3.07,
-      "learning_rate": 0.00021170731707317072,
-      "loss": 0.6055,
+      "learning_rate": 0.0002682658959537572,
+      "loss": 0.6246,
       "step": 685
     },
     {
       "epoch": 3.08,
-      "learning_rate": 0.00021121951219512194,
-      "loss": 0.5535,
+      "learning_rate": 0.00026809248554913296,
+      "loss": 0.6005,
       "step": 686
     },
     {
       "epoch": 3.08,
-      "learning_rate": 0.00021073170731707313,
-      "loss": 0.5278,
+      "learning_rate": 0.00026791907514450865,
+      "loss": 0.5905,
       "step": 687
     },
     {
       "epoch": 3.09,
-      "learning_rate": 0.00021024390243902437,
-      "loss": 0.5102,
+      "learning_rate": 0.00026774566473988435,
+      "loss": 0.5917,
       "step": 688
     },
     {
       "epoch": 3.09,
-      "learning_rate": 0.0002097560975609756,
-      "loss": 0.5113,
+      "learning_rate": 0.0002675722543352601,
+      "loss": 0.5212,
       "step": 689
     },
     {
       "epoch": 3.09,
-      "learning_rate": 0.0002092682926829268,
-      "loss": 0.4592,
+      "learning_rate": 0.00026739884393063584,
+      "loss": 0.5089,
       "step": 690
     },
     {
       "epoch": 3.1,
-      "learning_rate": 0.00020878048780487802,
-      "loss": 0.4637,
+      "learning_rate": 0.00026722543352601153,
+      "loss": 0.4883,
       "step": 691
     },
     {
       "epoch": 3.1,
-      "learning_rate": 0.00020829268292682927,
-      "loss": 0.4902,
+      "learning_rate": 0.0002670520231213873,
+      "loss": 0.5311,
       "step": 692
     },
     {
       "epoch": 3.11,
-      "learning_rate": 0.00020780487804878046,
-      "loss": 0.4212,
+      "learning_rate": 0.000266878612716763,
+      "loss": 0.4886,
       "step": 693
     },
     {
       "epoch": 3.11,
-      "learning_rate": 0.00020731707317073167,
-      "loss": 0.4117,
+      "learning_rate": 0.0002667052023121387,
+      "loss": 0.4564,
       "step": 694
     },
     {
       "epoch": 3.12,
-      "learning_rate": 0.00020682926829268292,
-      "loss": 0.4027,
+      "learning_rate": 0.0002665317919075144,
+      "loss": 0.4356,
       "step": 695
     },
     {
       "epoch": 3.12,
-      "learning_rate": 0.0002063414634146341,
-      "loss": 0.4618,
+      "learning_rate": 0.00026635838150289016,
+      "loss": 0.5206,
       "step": 696
     },
     {
       "epoch": 3.13,
-      "learning_rate": 0.00020585365853658535,
-      "loss": 0.3627,
+      "learning_rate": 0.00026618497109826586,
+      "loss": 0.409,
       "step": 697
     },
     {
       "epoch": 3.13,
-      "learning_rate": 0.00020536585365853657,
-      "loss": 0.4113,
+      "learning_rate": 0.00026601156069364155,
+      "loss": 0.431,
       "step": 698
     },
     {
       "epoch": 3.13,
-      "learning_rate": 0.0002048780487804878,
-      "loss": 0.3982,
+      "learning_rate": 0.0002658381502890173,
+      "loss": 0.448,
       "step": 699
     },
     {
       "epoch": 3.14,
-      "learning_rate": 0.000204390243902439,
-      "loss": 0.3762,
+      "learning_rate": 0.00026566473988439305,
+      "loss": 0.4387,
       "step": 700
     },
     {
       "epoch": 3.14,
-      "learning_rate": 0.00020390243902439022,
-      "loss": 0.3667,
+      "learning_rate": 0.00026549132947976874,
+      "loss": 0.3993,
       "step": 701
     },
     {
       "epoch": 3.15,
-      "learning_rate": 0.00020341463414634146,
-      "loss": 0.3957,
+      "learning_rate": 0.0002653179190751445,
+      "loss": 0.4169,
       "step": 702
     },
     {
       "epoch": 3.15,
-      "learning_rate": 0.00020292682926829265,
-      "loss": 0.3483,
+      "learning_rate": 0.00026514450867052024,
+      "loss": 0.4112,
       "step": 703
     },
     {
       "epoch": 3.16,
-      "learning_rate": 0.0002024390243902439,
-      "loss": 0.3318,
+      "learning_rate": 0.00026497109826589593,
+      "loss": 0.419,
       "step": 704
     },
     {
       "epoch": 3.16,
-      "learning_rate": 0.00020195121951219511,
-      "loss": 0.4213,
+      "learning_rate": 0.0002647976878612716,
+      "loss": 0.4739,
       "step": 705
     },
     {
       "epoch": 3.17,
-      "learning_rate": 0.0002014634146341463,
-      "loss": 0.368,
+      "learning_rate": 0.00026462427745664737,
+      "loss": 0.4217,
       "step": 706
     },
     {
       "epoch": 3.17,
-      "learning_rate": 0.00020097560975609755,
-      "loss": 0.3552,
+      "learning_rate": 0.0002644508670520231,
+      "loss": 0.4082,
       "step": 707
     },
     {
       "epoch": 3.17,
-      "learning_rate": 0.00020048780487804877,
-      "loss": 0.3594,
+      "learning_rate": 0.0002642774566473988,
+      "loss": 0.4224,
       "step": 708
     },
     {
       "epoch": 3.18,
-      "learning_rate": 0.00019999999999999998,
-      "loss": 0.3226,
+      "learning_rate": 0.00026410404624277456,
+      "loss": 0.3968,
       "step": 709
     },
     {
       "epoch": 3.18,
-      "learning_rate": 0.0001995121951219512,
-      "loss": 0.3709,
+      "learning_rate": 0.00026393063583815025,
+      "loss": 0.4368,
       "step": 710
     },
     {
       "epoch": 3.19,
-      "learning_rate": 0.00019902439024390244,
-      "loss": 0.3987,
+      "learning_rate": 0.000263757225433526,
+      "loss": 0.4774,
       "step": 711
     },
     {
       "epoch": 3.19,
-      "learning_rate": 0.00019853658536585363,
-      "loss": 0.3575,
+      "learning_rate": 0.0002635838150289017,
+      "loss": 0.3729,
       "step": 712
     },
     {
       "epoch": 3.2,
-      "learning_rate": 0.00019804878048780485,
-      "loss": 0.3317,
+      "learning_rate": 0.00026341040462427744,
+      "loss": 0.3777,
       "step": 713
     },
     {
       "epoch": 3.2,
-      "learning_rate": 0.0001975609756097561,
-      "loss": 0.2524,
+      "learning_rate": 0.00026323699421965314,
+      "loss": 0.3002,
       "step": 714
     },
     {
       "epoch": 3.21,
-      "learning_rate": 0.00019707317073170728,
-      "loss": 0.3431,
+      "learning_rate": 0.0002630635838150289,
+      "loss": 0.3403,
       "step": 715
     },
     {
       "epoch": 3.21,
-      "learning_rate": 0.00019658536585365853,
-      "loss": 0.3492,
+      "learning_rate": 0.0002628901734104046,
+      "loss": 0.3768,
       "step": 716
     },
     {
       "epoch": 3.22,
-      "learning_rate": 0.00019609756097560974,
-      "loss": 0.2618,
+      "learning_rate": 0.0002627167630057803,
+      "loss": 0.3403,
       "step": 717
     },
     {
       "epoch": 3.22,
-      "learning_rate": 0.00019560975609756093,
-      "loss": 0.3719,
+      "learning_rate": 0.000262543352601156,
+      "loss": 0.4646,
       "step": 718
     },
     {
       "epoch": 3.22,
-      "learning_rate": 0.00019512195121951218,
-      "loss": 0.3291,
+      "learning_rate": 0.00026236994219653177,
+      "loss": 0.3688,
       "step": 719
     },
     {
       "epoch": 3.23,
-      "learning_rate": 0.0001946341463414634,
-      "loss": 3.3472,
+      "learning_rate": 0.0002621965317919075,
+      "loss": 3.746,
       "step": 720
     },
     {
       "epoch": 3.23,
-      "learning_rate": 0.0001941463414634146,
-      "loss": 2.2989,
+      "learning_rate": 0.0002620231213872832,
+      "loss": 2.5798,
       "step": 721
     },
     {
       "epoch": 3.24,
-      "learning_rate": 0.00019365853658536583,
-      "loss": 1.2984,
+      "learning_rate": 0.0002618497109826589,
+      "loss": 1.5031,
       "step": 722
     },
     {
       "epoch": 3.24,
-      "learning_rate": 0.00019317073170731707,
-      "loss": 0.9822,
+      "learning_rate": 0.00026167630057803465,
+      "loss": 1.2022,
       "step": 723
     },
     {
       "epoch": 3.25,
-      "learning_rate": 0.0001926829268292683,
-      "loss": 0.9058,
+      "learning_rate": 0.0002615028901734104,
+      "loss": 1.0921,
       "step": 724
     },
     {
       "epoch": 3.25,
-      "learning_rate": 0.00019219512195121948,
-      "loss": 0.8659,
+      "learning_rate": 0.0002613294797687861,
+      "loss": 1.013,
       "step": 725
     },
     {
       "epoch": 3.26,
-      "learning_rate": 0.00019170731707317072,
-      "loss": 0.7585,
+      "learning_rate": 0.00026115606936416184,
+      "loss": 0.9056,
       "step": 726
     },
     {
       "epoch": 3.26,
-      "learning_rate": 0.00019121951219512194,
-      "loss": 0.7159,
+      "learning_rate": 0.0002609826589595376,
+      "loss": 0.8245,
       "step": 727
     },
     {
       "epoch": 3.26,
-      "learning_rate": 0.00019073170731707316,
-      "loss": 0.643,
+      "learning_rate": 0.0002608092485549133,
+      "loss": 0.7315,
       "step": 728
     },
     {
       "epoch": 3.27,
-      "learning_rate": 0.00019024390243902437,
-      "loss": 0.5668,
+      "learning_rate": 0.000260635838150289,
+      "loss": 0.6466,
       "step": 729
     },
     {
       "epoch": 3.27,
-      "learning_rate": 0.00018975609756097562,
-      "loss": 0.5185,
+      "learning_rate": 0.0002604624277456647,
+      "loss": 0.5553,
       "step": 730
     },
     {
       "epoch": 3.28,
-      "learning_rate": 0.0001892682926829268,
-      "loss": 0.5032,
+      "learning_rate": 0.00026028901734104047,
+      "loss": 0.5448,
       "step": 731
     },
     {
       "epoch": 3.28,
-      "learning_rate": 0.00018878048780487803,
-      "loss": 0.478,
+      "learning_rate": 0.00026011560693641616,
+      "loss": 0.5247,
       "step": 732
     },
     {
       "epoch": 3.29,
-      "learning_rate": 0.00018829268292682927,
-      "loss": 0.4333,
+      "learning_rate": 0.00025994219653179186,
+      "loss": 0.5081,
       "step": 733
     },
     {
       "epoch": 3.29,
-      "learning_rate": 0.00018780487804878046,
-      "loss": 0.4642,
+      "learning_rate": 0.0002597687861271676,
+      "loss": 0.4833,
       "step": 734
     },
     {
       "epoch": 3.3,
-      "learning_rate": 0.0001873170731707317,
-      "loss": 0.4117,
+      "learning_rate": 0.00025959537572254335,
+      "loss": 0.4507,
       "step": 735
     },
     {
       "epoch": 3.3,
-      "learning_rate": 0.00018682926829268292,
-      "loss": 0.4252,
+      "learning_rate": 0.00025942196531791905,
+      "loss": 0.5001,
       "step": 736
     },
     {
       "epoch": 3.3,
-      "learning_rate": 0.0001863414634146341,
-      "loss": 0.3947,
+      "learning_rate": 0.0002592485549132948,
+      "loss": 0.4418,
       "step": 737
     },
     {
       "epoch": 3.31,
-      "learning_rate": 0.00018585365853658535,
-      "loss": 0.4007,
+      "learning_rate": 0.0002590751445086705,
+      "loss": 0.418,
       "step": 738
     },
     {
       "epoch": 3.31,
-      "learning_rate": 0.00018536585365853657,
-      "loss": 0.4026,
+      "learning_rate": 0.00025890173410404624,
+      "loss": 0.4375,
       "step": 739
     },
     {
       "epoch": 3.32,
-      "learning_rate": 0.0001848780487804878,
-      "loss": 0.3837,
+      "learning_rate": 0.00025872832369942193,
+      "loss": 0.4326,
       "step": 740
     },
     {
       "epoch": 3.32,
-      "learning_rate": 0.000184390243902439,
-      "loss": 0.3415,
+      "learning_rate": 0.0002585549132947977,
+      "loss": 0.3794,
       "step": 741
     },
     {
       "epoch": 3.33,
-      "learning_rate": 0.00018390243902439025,
-      "loss": 0.4264,
+      "learning_rate": 0.00025838150289017337,
+      "loss": 0.4532,
       "step": 742
     },
     {
       "epoch": 3.33,
-      "learning_rate": 0.00018341463414634144,
-      "loss": 0.3209,
+      "learning_rate": 0.0002582080924855491,
+      "loss": 0.3637,
       "step": 743
     },
     {
       "epoch": 3.34,
-      "learning_rate": 0.00018292682926829266,
-      "loss": 0.3521,
+      "learning_rate": 0.0002580346820809248,
+      "loss": 0.3888,
       "step": 744
     },
     {
       "epoch": 3.34,
-      "learning_rate": 0.0001824390243902439,
-      "loss": 0.3439,
+      "learning_rate": 0.00025786127167630056,
+      "loss": 0.3813,
       "step": 745
     },
     {
       "epoch": 3.35,
-      "learning_rate": 0.0001819512195121951,
-      "loss": 0.3283,
+      "learning_rate": 0.00025768786127167625,
+      "loss": 0.3921,
       "step": 746
     },
     {
       "epoch": 3.35,
-      "learning_rate": 0.00018146341463414633,
-      "loss": 0.3709,
+      "learning_rate": 0.000257514450867052,
+      "loss": 0.387,
       "step": 747
     },
     {
       "epoch": 3.35,
-      "learning_rate": 0.00018097560975609755,
-      "loss": 0.3629,
+      "learning_rate": 0.00025734104046242775,
+      "loss": 0.4,
       "step": 748
     },
     {
       "epoch": 3.36,
-      "learning_rate": 0.00018048780487804877,
-      "loss": 0.3261,
+      "learning_rate": 0.00025716763005780344,
+      "loss": 0.3746,
       "step": 749
     },
     {
       "epoch": 3.36,
-      "learning_rate": 0.00017999999999999998,
-      "loss": 0.3054,
+      "learning_rate": 0.00025699421965317914,
+      "loss": 0.3611,
       "step": 750
     },
     {
       "epoch": 3.37,
-      "learning_rate": 0.0001795121951219512,
-      "loss": 0.3341,
+      "learning_rate": 0.0002568208092485549,
+      "loss": 0.3834,
       "step": 751
     },
     {
       "epoch": 3.37,
-      "learning_rate": 0.00017902439024390245,
-      "loss": 0.3435,
+      "learning_rate": 0.00025664739884393063,
+      "loss": 0.3827,
       "step": 752
     },
     {
       "epoch": 3.38,
-      "learning_rate": 0.00017853658536585363,
-      "loss": 0.3293,
+      "learning_rate": 0.0002564739884393063,
+      "loss": 0.4217,
       "step": 753
     },
     {
       "epoch": 3.38,
-      "learning_rate": 0.00017804878048780485,
-      "loss": 0.3783,
+      "learning_rate": 0.00025630057803468207,
+      "loss": 0.4256,
       "step": 754
     },
     {
       "epoch": 3.39,
-      "learning_rate": 0.0001775609756097561,
-      "loss": 0.3352,
+      "learning_rate": 0.0002561271676300578,
+      "loss": 0.3919,
       "step": 755
     },
     {
       "epoch": 3.39,
-      "learning_rate": 0.00017707317073170729,
-      "loss": 0.275,
+      "learning_rate": 0.0002559537572254335,
+      "loss": 0.329,
       "step": 756
     },
     {
       "epoch": 3.39,
-      "learning_rate": 0.00017658536585365853,
-      "loss": 0.3101,
+      "learning_rate": 0.0002557803468208092,
+      "loss": 0.364,
       "step": 757
     },
     {
       "epoch": 3.4,
-      "learning_rate": 0.00017609756097560975,
-      "loss": 0.322,
+      "learning_rate": 0.00025560693641618496,
+      "loss": 0.3676,
       "step": 758
     },
     {
       "epoch": 3.4,
-      "learning_rate": 0.00017560975609756094,
-      "loss": 0.2583,
+      "learning_rate": 0.0002554335260115607,
+      "loss": 0.3192,
       "step": 759
     },
     {
       "epoch": 3.41,
-      "learning_rate": 0.00017512195121951218,
-      "loss": 0.2986,
+      "learning_rate": 0.0002552601156069364,
+      "loss": 0.386,
       "step": 760
     },
     {
       "epoch": 3.41,
-      "learning_rate": 0.0001746341463414634,
-      "loss": 0.2965,
+      "learning_rate": 0.0002550867052023121,
+      "loss": 0.3516,
       "step": 761
     },
     {
       "epoch": 3.42,
-      "learning_rate": 0.00017414634146341461,
-      "loss": 0.3305,
+      "learning_rate": 0.00025491329479768784,
+      "loss": 0.3816,
       "step": 762
     },
     {
       "epoch": 3.42,
-      "learning_rate": 0.00017365853658536583,
-      "loss": 0.2637,
+      "learning_rate": 0.0002547398843930636,
+      "loss": 0.3091,
       "step": 763
     },
     {
       "epoch": 3.43,
-      "learning_rate": 0.00017317073170731708,
-      "loss": 0.2815,
+      "learning_rate": 0.0002545664739884393,
+      "loss": 0.3428,
       "step": 764
     },
     {
       "epoch": 3.43,
-      "learning_rate": 0.00017268292682926826,
-      "loss": 0.3337,
+      "learning_rate": 0.00025439306358381503,
+      "loss": 0.3129,
       "step": 765
     },
     {
       "epoch": 3.43,
-      "learning_rate": 0.00017219512195121948,
-      "loss": 0.3071,
+      "learning_rate": 0.0002542196531791907,
+      "loss": 0.3559,
       "step": 766
     },
     {
       "epoch": 3.44,
-      "learning_rate": 0.00017170731707317073,
-      "loss": 0.2576,
+      "learning_rate": 0.00025404624277456647,
+      "loss": 0.3588,
       "step": 767
     },
     {
       "epoch": 3.44,
-      "learning_rate": 0.00017121951219512192,
-      "loss": 0.2627,
+      "learning_rate": 0.00025387283236994216,
+      "loss": 0.3518,
       "step": 768
     },
     {
       "epoch": 3.45,
-      "learning_rate": 0.00017073170731707316,
-      "loss": 0.2915,
+      "learning_rate": 0.0002536994219653179,
+      "loss": 0.3134,
       "step": 769
     },
     {
       "epoch": 3.45,
-      "learning_rate": 0.00017024390243902438,
-      "loss": 1.9187,
+      "learning_rate": 0.0002535260115606936,
+      "loss": 2.7246,
       "step": 770
     },
     {
       "epoch": 3.46,
-      "learning_rate": 0.00016975609756097557,
-      "loss": 1.4796,
+      "learning_rate": 0.00025335260115606935,
+      "loss": 2.0574,
       "step": 771
     },
     {
       "epoch": 3.46,
-      "learning_rate": 0.0001692682926829268,
-      "loss": 0.9913,
+      "learning_rate": 0.00025317919075144504,
+      "loss": 1.2423,
       "step": 772
     },
     {
       "epoch": 3.47,
-      "learning_rate": 0.00016878048780487803,
-      "loss": 0.6812,
+      "learning_rate": 0.0002530057803468208,
+      "loss": 0.8525,
       "step": 773
     },
     {
       "epoch": 3.47,
-      "learning_rate": 0.00016829268292682927,
-      "loss": 0.6404,
+      "learning_rate": 0.0002528323699421965,
+      "loss": 0.8248,
       "step": 774
     },
     {
       "epoch": 3.48,
-      "learning_rate": 0.00016780487804878046,
-      "loss": 0.6294,
+      "learning_rate": 0.00025265895953757223,
+      "loss": 0.6959,
       "step": 775
     },
     {
       "epoch": 3.48,
-      "learning_rate": 0.0001673170731707317,
-      "loss": 0.5524,
+      "learning_rate": 0.000252485549132948,
+      "loss": 0.6066,
       "step": 776
     },
     {
       "epoch": 3.48,
-      "learning_rate": 0.00016682926829268292,
-      "loss": 0.4834,
+      "learning_rate": 0.0002523121387283237,
+      "loss": 0.5454,
       "step": 777
     },
     {
       "epoch": 3.49,
-      "learning_rate": 0.0001663414634146341,
-      "loss": 0.4867,
+      "learning_rate": 0.00025213872832369937,
+      "loss": 0.5132,
       "step": 778
     },
     {
       "epoch": 3.49,
-      "learning_rate": 0.00016585365853658536,
-      "loss": 0.4394,
+      "learning_rate": 0.0002519653179190751,
+      "loss": 0.4923,
       "step": 779
     },
     {
       "epoch": 3.5,
-      "learning_rate": 0.00016536585365853657,
-      "loss": 0.4228,
+      "learning_rate": 0.00025179190751445086,
+      "loss": 0.4915,
       "step": 780
     },
     {
       "epoch": 3.5,
-      "learning_rate": 0.0001648780487804878,
-      "loss": 0.3963,
+      "learning_rate": 0.00025161849710982656,
+      "loss": 0.4586,
       "step": 781
     },
     {
       "epoch": 3.51,
-      "learning_rate": 0.000164390243902439,
-      "loss": 0.3693,
+      "learning_rate": 0.0002514450867052023,
+      "loss": 0.4383,
       "step": 782
     },
     {
       "epoch": 3.51,
-      "learning_rate": 0.00016390243902439025,
-      "loss": 0.3423,
+      "learning_rate": 0.00025127167630057805,
+      "loss": 0.3949,
       "step": 783
     },
     {
       "epoch": 3.52,
-      "learning_rate": 0.00016341463414634144,
-      "loss": 0.3681,
+      "learning_rate": 0.00025109826589595375,
+      "loss": 0.4295,
       "step": 784
     },
     {
       "epoch": 3.52,
-      "learning_rate": 0.00016292682926829266,
-      "loss": 0.3444,
+      "learning_rate": 0.00025092485549132944,
+      "loss": 0.3954,
       "step": 785
     },
     {
       "epoch": 3.52,
-      "learning_rate": 0.0001624390243902439,
-      "loss": 0.3625,
+      "learning_rate": 0.0002507514450867052,
+      "loss": 0.4317,
       "step": 786
     },
     {
       "epoch": 3.53,
-      "learning_rate": 0.0001619512195121951,
-      "loss": 0.335,
+      "learning_rate": 0.00025057803468208094,
+      "loss": 0.4151,
       "step": 787
     },
     {
       "epoch": 3.53,
-      "learning_rate": 0.00016146341463414634,
-      "loss": 0.3001,
+      "learning_rate": 0.00025040462427745663,
+      "loss": 0.349,
       "step": 788
     },
     {
       "epoch": 3.54,
-      "learning_rate": 0.00016097560975609755,
-      "loss": 0.3523,
+      "learning_rate": 0.0002502312138728323,
+      "loss": 0.3537,
       "step": 789
     },
     {
       "epoch": 3.54,
-      "learning_rate": 0.00016048780487804874,
-      "loss": 0.3084,
+      "learning_rate": 0.00025005780346820807,
+      "loss": 0.3305,
       "step": 790
     },
     {
       "epoch": 3.55,
-      "learning_rate": 0.00015999999999999999,
-      "loss": 0.3502,
+      "learning_rate": 0.00024988439306358376,
+      "loss": 0.3834,
       "step": 791
     },
     {
       "epoch": 3.55,
-      "learning_rate": 0.0001595121951219512,
-      "loss": 0.3246,
+      "learning_rate": 0.0002497109826589595,
+      "loss": 0.3788,
       "step": 792
     },
     {
       "epoch": 3.56,
-      "learning_rate": 0.00015902439024390242,
-      "loss": 0.3061,
+      "learning_rate": 0.00024953757225433526,
+      "loss": 0.3858,
       "step": 793
     },
     {
       "epoch": 3.56,
-      "learning_rate": 0.00015853658536585364,
-      "loss": 0.3581,
+      "learning_rate": 0.00024936416184971095,
+      "loss": 0.3739,
       "step": 794
     },
     {
       "epoch": 3.57,
-      "learning_rate": 0.00015804878048780488,
-      "loss": 0.3109,
+      "learning_rate": 0.00024919075144508665,
+      "loss": 0.3757,
       "step": 795
     },
     {
       "epoch": 3.57,
-      "learning_rate": 0.00015756097560975607,
-      "loss": 0.2814,
+      "learning_rate": 0.0002490173410404624,
+      "loss": 0.3295,
       "step": 796
     },
     {
       "epoch": 3.57,
-      "learning_rate": 0.0001570731707317073,
-      "loss": 0.2509,
+      "learning_rate": 0.00024884393063583814,
+      "loss": 0.2999,
       "step": 797
     },
     {
       "epoch": 3.58,
-      "learning_rate": 0.00015658536585365853,
-      "loss": 0.3046,
+      "learning_rate": 0.00024867052023121384,
+      "loss": 0.3492,
       "step": 798
     },
     {
       "epoch": 3.58,
-      "learning_rate": 0.00015609756097560975,
-      "loss": 0.2995,
+      "learning_rate": 0.0002484971098265896,
+      "loss": 0.3537,
       "step": 799
     },
     {
       "epoch": 3.59,
-      "learning_rate": 0.00015560975609756097,
-      "loss": 0.2679,
+      "learning_rate": 0.00024832369942196533,
+      "loss": 0.3539,
       "step": 800
     },
     {
       "epoch": 3.59,
-      "learning_rate": 0.00015512195121951218,
-      "loss": 0.3044,
+      "learning_rate": 0.000248150289017341,
+      "loss": 0.3874,
       "step": 801
     },
     {
       "epoch": 3.6,
-      "learning_rate": 0.00015463414634146343,
-      "loss": 0.286,
+      "learning_rate": 0.0002479768786127167,
+      "loss": 0.3529,
       "step": 802
     },
     {
       "epoch": 3.6,
-      "learning_rate": 0.00015414634146341462,
-      "loss": 0.2883,
+      "learning_rate": 0.00024780346820809247,
+      "loss": 0.3646,
       "step": 803
     },
     {
       "epoch": 3.61,
-      "learning_rate": 0.00015365853658536583,
-      "loss": 0.3195,
+      "learning_rate": 0.0002476300578034682,
+      "loss": 0.3855,
       "step": 804
     },
     {
       "epoch": 3.61,
-      "learning_rate": 0.00015317073170731708,
-      "loss": 0.284,
+      "learning_rate": 0.0002474566473988439,
+      "loss": 0.3391,
       "step": 805
     },
     {
       "epoch": 3.61,
-      "learning_rate": 0.00015268292682926827,
-      "loss": 0.2843,
+      "learning_rate": 0.0002472832369942196,
+      "loss": 0.3225,
       "step": 806
     },
     {
       "epoch": 3.62,
-      "learning_rate": 0.0001521951219512195,
-      "loss": 0.2142,
+      "learning_rate": 0.00024710982658959535,
+      "loss": 0.2516,
       "step": 807
     },
     {
       "epoch": 3.62,
-      "learning_rate": 0.00015170731707317073,
-      "loss": 0.2466,
+      "learning_rate": 0.0002469364161849711,
+      "loss": 0.3274,
       "step": 808
     },
     {
       "epoch": 3.63,
-      "learning_rate": 0.00015121951219512192,
-      "loss": 0.2987,
+      "learning_rate": 0.0002467630057803468,
+      "loss": 0.3606,
       "step": 809
     },
     {
       "epoch": 3.63,
-      "learning_rate": 0.00015073170731707316,
-      "loss": 0.2849,
+      "learning_rate": 0.00024658959537572254,
+      "loss": 0.3471,
       "step": 810
     },
     {
       "epoch": 3.64,
-      "learning_rate": 0.00015024390243902438,
-      "loss": 0.2537,
+      "learning_rate": 0.00024641618497109823,
+      "loss": 0.329,
       "step": 811
     },
     {
       "epoch": 3.64,
-      "learning_rate": 0.0001497560975609756,
-      "loss": 0.2193,
+      "learning_rate": 0.000246242774566474,
+      "loss": 0.2733,
       "step": 812
     },
     {
       "epoch": 3.65,
-      "learning_rate": 0.0001492682926829268,
-      "loss": 0.2516,
+      "learning_rate": 0.0002460693641618497,
+      "loss": 0.3286,
       "step": 813
     },
     {
       "epoch": 3.65,
-      "learning_rate": 0.00014878048780487803,
-      "loss": 0.2595,
+      "learning_rate": 0.0002458959537572254,
+      "loss": 0.2962,
       "step": 814
     },
     {
       "epoch": 3.65,
-      "learning_rate": 0.00014829268292682925,
-      "loss": 0.2648,
+      "learning_rate": 0.0002457225433526011,
+      "loss": 0.2743,
       "step": 815
     },
     {
       "epoch": 3.66,
-      "learning_rate": 0.00014780487804878046,
-      "loss": 0.2563,
+      "learning_rate": 0.00024554913294797686,
+      "loss": 0.3511,
       "step": 816
     },
     {
       "epoch": 3.66,
-      "learning_rate": 0.00014731707317073168,
-      "loss": 0.2177,
+      "learning_rate": 0.00024537572254335256,
+      "loss": 0.2911,
       "step": 817
     },
     {
       "epoch": 3.67,
-      "learning_rate": 0.00014682926829268292,
-      "loss": 0.2533,
+      "learning_rate": 0.0002452023121387283,
+      "loss": 0.2989,
       "step": 818
     },
     {
       "epoch": 3.67,
-      "learning_rate": 0.00014634146341463414,
-      "loss": 0.292,
+      "learning_rate": 0.000245028901734104,
+      "loss": 0.3634,
       "step": 819
     },
     {
       "epoch": 3.68,
-      "learning_rate": 0.00014585365853658536,
-      "loss": 1.3898,
+      "learning_rate": 0.00024485549132947975,
+      "loss": 1.8232,
       "step": 820
     },
     {
       "epoch": 3.68,
-      "learning_rate": 0.00014536585365853657,
-      "loss": 1.0223,
+      "learning_rate": 0.0002446820809248555,
+      "loss": 1.0919,
       "step": 821
     },
     {
       "epoch": 3.69,
-      "learning_rate": 0.0001448780487804878,
-      "loss": 0.6619,
+      "learning_rate": 0.0002445086705202312,
+      "loss": 0.7213,
       "step": 822
     },
     {
       "epoch": 3.69,
-      "learning_rate": 0.000144390243902439,
-      "loss": 0.5068,
+      "learning_rate": 0.0002443352601156069,
+      "loss": 0.7192,
       "step": 823
     },
     {
       "epoch": 3.7,
-      "learning_rate": 0.00014390243902439023,
-      "loss": 0.473,
+      "learning_rate": 0.00024416184971098263,
+      "loss": 0.7089,
       "step": 824
     },
     {
       "epoch": 3.7,
-      "learning_rate": 0.00014341463414634144,
-      "loss": 0.4847,
+      "learning_rate": 0.00024398843930635838,
+      "loss": 0.5815,
       "step": 825
     },
     {
       "epoch": 3.7,
-      "learning_rate": 0.00014292682926829269,
-      "loss": 0.4595,
+      "learning_rate": 0.00024381502890173407,
+      "loss": 0.5092,
       "step": 826
     },
     {
       "epoch": 3.71,
-      "learning_rate": 0.00014243902439024388,
-      "loss": 0.3926,
+      "learning_rate": 0.0002436416184971098,
+      "loss": 0.5179,
       "step": 827
     },
     {
       "epoch": 3.71,
-      "learning_rate": 0.0001419512195121951,
-      "loss": 0.3979,
+      "learning_rate": 0.00024346820809248554,
+      "loss": 0.4872,
       "step": 828
     },
     {
       "epoch": 3.72,
-      "learning_rate": 0.00014146341463414634,
-      "loss": 0.3735,
+      "learning_rate": 0.00024329479768786126,
+      "loss": 0.4405,
       "step": 829
     },
     {
       "epoch": 3.72,
-      "learning_rate": 0.00014097560975609755,
-      "loss": 0.3384,
+      "learning_rate": 0.00024312138728323698,
+      "loss": 0.4115,
       "step": 830
     },
     {
       "epoch": 3.73,
-      "learning_rate": 0.00014048780487804877,
-      "loss": 0.3536,
+      "learning_rate": 0.00024294797687861267,
+      "loss": 0.4027,
       "step": 831
     },
     {
       "epoch": 3.73,
-      "learning_rate": 0.00014,
-      "loss": 0.3151,
+      "learning_rate": 0.00024277456647398842,
+      "loss": 0.3596,
       "step": 832
     },
     {
       "epoch": 3.74,
-      "learning_rate": 0.0001395121951219512,
-      "loss": 0.3147,
+      "learning_rate": 0.00024260115606936414,
+      "loss": 0.3759,
       "step": 833
     },
     {
       "epoch": 3.74,
-      "learning_rate": 0.00013902439024390242,
-      "loss": 0.3133,
+      "learning_rate": 0.00024242774566473986,
+      "loss": 0.3514,
       "step": 834
     },
     {
       "epoch": 3.74,
-      "learning_rate": 0.00013853658536585364,
-      "loss": 0.255,
+      "learning_rate": 0.00024225433526011558,
+      "loss": 0.3,
       "step": 835
     },
     {
       "epoch": 3.75,
-      "learning_rate": 0.00013804878048780486,
-      "loss": 0.2868,
+      "learning_rate": 0.00024208092485549133,
+      "loss": 0.3155,
       "step": 836
     },
     {
       "epoch": 3.75,
-      "learning_rate": 0.0001375609756097561,
-      "loss": 0.2783,
+      "learning_rate": 0.00024190751445086702,
+      "loss": 0.3328,
       "step": 837
     },
     {
       "epoch": 3.76,
-      "learning_rate": 0.00013707317073170732,
-      "loss": 0.2877,
+      "learning_rate": 0.00024173410404624275,
+      "loss": 0.3285,
       "step": 838
     },
     {
       "epoch": 3.76,
-      "learning_rate": 0.0001365853658536585,
-      "loss": 0.2775,
+      "learning_rate": 0.00024156069364161847,
+      "loss": 0.3116,
       "step": 839
     },
     {
       "epoch": 3.77,
-      "learning_rate": 0.00013609756097560975,
-      "loss": 0.2804,
+      "learning_rate": 0.0002413872832369942,
+      "loss": 0.4866,
       "step": 840
     },
     {
       "epoch": 3.77,
-      "learning_rate": 0.00013560975609756097,
-      "loss": 0.2744,
+      "learning_rate": 0.00024121387283236993,
+      "loss": 0.3259,
       "step": 841
     },
     {
       "epoch": 3.78,
-      "learning_rate": 0.00013512195121951218,
-      "loss": 0.2385,
+      "learning_rate": 0.00024104046242774563,
+      "loss": 0.3017,
       "step": 842
     },
     {
       "epoch": 3.78,
-      "learning_rate": 0.0001346341463414634,
-      "loss": 0.2474,
+      "learning_rate": 0.00024086705202312135,
+      "loss": 0.3129,
       "step": 843
     },
     {
       "epoch": 3.78,
-      "learning_rate": 0.00013414634146341462,
-      "loss": 0.2529,
+      "learning_rate": 0.0002406936416184971,
+      "loss": 0.328,
       "step": 844
     },
     {
       "epoch": 3.79,
-      "learning_rate": 0.00013365853658536586,
-      "loss": 0.2812,
+      "learning_rate": 0.00024052023121387282,
+      "loss": 0.3282,
       "step": 845
     },
     {
       "epoch": 3.79,
-      "learning_rate": 0.00013317073170731705,
-      "loss": 0.2556,
+      "learning_rate": 0.00024034682080924854,
+      "loss": 0.2992,
       "step": 846
     },
     {
       "epoch": 3.8,
-      "learning_rate": 0.00013268292682926827,
-      "loss": 0.2324,
+      "learning_rate": 0.00024017341040462423,
+      "loss": 0.2855,
       "step": 847
     },
     {
       "epoch": 3.8,
-      "learning_rate": 0.0001321951219512195,
-      "loss": 0.2562,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 0.3201,
       "step": 848
     },
     {
       "epoch": 3.81,
-      "learning_rate": 0.00013170731707317073,
-      "loss": 0.2345,
+      "learning_rate": 0.0002398265895953757,
+      "loss": 0.264,
       "step": 849
     },
     {
       "epoch": 3.81,
-      "learning_rate": 0.00013121951219512195,
-      "loss": 0.2582,
+      "learning_rate": 0.00023965317919075142,
+      "loss": 0.3239,
       "step": 850
     },
     {
       "epoch": 3.82,
-      "learning_rate": 0.00013073170731707316,
-      "loss": 0.2171,
+      "learning_rate": 0.00023947976878612714,
+      "loss": 0.2577,
       "step": 851
     },
     {
       "epoch": 3.82,
-      "learning_rate": 0.00013024390243902438,
-      "loss": 0.2607,
+      "learning_rate": 0.0002393063583815029,
+      "loss": 0.316,
       "step": 852
     },
     {
       "epoch": 3.83,
-      "learning_rate": 0.0001297560975609756,
-      "loss": 0.227,
+      "learning_rate": 0.0002391329479768786,
+      "loss": 0.2993,
       "step": 853
     },
     {
       "epoch": 3.83,
-      "learning_rate": 0.00012926829268292681,
-      "loss": 0.2222,
+      "learning_rate": 0.0002389595375722543,
+      "loss": 0.2835,
       "step": 854
     },
     {
       "epoch": 3.83,
-      "learning_rate": 0.00012878048780487803,
-      "loss": 0.2358,
+      "learning_rate": 0.00023878612716763002,
+      "loss": 0.3059,
       "step": 855
     },
     {
       "epoch": 3.84,
-      "learning_rate": 0.00012829268292682925,
-      "loss": 0.235,
+      "learning_rate": 0.00023861271676300577,
+      "loss": 0.2899,
       "step": 856
     },
     {
       "epoch": 3.84,
-      "learning_rate": 0.0001278048780487805,
-      "loss": 0.2213,
+      "learning_rate": 0.0002384393063583815,
+      "loss": 0.2627,
       "step": 857
     },
     {
       "epoch": 3.85,
-      "learning_rate": 0.00012731707317073168,
-      "loss": 0.2178,
+      "learning_rate": 0.0002382658959537572,
+      "loss": 0.2783,
       "step": 858
     },
     {
       "epoch": 3.85,
-      "learning_rate": 0.00012682926829268293,
-      "loss": 0.2105,
+      "learning_rate": 0.0002380924855491329,
+      "loss": 0.2801,
       "step": 859
     },
     {
       "epoch": 3.86,
-      "learning_rate": 0.00012634146341463414,
-      "loss": 0.2181,
+      "learning_rate": 0.00023791907514450865,
+      "loss": 0.2571,
       "step": 860
     },
     {
       "epoch": 3.86,
-      "learning_rate": 0.00012585365853658536,
-      "loss": 0.2648,
+      "learning_rate": 0.00023774566473988437,
+      "loss": 0.3506,
       "step": 861
     },
     {
       "epoch": 3.87,
-      "learning_rate": 0.00012536585365853658,
-      "loss": 0.2673,
+      "learning_rate": 0.0002375722543352601,
+      "loss": 0.3228,
       "step": 862
     },
     {
       "epoch": 3.87,
-      "learning_rate": 0.0001248780487804878,
-      "loss": 0.2179,
+      "learning_rate": 0.00023739884393063582,
+      "loss": 0.2565,
       "step": 863
     },
     {
       "epoch": 3.87,
-      "learning_rate": 0.000124390243902439,
-      "loss": 0.1762,
+      "learning_rate": 0.00023722543352601156,
+      "loss": 0.2523,
       "step": 864
     },
     {
       "epoch": 3.88,
-      "learning_rate": 0.00012390243902439023,
-      "loss": 0.2107,
+      "learning_rate": 0.00023705202312138726,
+      "loss": 0.2447,
       "step": 865
     },
     {
       "epoch": 3.88,
-      "learning_rate": 0.00012341463414634144,
-      "loss": 0.2415,
+      "learning_rate": 0.00023687861271676298,
+      "loss": 0.3134,
       "step": 866
     },
     {
       "epoch": 3.89,
-      "learning_rate": 0.00012292682926829266,
-      "loss": 0.265,
+      "learning_rate": 0.0002367052023121387,
+      "loss": 0.3882,
       "step": 867
     },
     {
       "epoch": 3.89,
-      "learning_rate": 0.0001224390243902439,
-      "loss": 0.2246,
+      "learning_rate": 0.00023653179190751445,
+      "loss": 0.3079,
       "step": 868
     },
     {
       "epoch": 3.9,
-      "learning_rate": 0.00012195121951219511,
-      "loss": 0.3635,
+      "learning_rate": 0.00023635838150289017,
+      "loss": 0.4326,
       "step": 869
     },
     {
       "epoch": 3.9,
-      "learning_rate": 0.00012146341463414634,
-      "loss": 0.7013,
+      "learning_rate": 0.00023618497109826586,
+      "loss": 1.5025,
       "step": 870
     },
     {
       "epoch": 3.91,
-      "learning_rate": 0.00012097560975609756,
-      "loss": 0.536,
+      "learning_rate": 0.00023601156069364158,
+      "loss": 0.9483,
       "step": 871
     },
     {
       "epoch": 3.91,
-      "learning_rate": 0.00012048780487804877,
-      "loss": 0.398,
+      "learning_rate": 0.00023583815028901733,
+      "loss": 0.5595,
       "step": 872
     },
     {
       "epoch": 3.91,
-      "learning_rate": 0.00011999999999999999,
-      "loss": 0.3576,
+      "learning_rate": 0.00023566473988439305,
+      "loss": 0.5277,
       "step": 873
     },
     {
       "epoch": 3.92,
-      "learning_rate": 0.0001195121951219512,
-      "loss": 0.3146,
+      "learning_rate": 0.00023549132947976877,
+      "loss": 0.4715,
       "step": 874
     },
     {
       "epoch": 3.92,
-      "learning_rate": 0.00011902439024390242,
-      "loss": 0.3146,
+      "learning_rate": 0.0002353179190751445,
+      "loss": 0.4018,
       "step": 875
     },
     {
       "epoch": 3.93,
-      "learning_rate": 0.00011853658536585365,
-      "loss": 0.3472,
+      "learning_rate": 0.00023514450867052024,
+      "loss": 0.4456,
       "step": 876
     },
     {
       "epoch": 3.93,
-      "learning_rate": 0.00011804878048780487,
-      "loss": 0.2634,
+      "learning_rate": 0.00023497109826589593,
+      "loss": 0.3791,
       "step": 877
     },
     {
       "epoch": 3.94,
-      "learning_rate": 0.00011756097560975607,
-      "loss": 0.3482,
+      "learning_rate": 0.00023479768786127165,
+      "loss": 0.4253,
       "step": 878
     },
     {
       "epoch": 3.94,
-      "learning_rate": 0.0001170731707317073,
-      "loss": 0.2643,
+      "learning_rate": 0.00023462427745664737,
+      "loss": 0.3124,
       "step": 879
     },
     {
       "epoch": 3.95,
-      "learning_rate": 0.00011658536585365852,
-      "loss": 0.2699,
+      "learning_rate": 0.00023445086705202312,
+      "loss": 0.3515,
       "step": 880
     },
     {
       "epoch": 3.95,
-      "learning_rate": 0.00011609756097560974,
-      "loss": 0.2764,
+      "learning_rate": 0.00023427745664739884,
+      "loss": 0.3239,
       "step": 881
     },
     {
       "epoch": 3.96,
-      "learning_rate": 0.00011560975609756097,
-      "loss": 0.2324,
+      "learning_rate": 0.00023410404624277454,
+      "loss": 0.2906,
       "step": 882
     },
     {
       "epoch": 3.96,
-      "learning_rate": 0.00011512195121951219,
-      "loss": 0.2683,
+      "learning_rate": 0.00023393063583815026,
+      "loss": 0.3742,
       "step": 883
     },
     {
       "epoch": 3.96,
-      "learning_rate": 0.00011463414634146342,
-      "loss": 0.2209,
+      "learning_rate": 0.000233757225433526,
+      "loss": 0.2932,
       "step": 884
     },
     {
       "epoch": 3.97,
-      "learning_rate": 0.00011414634146341462,
-      "loss": 0.2287,
+      "learning_rate": 0.00023358381502890173,
+      "loss": 0.3035,
       "step": 885
     },
     {
       "epoch": 3.97,
-      "learning_rate": 0.00011365853658536584,
-      "loss": 0.2358,
+      "learning_rate": 0.00023341040462427745,
+      "loss": 0.2989,
       "step": 886
     },
     {
       "epoch": 3.98,
-      "learning_rate": 0.00011317073170731707,
-      "loss": 0.2162,
+      "learning_rate": 0.00023323699421965314,
+      "loss": 0.2559,
       "step": 887
     },
     {
       "epoch": 3.98,
-      "learning_rate": 0.00011268292682926828,
-      "loss": 0.2329,
+      "learning_rate": 0.00023306358381502886,
+      "loss": 0.3164,
       "step": 888
     },
     {
       "epoch": 3.99,
-      "learning_rate": 0.0001121951219512195,
-      "loss": 0.2235,
+      "learning_rate": 0.0002328901734104046,
+      "loss": 0.2812,
       "step": 889
     },
     {
       "epoch": 3.99,
-      "learning_rate": 0.00011170731707317073,
-      "loss": 0.2409,
+      "learning_rate": 0.00023271676300578033,
+      "loss": 0.3021,
       "step": 890
     },
     {
       "epoch": 4.0,
-      "learning_rate": 0.00011121951219512193,
-      "loss": 0.1879,
+      "learning_rate": 0.00023254335260115605,
+      "loss": 0.233,
       "step": 891
     },
     {
       "epoch": 4.0,
-      "learning_rate": 0.00011073170731707315,
-      "loss": 0.263,
+      "learning_rate": 0.00023236994219653174,
+      "loss": 0.3204,
       "step": 892
     },
     {
       "epoch": 4.0,
-      "learning_rate": 0.00011024390243902438,
-      "loss": 0.4579,
+      "learning_rate": 0.0002321965317919075,
+      "loss": 0.771,
       "step": 893
     },
     {
       "epoch": 4.01,
-      "learning_rate": 0.0001097560975609756,
-      "loss": 0.4033,
+      "learning_rate": 0.0002320231213872832,
+      "loss": 0.5956,
       "step": 894
     },
     {
       "epoch": 4.01,
-      "learning_rate": 0.00010926829268292683,
-      "loss": 0.313,
+      "learning_rate": 0.00023184971098265893,
+      "loss": 0.4446,
       "step": 895
     },
     {
       "epoch": 4.02,
-      "learning_rate": 0.00010878048780487805,
-      "loss": 0.2394,
+      "learning_rate": 0.00023167630057803465,
+      "loss": 0.3546,
       "step": 896
     },
     {
       "epoch": 4.02,
-      "learning_rate": 0.00010829268292682925,
-      "loss": 0.2394,
+      "learning_rate": 0.0002315028901734104,
+      "loss": 0.3401,
       "step": 897
     },
     {
       "epoch": 4.03,
-      "learning_rate": 0.00010780487804878048,
-      "loss": 0.2715,
+      "learning_rate": 0.00023132947976878612,
+      "loss": 0.3294,
       "step": 898
     },
     {
       "epoch": 4.03,
-      "learning_rate": 0.0001073170731707317,
-      "loss": 0.24,
+      "learning_rate": 0.00023115606936416181,
+      "loss": 0.3047,
       "step": 899
     },
     {
       "epoch": 4.04,
-      "learning_rate": 0.00010682926829268291,
-      "loss": 0.2086,
+      "learning_rate": 0.00023098265895953754,
+      "loss": 0.2599,
       "step": 900
     },
     {
       "epoch": 4.04,
-      "learning_rate": 0.00010634146341463414,
-      "loss": 0.2325,
+      "learning_rate": 0.00023080924855491328,
+      "loss": 0.2687,
       "step": 901
     },
     {
       "epoch": 4.04,
-      "learning_rate": 0.00010585365853658536,
-      "loss": 0.1837,
+      "learning_rate": 0.000230635838150289,
+      "loss": 0.2312,
       "step": 902
     },
     {
       "epoch": 4.05,
-      "learning_rate": 0.00010536585365853656,
-      "loss": 0.1641,
+      "learning_rate": 0.00023046242774566472,
+      "loss": 0.2279,
       "step": 903
     },
     {
       "epoch": 4.05,
-      "learning_rate": 0.0001048780487804878,
-      "loss": 0.169,
+      "learning_rate": 0.00023028901734104042,
+      "loss": 0.2295,
       "step": 904
     },
     {
       "epoch": 4.06,
-      "learning_rate": 0.00010439024390243901,
-      "loss": 0.1444,
+      "learning_rate": 0.00023011560693641617,
+      "loss": 0.1895,
       "step": 905
     },
     {
       "epoch": 4.06,
-      "learning_rate": 0.00010390243902439023,
-      "loss": 0.1955,
+      "learning_rate": 0.0002299421965317919,
+      "loss": 0.236,
       "step": 906
     },
     {
       "epoch": 4.07,
-      "learning_rate": 0.00010341463414634146,
-      "loss": 0.1712,
+      "learning_rate": 0.0002297687861271676,
+      "loss": 0.2184,
       "step": 907
     },
     {
       "epoch": 4.07,
-      "learning_rate": 0.00010292682926829268,
-      "loss": 0.1723,
+      "learning_rate": 0.00022959537572254333,
+      "loss": 0.1909,
       "step": 908
     },
     {
       "epoch": 4.08,
-      "learning_rate": 0.0001024390243902439,
-      "loss": 0.1577,
+      "learning_rate": 0.00022942196531791908,
+      "loss": 0.1998,
       "step": 909
     },
     {
       "epoch": 4.08,
-      "learning_rate": 0.00010195121951219511,
-      "loss": 0.1411,
+      "learning_rate": 0.00022924855491329477,
+      "loss": 0.1957,
       "step": 910
     },
     {
       "epoch": 4.09,
-      "learning_rate": 0.00010146341463414633,
-      "loss": 0.147,
+      "learning_rate": 0.0002290751445086705,
+      "loss": 0.1838,
       "step": 911
     },
     {
       "epoch": 4.09,
-      "learning_rate": 0.00010097560975609756,
-      "loss": 0.1688,
+      "learning_rate": 0.0002289017341040462,
+      "loss": 0.1752,
       "step": 912
     },
     {
       "epoch": 4.09,
-      "learning_rate": 0.00010048780487804877,
-      "loss": 0.1425,
+      "learning_rate": 0.00022872832369942196,
+      "loss": 0.202,
       "step": 913
     },
     {
       "epoch": 4.1,
-      "learning_rate": 9.999999999999999e-05,
-      "loss": 0.184,
+      "learning_rate": 0.00022855491329479768,
+      "loss": 0.1954,
       "step": 914
     },
     {
       "epoch": 4.1,
-      "learning_rate": 9.951219512195122e-05,
-      "loss": 0.1604,
+      "learning_rate": 0.00022838150289017337,
+      "loss": 0.1885,
       "step": 915
     },
     {
       "epoch": 4.11,
-      "learning_rate": 9.902439024390243e-05,
-      "loss": 0.1326,
+      "learning_rate": 0.0002282080924855491,
+      "loss": 0.1631,
       "step": 916
     },
     {
       "epoch": 4.11,
-      "learning_rate": 9.853658536585364e-05,
-      "loss": 0.1505,
+      "learning_rate": 0.00022803468208092484,
+      "loss": 0.1711,
       "step": 917
     },
     {
       "epoch": 4.12,
-      "learning_rate": 9.804878048780487e-05,
-      "loss": 0.1165,
+      "learning_rate": 0.00022786127167630056,
+      "loss": 0.1738,
       "step": 918
     },
     {
       "epoch": 4.12,
-      "learning_rate": 9.756097560975609e-05,
-      "loss": 0.1217,
+      "learning_rate": 0.00022768786127167628,
+      "loss": 0.1571,
       "step": 919
     },
     {
       "epoch": 4.13,
-      "learning_rate": 9.70731707317073e-05,
-      "loss": 0.1376,
+      "learning_rate": 0.00022751445086705198,
+      "loss": 0.1176,
       "step": 920
     },
     {
       "epoch": 4.13,
-      "learning_rate": 9.658536585365854e-05,
-      "loss": 0.1416,
+      "learning_rate": 0.00022734104046242772,
+      "loss": 0.1831,
       "step": 921
     },
     {
       "epoch": 4.13,
-      "learning_rate": 9.609756097560974e-05,
-      "loss": 0.1483,
+      "learning_rate": 0.00022716763005780344,
+      "loss": 0.1878,
       "step": 922
     },
     {
       "epoch": 4.14,
-      "learning_rate": 9.560975609756097e-05,
-      "loss": 0.1441,
+      "learning_rate": 0.00022699421965317917,
+      "loss": 0.1848,
       "step": 923
     },
     {
       "epoch": 4.14,
-      "learning_rate": 9.512195121951219e-05,
-      "loss": 0.1459,
+      "learning_rate": 0.00022682080924855489,
+      "loss": 0.2019,
       "step": 924
     },
     {
       "epoch": 4.15,
-      "learning_rate": 9.46341463414634e-05,
-      "loss": 0.138,
+      "learning_rate": 0.00022664739884393063,
+      "loss": 0.1753,
       "step": 925
     },
     {
       "epoch": 4.15,
-      "learning_rate": 9.414634146341463e-05,
-      "loss": 0.1196,
+      "learning_rate": 0.00022647398843930635,
+      "loss": 0.1706,
       "step": 926
     },
     {
       "epoch": 4.16,
-      "learning_rate": 9.365853658536585e-05,
-      "loss": 0.1039,
+      "learning_rate": 0.00022630057803468205,
+      "loss": 0.1571,
       "step": 927
     },
     {
       "epoch": 4.16,
-      "learning_rate": 9.317073170731706e-05,
-      "loss": 0.1461,
+      "learning_rate": 0.00022612716763005777,
+      "loss": 0.1915,
       "step": 928
     },
     {
       "epoch": 4.17,
-      "learning_rate": 9.268292682926829e-05,
-      "loss": 0.156,
+      "learning_rate": 0.00022595375722543352,
+      "loss": 0.1784,
       "step": 929
     },
     {
       "epoch": 4.17,
-      "learning_rate": 9.21951219512195e-05,
-      "loss": 0.1005,
+      "learning_rate": 0.00022578034682080924,
+      "loss": 0.1372,
       "step": 930
     },
     {
       "epoch": 4.17,
-      "learning_rate": 9.170731707317072e-05,
-      "loss": 0.1227,
+      "learning_rate": 0.00022560693641618496,
+      "loss": 0.1602,
       "step": 931
     },
     {
       "epoch": 4.18,
-      "learning_rate": 9.121951219512195e-05,
-      "loss": 0.1158,
+      "learning_rate": 0.00022543352601156065,
+      "loss": 0.1511,
       "step": 932
     },
     {
       "epoch": 4.18,
-      "learning_rate": 9.073170731707317e-05,
-      "loss": 0.1069,
+      "learning_rate": 0.0002252601156069364,
+      "loss": 0.1482,
       "step": 933
     },
     {
       "epoch": 4.19,
-      "learning_rate": 9.024390243902438e-05,
-      "loss": 0.1131,
+      "learning_rate": 0.00022508670520231212,
+      "loss": 0.1547,
       "step": 934
     },
     {
       "epoch": 4.19,
-      "learning_rate": 8.97560975609756e-05,
-      "loss": 0.0864,
+      "learning_rate": 0.00022491329479768784,
+      "loss": 0.1249,
       "step": 935
     },
     {
       "epoch": 4.2,
-      "learning_rate": 8.926829268292682e-05,
-      "loss": 0.0965,
+      "learning_rate": 0.00022473988439306356,
+      "loss": 0.1547,
       "step": 936
     },
     {
       "epoch": 4.2,
-      "learning_rate": 8.878048780487805e-05,
-      "loss": 0.0969,
+      "learning_rate": 0.0002245664739884393,
+      "loss": 0.1549,
       "step": 937
     },
     {
       "epoch": 4.21,
-      "learning_rate": 8.829268292682926e-05,
-      "loss": 0.1018,
+      "learning_rate": 0.000224393063583815,
+      "loss": 0.138,
       "step": 938
     },
     {
       "epoch": 4.21,
-      "learning_rate": 8.780487804878047e-05,
-      "loss": 0.0772,
+      "learning_rate": 0.00022421965317919072,
+      "loss": 0.1416,
       "step": 939
     },
     {
       "epoch": 4.22,
-      "learning_rate": 8.73170731707317e-05,
-      "loss": 0.0811,
+      "learning_rate": 0.00022404624277456644,
+      "loss": 0.1237,
       "step": 940
     },
     {
       "epoch": 4.22,
-      "learning_rate": 8.682926829268292e-05,
-      "loss": 0.108,
+      "learning_rate": 0.0002238728323699422,
+      "loss": 0.1804,
       "step": 941
     },
     {
       "epoch": 4.22,
-      "learning_rate": 8.634146341463413e-05,
-      "loss": 0.1689,
+      "learning_rate": 0.0002236994219653179,
+      "loss": 0.2177,
       "step": 942
     },
     {
       "epoch": 4.23,
-      "learning_rate": 8.585365853658536e-05,
-      "loss": 0.4178,
+      "learning_rate": 0.0002235260115606936,
+      "loss": 0.8002,
       "step": 943
     },
     {
       "epoch": 4.23,
-      "learning_rate": 8.536585365853658e-05,
-      "loss": 0.3293,
+      "learning_rate": 0.00022335260115606933,
+      "loss": 0.5931,
       "step": 944
     },
     {
       "epoch": 4.24,
-      "learning_rate": 8.487804878048778e-05,
-      "loss": 0.3048,
+      "learning_rate": 0.00022317919075144507,
+      "loss": 0.4373,
       "step": 945
     },
     {
       "epoch": 4.24,
-      "learning_rate": 8.439024390243901e-05,
-      "loss": 0.2403,
+      "learning_rate": 0.0002230057803468208,
+      "loss": 0.3612,
       "step": 946
     },
     {
       "epoch": 4.25,
-      "learning_rate": 8.390243902439023e-05,
-      "loss": 0.2498,
+      "learning_rate": 0.00022283236994219652,
+      "loss": 0.37,
       "step": 947
     },
     {
       "epoch": 4.25,
-      "learning_rate": 8.341463414634146e-05,
-      "loss": 0.1854,
+      "learning_rate": 0.00022265895953757224,
+      "loss": 0.2676,
       "step": 948
     },
     {
       "epoch": 4.26,
-      "learning_rate": 8.292682926829268e-05,
-      "loss": 0.1997,
+      "learning_rate": 0.00022248554913294798,
+      "loss": 0.2842,
       "step": 949
     },
     {
       "epoch": 4.26,
-      "learning_rate": 8.24390243902439e-05,
-      "loss": 0.2298,
+      "learning_rate": 0.00022231213872832368,
+      "loss": 0.3132,
       "step": 950
     },
     {
       "epoch": 4.26,
-      "learning_rate": 8.195121951219513e-05,
-      "loss": 0.1632,
+      "learning_rate": 0.0002221387283236994,
+      "loss": 0.2282,
       "step": 951
     },
     {
       "epoch": 4.27,
-      "learning_rate": 8.146341463414633e-05,
-      "loss": 0.1737,
+      "learning_rate": 0.00022196531791907512,
+      "loss": 0.238,
       "step": 952
     },
     {
       "epoch": 4.27,
-      "learning_rate": 8.097560975609755e-05,
-      "loss": 0.1601,
+      "learning_rate": 0.00022179190751445087,
+      "loss": 0.2261,
       "step": 953
     },
     {
       "epoch": 4.28,
-      "learning_rate": 8.048780487804878e-05,
-      "loss": 0.1663,
+      "learning_rate": 0.0002216184971098266,
+      "loss": 0.2274,
       "step": 954
     },
     {
       "epoch": 4.28,
-      "learning_rate": 7.999999999999999e-05,
-      "loss": 0.187,
+      "learning_rate": 0.00022144508670520228,
+      "loss": 0.2562,
       "step": 955
     },
     {
       "epoch": 4.29,
-      "learning_rate": 7.951219512195121e-05,
-      "loss": 0.1445,
+      "learning_rate": 0.000221271676300578,
+      "loss": 0.1999,
       "step": 956
     },
     {
       "epoch": 4.29,
-      "learning_rate": 7.902439024390244e-05,
-      "loss": 0.1499,
+      "learning_rate": 0.00022109826589595375,
+      "loss": 0.1859,
       "step": 957
     },
     {
       "epoch": 4.3,
-      "learning_rate": 7.853658536585364e-05,
-      "loss": 0.1406,
+      "learning_rate": 0.00022092485549132947,
+      "loss": 0.1847,
       "step": 958
     },
     {
       "epoch": 4.3,
-      "learning_rate": 7.804878048780487e-05,
-      "loss": 0.1639,
+      "learning_rate": 0.0002207514450867052,
+      "loss": 0.2134,
       "step": 959
     },
     {
       "epoch": 4.3,
-      "learning_rate": 7.756097560975609e-05,
-      "loss": 0.1295,
+      "learning_rate": 0.00022057803468208088,
+      "loss": 0.188,
       "step": 960
     },
     {
       "epoch": 4.31,
-      "learning_rate": 7.707317073170731e-05,
-      "loss": 0.1336,
+      "learning_rate": 0.00022040462427745663,
+      "loss": 0.2077,
       "step": 961
     },
     {
       "epoch": 4.31,
-      "learning_rate": 7.658536585365854e-05,
-      "loss": 0.1529,
+      "learning_rate": 0.00022023121387283235,
+      "loss": 0.1951,
       "step": 962
     },
     {
       "epoch": 4.32,
-      "learning_rate": 7.609756097560976e-05,
-      "loss": 0.1371,
+      "learning_rate": 0.00022005780346820807,
+      "loss": 0.1824,
       "step": 963
     },
     {
       "epoch": 4.32,
-      "learning_rate": 7.560975609756096e-05,
-      "loss": 0.1307,
+      "learning_rate": 0.0002198843930635838,
+      "loss": 0.1861,
       "step": 964
     },
     {
       "epoch": 4.33,
-      "learning_rate": 7.512195121951219e-05,
-      "loss": 0.1316,
+      "learning_rate": 0.00021971098265895954,
+      "loss": 0.1701,
       "step": 965
     },
     {
       "epoch": 4.33,
-      "learning_rate": 7.46341463414634e-05,
-      "loss": 0.1358,
+      "learning_rate": 0.00021953757225433524,
+      "loss": 0.1884,
       "step": 966
     },
     {
       "epoch": 4.34,
-      "learning_rate": 7.414634146341462e-05,
-      "loss": 0.1215,
+      "learning_rate": 0.00021936416184971096,
+      "loss": 0.152,
       "step": 967
     },
     {
       "epoch": 4.34,
-      "learning_rate": 7.365853658536584e-05,
-      "loss": 0.1351,
+      "learning_rate": 0.00021919075144508668,
+      "loss": 0.1824,
       "step": 968
     },
     {
       "epoch": 4.35,
-      "learning_rate": 7.317073170731707e-05,
-      "loss": 0.121,
+      "learning_rate": 0.00021901734104046243,
+      "loss": 0.1769,
       "step": 969
     },
     {
       "epoch": 4.35,
-      "learning_rate": 7.268292682926829e-05,
-      "loss": 0.1392,
+      "learning_rate": 0.00021884393063583815,
+      "loss": 0.1704,
       "step": 970
     },
     {
       "epoch": 4.35,
-      "learning_rate": 7.21951219512195e-05,
-      "loss": 0.106,
+      "learning_rate": 0.00021867052023121384,
+      "loss": 0.1562,
       "step": 971
     },
     {
       "epoch": 4.36,
-      "learning_rate": 7.170731707317072e-05,
-      "loss": 0.1224,
+      "learning_rate": 0.00021849710982658956,
+      "loss": 0.1738,
       "step": 972
     },
     {
       "epoch": 4.36,
-      "learning_rate": 7.121951219512194e-05,
-      "loss": 0.1107,
+      "learning_rate": 0.0002183236994219653,
+      "loss": 0.1492,
       "step": 973
     },
     {
       "epoch": 4.37,
-      "learning_rate": 7.073170731707317e-05,
-      "loss": 0.1096,
+      "learning_rate": 0.00021815028901734103,
+      "loss": 0.1489,
       "step": 974
     },
     {
       "epoch": 4.37,
-      "learning_rate": 7.024390243902439e-05,
-      "loss": 0.1294,
+      "learning_rate": 0.00021797687861271675,
+      "loss": 0.1694,
       "step": 975
     },
     {
       "epoch": 4.38,
-      "learning_rate": 6.97560975609756e-05,
-      "loss": 0.1206,
+      "learning_rate": 0.00021780346820809247,
+      "loss": 0.1627,
       "step": 976
     },
     {
       "epoch": 4.38,
-      "learning_rate": 6.926829268292682e-05,
-      "loss": 0.1198,
+      "learning_rate": 0.00021763005780346822,
+      "loss": 0.1595,
       "step": 977
     },
     {
       "epoch": 4.39,
-      "learning_rate": 6.878048780487805e-05,
-      "loss": 0.1112,
+      "learning_rate": 0.0002174566473988439,
+      "loss": 0.165,
       "step": 978
     },
     {
       "epoch": 4.39,
-      "learning_rate": 6.829268292682925e-05,
-      "loss": 0.1085,
+      "learning_rate": 0.00021728323699421963,
+      "loss": 0.1581,
       "step": 979
     },
     {
       "epoch": 4.39,
-      "learning_rate": 6.780487804878048e-05,
-      "loss": 0.089,
+      "learning_rate": 0.00021710982658959535,
+      "loss": 0.1361,
       "step": 980
     },
     {
       "epoch": 4.4,
-      "learning_rate": 6.73170731707317e-05,
-      "loss": 0.0981,
+      "learning_rate": 0.0002169364161849711,
+      "loss": 0.1503,
       "step": 981
     },
     {
       "epoch": 4.4,
-      "learning_rate": 6.682926829268293e-05,
-      "loss": 0.1129,
+      "learning_rate": 0.00021676300578034682,
+      "loss": 0.1666,
       "step": 982
     },
     {
       "epoch": 4.41,
-      "learning_rate": 6.634146341463413e-05,
-      "loss": 0.1026,
+      "learning_rate": 0.00021658959537572251,
+      "loss": 0.1467,
       "step": 983
     },
     {
       "epoch": 4.41,
-      "learning_rate": 6.585365853658536e-05,
-      "loss": 0.0927,
+      "learning_rate": 0.00021641618497109824,
+      "loss": 0.1599,
       "step": 984
     },
     {
       "epoch": 4.42,
-      "learning_rate": 6.536585365853658e-05,
-      "loss": 0.0987,
+      "learning_rate": 0.00021624277456647396,
+      "loss": 0.1572,
       "step": 985
     },
     {
       "epoch": 4.42,
-      "learning_rate": 6.48780487804878e-05,
-      "loss": 0.1299,
+      "learning_rate": 0.0002160693641618497,
+      "loss": 0.1825,
       "step": 986
     },
     {
       "epoch": 4.43,
-      "learning_rate": 6.439024390243902e-05,
-      "loss": 0.0781,
+      "learning_rate": 0.00021589595375722542,
+      "loss": 0.1265,
       "step": 987
     },
     {
       "epoch": 4.43,
-      "learning_rate": 6.390243902439025e-05,
-      "loss": 0.1048,
+      "learning_rate": 0.00021572254335260112,
+      "loss": 0.136,
       "step": 988
     },
     {
       "epoch": 4.43,
-      "learning_rate": 6.341463414634146e-05,
-      "loss": 0.0572,
+      "learning_rate": 0.00021554913294797684,
+      "loss": 0.1099,
       "step": 989
     },
     {
       "epoch": 4.44,
-      "learning_rate": 6.292682926829268e-05,
-      "loss": 0.1033,
+      "learning_rate": 0.00021537572254335259,
+      "loss": 0.1753,
       "step": 990
     },
     {
       "epoch": 4.44,
-      "learning_rate": 6.24390243902439e-05,
-      "loss": 0.1029,
+      "learning_rate": 0.0002152023121387283,
+      "loss": 0.1497,
       "step": 991
     },
     {
       "epoch": 4.45,
-      "learning_rate": 6.195121951219511e-05,
-      "loss": 0.1846,
+      "learning_rate": 0.00021502890173410403,
+      "loss": 0.2148,
       "step": 992
     },
     {
       "epoch": 4.45,
-      "learning_rate": 6.146341463414633e-05,
-      "loss": 0.3495,
+      "learning_rate": 0.00021485549132947972,
+      "loss": 0.9389,
       "step": 993
     },
     {
       "epoch": 4.46,
-      "learning_rate": 6.0975609756097554e-05,
-      "loss": 0.3143,
+      "learning_rate": 0.00021468208092485547,
+      "loss": 0.619,
       "step": 994
     },
     {
       "epoch": 4.46,
-      "learning_rate": 6.048780487804878e-05,
-      "loss": 0.2743,
+      "learning_rate": 0.0002145086705202312,
+      "loss": 0.4313,
       "step": 995
     },
     {
       "epoch": 4.47,
-      "learning_rate": 5.9999999999999995e-05,
-      "loss": 0.1964,
+      "learning_rate": 0.0002143352601156069,
+      "loss": 0.3153,
       "step": 996
     },
     {
       "epoch": 4.47,
-      "learning_rate": 5.951219512195121e-05,
-      "loss": 0.2446,
+      "learning_rate": 0.00021416184971098263,
+      "loss": 0.38,
       "step": 997
     },
     {
       "epoch": 4.48,
-      "learning_rate": 5.9024390243902435e-05,
-      "loss": 0.1735,
+      "learning_rate": 0.00021398843930635838,
+      "loss": 0.2856,
       "step": 998
     },
     {
       "epoch": 4.48,
-      "learning_rate": 5.853658536585365e-05,
-      "loss": 0.2045,
+      "learning_rate": 0.0002138150289017341,
+      "loss": 0.3066,
       "step": 999
     },
     {
       "epoch": 4.48,
-      "learning_rate": 5.804878048780487e-05,
-      "loss": 0.1399,
+      "learning_rate": 0.0002136416184971098,
+      "loss": 0.2133,
       "step": 1000
     },
     {
       "epoch": 4.48,
-      "eval_loss": 0.36502909660339355,
-      "eval_runtime": 567.0865,
-      "eval_samples_per_second": 4.659,
-      "eval_steps_per_second": 0.584,
-      "eval_wer": 0.11207854026180088,
+      "eval_loss": 0.3967271149158478,
+      "eval_runtime": 570.6501,
+      "eval_samples_per_second": 4.63,
+      "eval_steps_per_second": 0.58,
+      "eval_wer": 0.12705275684252282,
       "step": 1000
     },
     {
       "epoch": 4.49,
-      "learning_rate": 5.756097560975609e-05,
-      "loss": 0.1683,
+      "learning_rate": 0.00021346820809248551,
+      "loss": 0.2367,
       "step": 1001
     },
     {
       "epoch": 4.49,
-      "learning_rate": 5.707317073170731e-05,
-      "loss": 0.1527,
+      "learning_rate": 0.00021329479768786126,
+      "loss": 0.219,
       "step": 1002
     },
     {
       "epoch": 4.5,
-      "learning_rate": 5.6585365853658533e-05,
-      "loss": 0.1248,
+      "learning_rate": 0.00021312138728323698,
+      "loss": 0.1921,
       "step": 1003
     },
     {
       "epoch": 4.5,
-      "learning_rate": 5.609756097560975e-05,
-      "loss": 0.1299,
+      "learning_rate": 0.0002129479768786127,
+      "loss": 0.1953,
       "step": 1004
     },
     {
       "epoch": 4.51,
-      "learning_rate": 5.560975609756097e-05,
-      "loss": 0.1563,
+      "learning_rate": 0.0002127745664739884,
+      "loss": 0.2085,
       "step": 1005
     },
     {
       "epoch": 4.51,
-      "learning_rate": 5.512195121951219e-05,
-      "loss": 0.1239,
+      "learning_rate": 0.00021260115606936414,
+      "loss": 0.1831,
       "step": 1006
     },
     {
       "epoch": 4.52,
-      "learning_rate": 5.4634146341463415e-05,
-      "loss": 0.1481,
+      "learning_rate": 0.00021242774566473987,
+      "loss": 0.2097,
       "step": 1007
     },
     {
       "epoch": 4.52,
-      "learning_rate": 5.4146341463414625e-05,
-      "loss": 0.1316,
+      "learning_rate": 0.00021225433526011559,
+      "loss": 0.1797,
       "step": 1008
     },
     {
       "epoch": 4.52,
-      "learning_rate": 5.365853658536585e-05,
-      "loss": 0.1313,
+      "learning_rate": 0.0002120809248554913,
+      "loss": 0.1994,
       "step": 1009
     },
     {
       "epoch": 4.53,
-      "learning_rate": 5.317073170731707e-05,
-      "loss": 0.1185,
+      "learning_rate": 0.00021190751445086705,
+      "loss": 0.173,
       "step": 1010
     },
     {
       "epoch": 4.53,
-      "learning_rate": 5.268292682926828e-05,
-      "loss": 0.1288,
+      "learning_rate": 0.00021173410404624275,
+      "loss": 0.1891,
       "step": 1011
     },
     {
       "epoch": 4.54,
-      "learning_rate": 5.2195121951219506e-05,
-      "loss": 0.1273,
+      "learning_rate": 0.00021156069364161847,
+      "loss": 0.1908,
       "step": 1012
     },
     {
       "epoch": 4.54,
-      "learning_rate": 5.170731707317073e-05,
-      "loss": 0.1418,
+      "learning_rate": 0.0002113872832369942,
+      "loss": 0.2069,
       "step": 1013
     },
     {
       "epoch": 4.55,
-      "learning_rate": 5.121951219512195e-05,
-      "loss": 0.1133,
+      "learning_rate": 0.00021121387283236994,
+      "loss": 0.1566,
       "step": 1014
     },
     {
       "epoch": 4.55,
-      "learning_rate": 5.0731707317073163e-05,
-      "loss": 0.1178,
+      "learning_rate": 0.00021104046242774566,
+      "loss": 0.1691,
       "step": 1015
     },
     {
       "epoch": 4.56,
-      "learning_rate": 5.024390243902439e-05,
-      "loss": 0.097,
+      "learning_rate": 0.00021086705202312135,
+      "loss": 0.1543,
       "step": 1016
     },
     {
       "epoch": 4.56,
-      "learning_rate": 4.975609756097561e-05,
-      "loss": 0.1185,
+      "learning_rate": 0.00021069364161849707,
+      "loss": 0.162,
       "step": 1017
     },
     {
       "epoch": 4.57,
-      "learning_rate": 4.926829268292682e-05,
-      "loss": 0.1349,
+      "learning_rate": 0.00021052023121387282,
+      "loss": 0.1871,
       "step": 1018
     },
     {
       "epoch": 4.57,
-      "learning_rate": 4.8780487804878045e-05,
-      "loss": 0.1152,
+      "learning_rate": 0.00021034682080924854,
+      "loss": 0.179,
       "step": 1019
     },
     {
       "epoch": 4.57,
-      "learning_rate": 4.829268292682927e-05,
-      "loss": 0.1227,
+      "learning_rate": 0.00021017341040462426,
+      "loss": 0.1806,
       "step": 1020
     },
     {
       "epoch": 4.58,
-      "learning_rate": 4.7804878048780485e-05,
-      "loss": 0.1017,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 0.1621,
       "step": 1021
     },
     {
       "epoch": 4.58,
-      "learning_rate": 4.73170731707317e-05,
-      "loss": 0.0982,
+      "learning_rate": 0.00020982658959537573,
+      "loss": 0.1616,
       "step": 1022
     },
     {
       "epoch": 4.59,
-      "learning_rate": 4.6829268292682926e-05,
-      "loss": 0.1289,
+      "learning_rate": 0.00020965317919075142,
+      "loss": 0.1854,
       "step": 1023
     },
     {
       "epoch": 4.59,
-      "learning_rate": 4.634146341463414e-05,
-      "loss": 0.0997,
+      "learning_rate": 0.00020947976878612714,
+      "loss": 0.1454,
       "step": 1024
     },
     {
       "epoch": 4.6,
-      "learning_rate": 4.585365853658536e-05,
-      "loss": 0.0879,
+      "learning_rate": 0.00020930635838150286,
+      "loss": 0.2193,
       "step": 1025
     },
     {
       "epoch": 4.6,
-      "learning_rate": 4.536585365853658e-05,
-      "loss": 0.0836,
+      "learning_rate": 0.0002091329479768786,
+      "loss": 0.1518,
       "step": 1026
     },
     {
       "epoch": 4.61,
-      "learning_rate": 4.48780487804878e-05,
-      "loss": 0.1062,
+      "learning_rate": 0.00020895953757225433,
+      "loss": 0.1593,
       "step": 1027
     },
     {
       "epoch": 4.61,
-      "learning_rate": 4.4390243902439024e-05,
-      "loss": 0.0964,
+      "learning_rate": 0.00020878612716763003,
+      "loss": 0.1579,
       "step": 1028
     },
     {
       "epoch": 4.61,
-      "learning_rate": 4.3902439024390234e-05,
-      "loss": 0.0981,
+      "learning_rate": 0.00020861271676300575,
+      "loss": 0.1378,
       "step": 1029
     },
     {
       "epoch": 4.62,
-      "learning_rate": 4.341463414634146e-05,
-      "loss": 0.068,
+      "learning_rate": 0.0002084393063583815,
+      "loss": 0.11,
       "step": 1030
     },
     {
       "epoch": 4.62,
-      "learning_rate": 4.292682926829268e-05,
-      "loss": 0.1051,
+      "learning_rate": 0.00020826589595375722,
+      "loss": 0.1537,
       "step": 1031
     },
     {
       "epoch": 4.63,
-      "learning_rate": 4.243902439024389e-05,
-      "loss": 0.0968,
+      "learning_rate": 0.00020809248554913294,
+      "loss": 0.1464,
       "step": 1032
     },
     {
       "epoch": 4.63,
-      "learning_rate": 4.1951219512195115e-05,
-      "loss": 0.1041,
+      "learning_rate": 0.00020791907514450863,
+      "loss": 0.1472,
       "step": 1033
     },
     {
       "epoch": 4.64,
-      "learning_rate": 4.146341463414634e-05,
-      "loss": 0.0835,
+      "learning_rate": 0.00020774566473988438,
+      "loss": 0.1454,
       "step": 1034
     },
     {
       "epoch": 4.64,
-      "learning_rate": 4.097560975609756e-05,
-      "loss": 0.1127,
+      "learning_rate": 0.0002075722543352601,
+      "loss": 0.1799,
       "step": 1035
     },
     {
       "epoch": 4.65,
-      "learning_rate": 4.048780487804877e-05,
-      "loss": 0.1029,
+      "learning_rate": 0.00020739884393063582,
+      "loss": 0.1408,
       "step": 1036
     },
     {
       "epoch": 4.65,
-      "learning_rate": 3.9999999999999996e-05,
-      "loss": 0.0855,
+      "learning_rate": 0.00020722543352601154,
+      "loss": 0.1415,
       "step": 1037
     },
     {
       "epoch": 4.65,
-      "learning_rate": 3.951219512195122e-05,
-      "loss": 0.0733,
+      "learning_rate": 0.0002070520231213873,
+      "loss": 0.1279,
       "step": 1038
     },
     {
       "epoch": 4.66,
-      "learning_rate": 3.902439024390244e-05,
-      "loss": 0.0936,
+      "learning_rate": 0.00020687861271676298,
+      "loss": 0.1413,
       "step": 1039
     },
     {
       "epoch": 4.66,
-      "learning_rate": 3.8536585365853654e-05,
-      "loss": 0.0868,
+      "learning_rate": 0.0002067052023121387,
+      "loss": 0.1423,
       "step": 1040
     },
     {
       "epoch": 4.67,
-      "learning_rate": 3.804878048780488e-05,
-      "loss": 0.094,
+      "learning_rate": 0.00020653179190751442,
+      "loss": 0.1867,
       "step": 1041
     },
     {
       "epoch": 4.67,
-      "learning_rate": 3.7560975609756095e-05,
-      "loss": 0.1837,
+      "learning_rate": 0.00020635838150289017,
+      "loss": 0.2485,
       "step": 1042
     },
     {
       "epoch": 4.68,
-      "learning_rate": 3.707317073170731e-05,
-      "loss": 0.3259,
+      "learning_rate": 0.0002061849710982659,
+      "loss": 0.5592,
       "step": 1043
     },
     {
       "epoch": 4.68,
-      "learning_rate": 3.6585365853658535e-05,
-      "loss": 0.2573,
+      "learning_rate": 0.00020601156069364158,
+      "loss": 0.4164,
       "step": 1044
     },
     {
       "epoch": 4.69,
-      "learning_rate": 3.609756097560975e-05,
-      "loss": 0.2468,
+      "learning_rate": 0.0002058381502890173,
+      "loss": 0.3318,
       "step": 1045
     },
     {
       "epoch": 4.69,
-      "learning_rate": 3.560975609756097e-05,
-      "loss": 0.2143,
+      "learning_rate": 0.00020566473988439305,
+      "loss": 0.3236,
       "step": 1046
     },
     {
       "epoch": 4.7,
-      "learning_rate": 3.512195121951219e-05,
-      "loss": 0.1886,
+      "learning_rate": 0.00020549132947976877,
+      "loss": 0.2769,
       "step": 1047
     },
     {
       "epoch": 4.7,
-      "learning_rate": 3.463414634146341e-05,
-      "loss": 0.2147,
+      "learning_rate": 0.0002053179190751445,
+      "loss": 0.317,
       "step": 1048
     },
     {
       "epoch": 4.7,
-      "learning_rate": 3.4146341463414627e-05,
-      "loss": 0.182,
+      "learning_rate": 0.00020514450867052021,
+      "loss": 0.2551,
       "step": 1049
     },
     {
       "epoch": 4.71,
-      "learning_rate": 3.365853658536585e-05,
-      "loss": 0.1671,
+      "learning_rate": 0.00020497109826589596,
+      "loss": 0.246,
       "step": 1050
     },
     {
       "epoch": 4.71,
-      "learning_rate": 3.317073170731707e-05,
-      "loss": 0.15,
+      "learning_rate": 0.00020479768786127166,
+      "loss": 0.2346,
       "step": 1051
     },
     {
       "epoch": 4.72,
-      "learning_rate": 3.268292682926829e-05,
-      "loss": 0.1564,
+      "learning_rate": 0.00020462427745664738,
+      "loss": 0.2146,
       "step": 1052
     },
     {
       "epoch": 4.72,
-      "learning_rate": 3.219512195121951e-05,
-      "loss": 0.1517,
+      "learning_rate": 0.0002044508670520231,
+      "loss": 0.1969,
       "step": 1053
     },
     {
       "epoch": 4.73,
-      "learning_rate": 3.170731707317073e-05,
-      "loss": 0.165,
+      "learning_rate": 0.00020427745664739885,
+      "loss": 0.243,
       "step": 1054
     },
     {
       "epoch": 4.73,
-      "learning_rate": 3.121951219512195e-05,
-      "loss": 0.1202,
+      "learning_rate": 0.00020410404624277457,
+      "loss": 0.1915,
       "step": 1055
     },
     {
       "epoch": 4.74,
-      "learning_rate": 3.0731707317073165e-05,
-      "loss": 0.1199,
+      "learning_rate": 0.00020393063583815026,
+      "loss": 0.1672,
       "step": 1056
     },
     {
       "epoch": 4.74,
-      "learning_rate": 3.024390243902439e-05,
-      "loss": 0.1425,
+      "learning_rate": 0.00020375722543352598,
+      "loss": 0.213,
       "step": 1057
     },
     {
       "epoch": 4.74,
-      "learning_rate": 2.9756097560975606e-05,
-      "loss": 0.1322,
+      "learning_rate": 0.00020358381502890173,
+      "loss": 0.178,
       "step": 1058
     },
     {
       "epoch": 4.75,
-      "learning_rate": 2.9268292682926826e-05,
-      "loss": 0.1356,
+      "learning_rate": 0.00020341040462427745,
+      "loss": 0.1977,
       "step": 1059
     },
     {
       "epoch": 4.75,
-      "learning_rate": 2.8780487804878046e-05,
-      "loss": 0.1227,
+      "learning_rate": 0.00020323699421965317,
+      "loss": 0.1956,
       "step": 1060
     },
     {
       "epoch": 4.76,
-      "learning_rate": 2.8292682926829267e-05,
-      "loss": 0.1052,
+      "learning_rate": 0.00020306358381502886,
+      "loss": 0.1645,
       "step": 1061
     },
     {
       "epoch": 4.76,
-      "learning_rate": 2.7804878048780484e-05,
-      "loss": 0.1255,
+      "learning_rate": 0.0002028901734104046,
+      "loss": 0.19,
       "step": 1062
     },
     {
       "epoch": 4.77,
-      "learning_rate": 2.7317073170731707e-05,
-      "loss": 0.0951,
+      "learning_rate": 0.00020271676300578033,
+      "loss": 0.1341,
       "step": 1063
     },
     {
       "epoch": 4.77,
-      "learning_rate": 2.6829268292682924e-05,
-      "loss": 0.1074,
+      "learning_rate": 0.00020254335260115605,
+      "loss": 0.1533,
       "step": 1064
     },
     {
       "epoch": 4.78,
-      "learning_rate": 2.634146341463414e-05,
-      "loss": 0.1143,
+      "learning_rate": 0.00020236994219653177,
+      "loss": 0.1639,
       "step": 1065
     },
     {
       "epoch": 4.78,
-      "learning_rate": 2.5853658536585365e-05,
-      "loss": 0.1112,
+      "learning_rate": 0.00020219653179190752,
+      "loss": 0.1612,
       "step": 1066
     },
     {
       "epoch": 4.78,
-      "learning_rate": 2.5365853658536582e-05,
-      "loss": 0.1014,
+      "learning_rate": 0.00020202312138728321,
+      "loss": 0.1496,
       "step": 1067
     },
     {
       "epoch": 4.79,
-      "learning_rate": 2.4878048780487805e-05,
-      "loss": 0.0987,
+      "learning_rate": 0.00020184971098265893,
+      "loss": 0.1454,
       "step": 1068
     },
     {
       "epoch": 4.79,
-      "learning_rate": 2.4390243902439022e-05,
-      "loss": 0.0907,
+      "learning_rate": 0.00020167630057803466,
+      "loss": 0.1465,
       "step": 1069
     },
     {
       "epoch": 4.8,
-      "learning_rate": 2.3902439024390243e-05,
-      "loss": 0.1075,
+      "learning_rate": 0.0002015028901734104,
+      "loss": 0.1555,
       "step": 1070
     },
     {
       "epoch": 4.8,
-      "learning_rate": 2.3414634146341463e-05,
-      "loss": 0.1121,
+      "learning_rate": 0.00020132947976878612,
+      "loss": 0.1481,
       "step": 1071
     },
     {
       "epoch": 4.81,
-      "learning_rate": 2.292682926829268e-05,
-      "loss": 0.0914,
+      "learning_rate": 0.00020115606936416184,
+      "loss": 0.1376,
       "step": 1072
     },
     {
       "epoch": 4.81,
-      "learning_rate": 2.24390243902439e-05,
-      "loss": 0.1099,
+      "learning_rate": 0.00020098265895953754,
+      "loss": 0.1617,
       "step": 1073
     },
     {
       "epoch": 4.82,
-      "learning_rate": 2.1951219512195117e-05,
-      "loss": 0.0843,
+      "learning_rate": 0.00020080924855491329,
+      "loss": 0.1281,
       "step": 1074
     },
     {
       "epoch": 4.82,
-      "learning_rate": 2.146341463414634e-05,
-      "loss": 0.1039,
+      "learning_rate": 0.000200635838150289,
+      "loss": 0.1483,
       "step": 1075
     },
     {
       "epoch": 4.83,
-      "learning_rate": 2.0975609756097558e-05,
-      "loss": 0.1008,
+      "learning_rate": 0.00020046242774566473,
+      "loss": 0.1589,
       "step": 1076
     },
     {
       "epoch": 4.83,
-      "learning_rate": 2.048780487804878e-05,
-      "loss": 0.0954,
+      "learning_rate": 0.00020028901734104045,
+      "loss": 0.1597,
       "step": 1077
     },
     {
       "epoch": 4.83,
-      "learning_rate": 1.9999999999999998e-05,
-      "loss": 0.0986,
+      "learning_rate": 0.0002001156069364162,
+      "loss": 0.1448,
       "step": 1078
     },
     {
       "epoch": 4.84,
-      "learning_rate": 1.951219512195122e-05,
-      "loss": 0.0843,
+      "learning_rate": 0.0001999421965317919,
+      "loss": 0.1361,
       "step": 1079
     },
     {
       "epoch": 4.84,
-      "learning_rate": 1.902439024390244e-05,
-      "loss": 0.0825,
+      "learning_rate": 0.0001997687861271676,
+      "loss": 0.1457,
       "step": 1080
     },
     {
       "epoch": 4.85,
-      "learning_rate": 1.8536585365853656e-05,
-      "loss": 0.0804,
+      "learning_rate": 0.00019959537572254333,
+      "loss": 0.1266,
       "step": 1081
     },
     {
       "epoch": 4.85,
-      "learning_rate": 1.8048780487804876e-05,
-      "loss": 0.0896,
+      "learning_rate": 0.00019942196531791905,
+      "loss": 0.14,
       "step": 1082
     },
     {
       "epoch": 4.86,
-      "learning_rate": 1.7560975609756096e-05,
-      "loss": 0.0916,
+      "learning_rate": 0.0001992485549132948,
+      "loss": 0.1396,
       "step": 1083
     },
     {
       "epoch": 4.86,
-      "learning_rate": 1.7073170731707313e-05,
-      "loss": 0.0811,
+      "learning_rate": 0.0001990751445086705,
+      "loss": 0.1292,
       "step": 1084
     },
     {
       "epoch": 4.87,
-      "learning_rate": 1.6585365853658534e-05,
-      "loss": 0.0858,
+      "learning_rate": 0.0001989017341040462,
+      "loss": 0.1202,
       "step": 1085
     },
     {
       "epoch": 4.87,
-      "learning_rate": 1.6097560975609754e-05,
-      "loss": 0.0718,
+      "learning_rate": 0.00019872832369942193,
+      "loss": 0.1194,
       "step": 1086
     },
     {
       "epoch": 4.87,
-      "learning_rate": 1.5609756097560974e-05,
-      "loss": 0.0726,
+      "learning_rate": 0.00019855491329479768,
+      "loss": 0.1251,
       "step": 1087
     },
     {
       "epoch": 4.88,
-      "learning_rate": 1.5121951219512194e-05,
-      "loss": 0.0752,
+      "learning_rate": 0.0001983815028901734,
+      "loss": 0.1142,
       "step": 1088
     },
     {
       "epoch": 4.88,
-      "learning_rate": 1.4634146341463413e-05,
-      "loss": 0.1133,
+      "learning_rate": 0.0001982080924855491,
+      "loss": 0.1977,
       "step": 1089
     },
     {
       "epoch": 4.89,
-      "learning_rate": 1.4146341463414633e-05,
-      "loss": 0.0928,
+      "learning_rate": 0.00019803468208092482,
+      "loss": 0.1501,
       "step": 1090
     },
     {
       "epoch": 4.89,
-      "learning_rate": 1.3658536585365854e-05,
-      "loss": 0.1092,
+      "learning_rate": 0.00019786127167630056,
+      "loss": 0.1647,
       "step": 1091
     },
     {
       "epoch": 4.9,
-      "learning_rate": 1.317073170731707e-05,
-      "loss": 0.1081,
+      "learning_rate": 0.00019768786127167629,
+      "loss": 0.2021,
       "step": 1092
     },
     {
       "epoch": 4.9,
-      "learning_rate": 1.2682926829268291e-05,
-      "loss": 0.2064,
+      "learning_rate": 0.000197514450867052,
+      "loss": 0.4591,
       "step": 1093
     },
     {
       "epoch": 4.91,
-      "learning_rate": 1.2195121951219511e-05,
-      "loss": 0.1688,
+      "learning_rate": 0.0001973410404624277,
+      "loss": 0.3462,
       "step": 1094
     },
     {
       "epoch": 4.91,
-      "learning_rate": 1.1707317073170731e-05,
-      "loss": 0.189,
+      "learning_rate": 0.00019716763005780345,
+      "loss": 0.2885,
       "step": 1095
     },
     {
       "epoch": 4.91,
-      "learning_rate": 1.121951219512195e-05,
-      "loss": 0.1708,
+      "learning_rate": 0.00019699421965317917,
+      "loss": 0.2618,
       "step": 1096
     },
     {
       "epoch": 4.92,
-      "learning_rate": 1.073170731707317e-05,
-      "loss": 0.1589,
+      "learning_rate": 0.0001968208092485549,
+      "loss": 0.2376,
       "step": 1097
     },
     {
       "epoch": 4.92,
-      "learning_rate": 1.024390243902439e-05,
-      "loss": 0.1198,
+      "learning_rate": 0.0001966473988439306,
+      "loss": 0.1721,
       "step": 1098
     },
     {
       "epoch": 4.93,
-      "learning_rate": 9.75609756097561e-06,
-      "loss": 0.1327,
+      "learning_rate": 0.00019647398843930636,
+      "loss": 0.1815,
       "step": 1099
     },
     {
       "epoch": 4.93,
-      "learning_rate": 9.268292682926828e-06,
-      "loss": 0.1172,
+      "learning_rate": 0.00019630057803468208,
+      "loss": 0.1885,
       "step": 1100
     },
     {
       "epoch": 4.94,
-      "learning_rate": 8.780487804878048e-06,
-      "loss": 0.1378,
+      "learning_rate": 0.00019612716763005777,
+      "loss": 0.1934,
       "step": 1101
     },
     {
       "epoch": 4.94,
-      "learning_rate": 8.292682926829267e-06,
-      "loss": 0.1146,
+      "learning_rate": 0.0001959537572254335,
+      "loss": 0.1607,
       "step": 1102
     },
     {
       "epoch": 4.95,
-      "learning_rate": 7.804878048780487e-06,
-      "loss": 0.1076,
+      "learning_rate": 0.00019578034682080924,
+      "loss": 0.1666,
       "step": 1103
     },
     {
       "epoch": 4.95,
-      "learning_rate": 7.3170731707317065e-06,
-      "loss": 0.0927,
+      "learning_rate": 0.00019560693641618496,
+      "loss": 0.1609,
       "step": 1104
     },
     {
       "epoch": 4.96,
-      "learning_rate": 6.829268292682927e-06,
-      "loss": 0.104,
+      "learning_rate": 0.00019543352601156068,
+      "loss": 0.1719,
       "step": 1105
     },
     {
       "epoch": 4.96,
-      "learning_rate": 6.3414634146341454e-06,
-      "loss": 0.1055,
+      "learning_rate": 0.00019526011560693637,
+      "loss": 0.1764,
       "step": 1106
     },
     {
       "epoch": 4.96,
-      "learning_rate": 5.853658536585366e-06,
-      "loss": 0.0898,
+      "learning_rate": 0.00019508670520231212,
+      "loss": 0.1512,
       "step": 1107
     },
     {
       "epoch": 4.97,
-      "learning_rate": 5.365853658536585e-06,
-      "loss": 0.1014,
+      "learning_rate": 0.00019491329479768784,
+      "loss": 0.1631,
       "step": 1108
     },
     {
       "epoch": 4.97,
-      "learning_rate": 4.878048780487805e-06,
-      "loss": 0.1008,
+      "learning_rate": 0.00019473988439306356,
+      "loss": 0.1507,
       "step": 1109
     },
     {
       "epoch": 4.98,
-      "learning_rate": 4.390243902439024e-06,
-      "loss": 0.0901,
+      "learning_rate": 0.00019456647398843928,
+      "loss": 0.1329,
       "step": 1110
     },
     {
       "epoch": 4.98,
-      "learning_rate": 3.9024390243902435e-06,
-      "loss": 0.1198,
+      "learning_rate": 0.00019439306358381503,
+      "loss": 0.1985,
       "step": 1111
     },
     {
       "epoch": 4.99,
-      "learning_rate": 3.4146341463414634e-06,
-      "loss": 0.0944,
+      "learning_rate": 0.00019421965317919073,
+      "loss": 0.1496,
       "step": 1112
     },
     {
       "epoch": 4.99,
-      "learning_rate": 2.926829268292683e-06,
-      "loss": 0.0886,
+      "learning_rate": 0.00019404624277456645,
+      "loss": 0.1347,
       "step": 1113
     },
     {
       "epoch": 5.0,
-      "learning_rate": 2.4390243902439023e-06,
-      "loss": 0.0888,
+      "learning_rate": 0.00019387283236994217,
+      "loss": 0.1417,
       "step": 1114
     },
     {
       "epoch": 5.0,
-      "learning_rate": 1.9512195121951218e-06,
-      "loss": 0.2069,
+      "learning_rate": 0.00019369942196531792,
+      "loss": 0.2601,
       "step": 1115
     },
     {
       "epoch": 5.0,
-      "step": 1115,
+      "learning_rate": 0.00019352601156069364,
+      "loss": 0.271,
+      "step": 1116
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 0.00019335260115606933,
+      "loss": 0.2395,
+      "step": 1117
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 0.00019317919075144505,
+      "loss": 0.1913,
+      "step": 1118
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 0.0001930057803468208,
+      "loss": 0.1692,
+      "step": 1119
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 0.00019283236994219652,
+      "loss": 0.1818,
+      "step": 1120
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 0.00019265895953757224,
+      "loss": 0.1619,
+      "step": 1121
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 0.00019248554913294796,
+      "loss": 0.1145,
+      "step": 1122
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 0.0001923121387283237,
+      "loss": 0.1634,
+      "step": 1123
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 0.0001921387283236994,
+      "loss": 0.1414,
+      "step": 1124
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 0.00019196531791907512,
+      "loss": 0.1077,
+      "step": 1125
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 0.00019179190751445084,
+      "loss": 0.1225,
+      "step": 1126
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 0.0001916184971098266,
+      "loss": 0.1166,
+      "step": 1127
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 0.0001914450867052023,
+      "loss": 0.115,
+      "step": 1128
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 0.000191271676300578,
+      "loss": 0.1072,
+      "step": 1129
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 0.00019109826589595373,
+      "loss": 0.1018,
+      "step": 1130
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 0.00019092485549132947,
+      "loss": 0.1254,
+      "step": 1131
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 0.0001907514450867052,
+      "loss": 0.1188,
+      "step": 1132
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 0.00019057803468208091,
+      "loss": 0.1018,
+      "step": 1133
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 0.0001904046242774566,
+      "loss": 0.0953,
+      "step": 1134
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 0.00019023121387283236,
+      "loss": 0.1073,
+      "step": 1135
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 0.00019005780346820808,
+      "loss": 0.0956,
+      "step": 1136
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 0.0001898843930635838,
+      "loss": 0.0952,
+      "step": 1137
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 0.00018971098265895952,
+      "loss": 0.1017,
+      "step": 1138
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 0.00018953757225433527,
+      "loss": 0.0923,
+      "step": 1139
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 0.00018936416184971096,
+      "loss": 0.0987,
+      "step": 1140
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 0.00018919075144508668,
+      "loss": 0.1126,
+      "step": 1141
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 0.0001890173410404624,
+      "loss": 0.0983,
+      "step": 1142
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 0.00018884393063583815,
+      "loss": 0.0825,
+      "step": 1143
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 0.00018867052023121387,
+      "loss": 0.0934,
+      "step": 1144
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 0.0001884971098265896,
+      "loss": 0.0904,
+      "step": 1145
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 0.00018832369942196528,
+      "loss": 0.0959,
+      "step": 1146
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 0.00018815028901734103,
+      "loss": 0.0749,
+      "step": 1147
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 0.00018797687861271675,
+      "loss": 0.1007,
+      "step": 1148
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 0.00018780346820809247,
+      "loss": 0.0995,
+      "step": 1149
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 0.0001876300578034682,
+      "loss": 0.0912,
+      "step": 1150
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 0.00018745664739884394,
+      "loss": 0.0994,
+      "step": 1151
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 0.00018728323699421963,
+      "loss": 0.0708,
+      "step": 1152
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 0.00018710982658959536,
+      "loss": 0.1036,
+      "step": 1153
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 0.00018693641618497108,
+      "loss": 0.0811,
+      "step": 1154
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 0.00018676300578034682,
+      "loss": 0.0847,
+      "step": 1155
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 0.00018658959537572254,
+      "loss": 0.0879,
+      "step": 1156
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 0.00018641618497109824,
+      "loss": 0.086,
+      "step": 1157
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 0.00018624277456647396,
+      "loss": 0.0875,
+      "step": 1158
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 0.0001860693641618497,
+      "loss": 0.0579,
+      "step": 1159
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 0.00018589595375722543,
+      "loss": 0.0755,
+      "step": 1160
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 0.00018572254335260115,
+      "loss": 0.0835,
+      "step": 1161
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 0.00018554913294797684,
+      "loss": 0.0806,
+      "step": 1162
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 0.0001853757225433526,
+      "loss": 0.0846,
+      "step": 1163
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 0.0001852023121387283,
+      "loss": 0.0802,
+      "step": 1164
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 0.00018502890173410403,
+      "loss": 0.1096,
+      "step": 1165
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 0.00018485549132947975,
+      "loss": 0.2806,
+      "step": 1166
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 0.0001846820809248555,
+      "loss": 0.2619,
+      "step": 1167
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 0.0001845086705202312,
+      "loss": 0.2016,
+      "step": 1168
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 0.0001843352601156069,
+      "loss": 0.1805,
+      "step": 1169
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 0.00018416184971098263,
+      "loss": 0.152,
+      "step": 1170
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 0.00018398843930635838,
+      "loss": 0.1561,
+      "step": 1171
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 0.0001838150289017341,
+      "loss": 0.135,
+      "step": 1172
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 0.00018364161849710982,
+      "loss": 0.1462,
+      "step": 1173
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 0.00018346820809248552,
+      "loss": 0.121,
+      "step": 1174
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 0.00018329479768786124,
+      "loss": 0.15,
+      "step": 1175
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 0.00018312138728323698,
+      "loss": 0.1087,
+      "step": 1176
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 0.0001829479768786127,
+      "loss": 0.1005,
+      "step": 1177
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 0.00018277456647398843,
+      "loss": 0.1302,
+      "step": 1178
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 0.00018260115606936412,
+      "loss": 0.1133,
+      "step": 1179
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 0.00018242774566473987,
+      "loss": 0.0955,
+      "step": 1180
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 0.0001822543352601156,
+      "loss": 0.0942,
+      "step": 1181
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 0.0001820809248554913,
+      "loss": 0.1154,
+      "step": 1182
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 0.00018190751445086703,
+      "loss": 0.1216,
+      "step": 1183
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 0.00018173410404624278,
+      "loss": 0.0994,
+      "step": 1184
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 0.00018156069364161847,
+      "loss": 0.0857,
+      "step": 1185
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 0.0001813872832369942,
+      "loss": 0.1136,
+      "step": 1186
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 0.0001812138728323699,
+      "loss": 0.1012,
+      "step": 1187
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 0.00018104046242774566,
+      "loss": 0.095,
+      "step": 1188
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 0.00018086705202312138,
+      "loss": 0.0799,
+      "step": 1189
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 0.00018069364161849707,
+      "loss": 0.0927,
+      "step": 1190
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 0.0001805202312138728,
+      "loss": 0.0887,
+      "step": 1191
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 0.00018034682080924854,
+      "loss": 0.0991,
+      "step": 1192
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 0.00018017341040462426,
+      "loss": 0.1128,
+      "step": 1193
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.0827,
+      "step": 1194
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 0.0001798265895953757,
+      "loss": 0.0856,
+      "step": 1195
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 0.00017965317919075145,
+      "loss": 0.085,
+      "step": 1196
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 0.00017947976878612715,
+      "loss": 0.1049,
+      "step": 1197
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 0.00017930635838150287,
+      "loss": 0.0995,
+      "step": 1198
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 0.0001791329479768786,
+      "loss": 0.0757,
+      "step": 1199
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 0.00017895953757225434,
+      "loss": 0.0879,
+      "step": 1200
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 0.00017878612716763006,
+      "loss": 0.0971,
+      "step": 1201
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 0.00017861271676300575,
+      "loss": 0.0995,
+      "step": 1202
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 0.00017843930635838147,
+      "loss": 0.0709,
+      "step": 1203
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 0.00017826589595375722,
+      "loss": 0.0798,
+      "step": 1204
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 0.00017809248554913294,
+      "loss": 0.0719,
+      "step": 1205
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 0.00017791907514450866,
+      "loss": 0.0798,
+      "step": 1206
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 0.00017774566473988435,
+      "loss": 0.0892,
+      "step": 1207
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 0.0001775722543352601,
+      "loss": 0.0759,
+      "step": 1208
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 0.00017739884393063582,
+      "loss": 0.0816,
+      "step": 1209
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 0.00017722543352601154,
+      "loss": 0.087,
+      "step": 1210
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 0.00017705202312138726,
+      "loss": 0.0928,
+      "step": 1211
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 0.000176878612716763,
+      "loss": 0.0925,
+      "step": 1212
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 0.0001767052023121387,
+      "loss": 0.061,
+      "step": 1213
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 0.00017653179190751442,
+      "loss": 0.1159,
+      "step": 1214
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 0.00017635838150289015,
+      "loss": 0.159,
+      "step": 1215
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 0.0001761849710982659,
+      "loss": 0.3672,
+      "step": 1216
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 0.00017601156069364161,
+      "loss": 0.2528,
+      "step": 1217
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 0.0001758381502890173,
+      "loss": 0.2057,
+      "step": 1218
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 0.00017566473988439303,
+      "loss": 0.1701,
+      "step": 1219
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 0.00017549132947976878,
+      "loss": 0.1412,
+      "step": 1220
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 0.0001753179190751445,
+      "loss": 0.1459,
+      "step": 1221
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 0.00017514450867052022,
+      "loss": 0.146,
+      "step": 1222
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 0.00017497109826589594,
+      "loss": 0.1425,
+      "step": 1223
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 0.00017479768786127169,
+      "loss": 0.1307,
+      "step": 1224
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 0.00017462427745664738,
+      "loss": 0.1416,
+      "step": 1225
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 0.0001744508670520231,
+      "loss": 0.1239,
+      "step": 1226
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 0.00017427745664739882,
+      "loss": 0.135,
+      "step": 1227
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 0.00017410404624277457,
+      "loss": 0.0945,
+      "step": 1228
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 0.0001739306358381503,
+      "loss": 0.1212,
+      "step": 1229
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 0.00017375722543352598,
+      "loss": 0.1213,
+      "step": 1230
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 0.0001735838150289017,
+      "loss": 0.11,
+      "step": 1231
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 0.00017341040462427745,
+      "loss": 0.1107,
+      "step": 1232
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 0.00017323699421965317,
+      "loss": 0.1042,
+      "step": 1233
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 0.0001730635838150289,
+      "loss": 0.1009,
+      "step": 1234
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 0.00017289017341040459,
+      "loss": 0.0995,
+      "step": 1235
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 0.00017271676300578033,
+      "loss": 0.1199,
+      "step": 1236
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 0.00017254335260115605,
+      "loss": 0.1462,
+      "step": 1237
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 0.00017236994219653178,
+      "loss": 0.0984,
+      "step": 1238
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 0.0001721965317919075,
+      "loss": 0.097,
+      "step": 1239
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 0.00017202312138728324,
+      "loss": 0.1193,
+      "step": 1240
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 0.00017184971098265894,
+      "loss": 0.1235,
+      "step": 1241
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 0.00017167630057803466,
+      "loss": 0.1089,
+      "step": 1242
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 0.00017150289017341038,
+      "loss": 0.1091,
+      "step": 1243
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 0.00017132947976878613,
+      "loss": 0.0913,
+      "step": 1244
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 0.00017115606936416185,
+      "loss": 0.1071,
+      "step": 1245
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 0.00017098265895953757,
+      "loss": 0.0796,
+      "step": 1246
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 0.00017080924855491326,
+      "loss": 0.0782,
+      "step": 1247
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 0.000170635838150289,
+      "loss": 0.0974,
+      "step": 1248
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 0.00017046242774566473,
+      "loss": 0.0765,
+      "step": 1249
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 0.00017028901734104045,
+      "loss": 0.0882,
+      "step": 1250
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 0.00017011560693641617,
+      "loss": 0.0782,
+      "step": 1251
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 0.00016994219653179192,
+      "loss": 0.085,
+      "step": 1252
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 0.0001697687861271676,
+      "loss": 0.0744,
+      "step": 1253
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 0.00016959537572254333,
+      "loss": 0.0714,
+      "step": 1254
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 0.00016942196531791905,
+      "loss": 0.0929,
+      "step": 1255
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 0.0001692485549132948,
+      "loss": 0.0866,
+      "step": 1256
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 0.00016907514450867052,
+      "loss": 0.1397,
+      "step": 1257
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 0.00016890173410404622,
+      "loss": 0.069,
+      "step": 1258
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 0.00016872832369942194,
+      "loss": 0.0713,
+      "step": 1259
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 0.00016855491329479768,
+      "loss": 0.0674,
+      "step": 1260
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 0.0001683815028901734,
+      "loss": 0.0674,
+      "step": 1261
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 0.00016820809248554913,
+      "loss": 0.0649,
+      "step": 1262
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 0.00016803468208092482,
+      "loss": 0.0891,
+      "step": 1263
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 0.00016786127167630057,
+      "loss": 0.079,
+      "step": 1264
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 0.0001676878612716763,
+      "loss": 0.1048,
+      "step": 1265
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 0.000167514450867052,
+      "loss": 0.2792,
+      "step": 1266
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 0.00016734104046242773,
+      "loss": 0.2596,
+      "step": 1267
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 0.00016716763005780348,
+      "loss": 0.2365,
+      "step": 1268
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 0.0001669942196531792,
+      "loss": 0.1353,
+      "step": 1269
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 0.0001668208092485549,
+      "loss": 0.1578,
+      "step": 1270
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 0.0001666473988439306,
+      "loss": 0.14,
+      "step": 1271
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 0.00016647398843930633,
+      "loss": 0.1199,
+      "step": 1272
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 0.00016630057803468208,
+      "loss": 0.1419,
+      "step": 1273
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 0.0001661271676300578,
+      "loss": 0.1321,
+      "step": 1274
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 0.0001659537572254335,
+      "loss": 0.1182,
+      "step": 1275
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 0.00016578034682080922,
+      "loss": 0.1181,
+      "step": 1276
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 0.00016560693641618496,
+      "loss": 0.1181,
+      "step": 1277
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 0.00016543352601156068,
+      "loss": 0.1041,
+      "step": 1278
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 0.0001652601156069364,
+      "loss": 0.1226,
+      "step": 1279
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 0.0001650867052023121,
+      "loss": 0.1154,
+      "step": 1280
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 0.00016491329479768785,
+      "loss": 0.1378,
+      "step": 1281
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 0.00016473988439306357,
+      "loss": 0.1154,
+      "step": 1282
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 0.0001645664739884393,
+      "loss": 0.1216,
+      "step": 1283
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 0.000164393063583815,
+      "loss": 0.1142,
+      "step": 1284
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 0.00016421965317919076,
+      "loss": 0.1157,
+      "step": 1285
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 0.00016404624277456645,
+      "loss": 0.0988,
+      "step": 1286
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 0.00016387283236994217,
+      "loss": 0.0949,
+      "step": 1287
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 0.0001636994219653179,
+      "loss": 0.0961,
+      "step": 1288
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 0.00016352601156069364,
+      "loss": 0.0861,
+      "step": 1289
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 0.00016335260115606936,
+      "loss": 0.0907,
+      "step": 1290
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 0.00016317919075144505,
+      "loss": 0.0893,
+      "step": 1291
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 0.00016300578034682077,
+      "loss": 0.0836,
+      "step": 1292
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 0.00016283236994219652,
+      "loss": 0.0774,
+      "step": 1293
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 0.00016265895953757224,
+      "loss": 0.0944,
+      "step": 1294
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 0.00016248554913294796,
+      "loss": 0.0802,
+      "step": 1295
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 0.00016231213872832368,
+      "loss": 0.0914,
+      "step": 1296
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 0.00016213872832369943,
+      "loss": 0.0849,
+      "step": 1297
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 0.00016196531791907512,
+      "loss": 0.0907,
+      "step": 1298
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 0.00016179190751445085,
+      "loss": 0.0839,
+      "step": 1299
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 0.00016161849710982657,
+      "loss": 0.0997,
+      "step": 1300
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 0.00016144508670520231,
+      "loss": 0.0792,
+      "step": 1301
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 0.00016127167630057803,
+      "loss": 0.0732,
+      "step": 1302
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 0.00016109826589595373,
+      "loss": 0.084,
+      "step": 1303
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 0.00016092485549132945,
+      "loss": 0.0786,
+      "step": 1304
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 0.0001607514450867052,
+      "loss": 0.0919,
+      "step": 1305
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 0.00016057803468208092,
+      "loss": 0.0896,
+      "step": 1306
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 0.00016040462427745664,
+      "loss": 0.1043,
+      "step": 1307
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 0.00016023121387283233,
+      "loss": 0.0694,
+      "step": 1308
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 0.00016005780346820808,
+      "loss": 0.0945,
+      "step": 1309
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 0.0001598843930635838,
+      "loss": 0.0835,
+      "step": 1310
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 0.00015971098265895952,
+      "loss": 0.0805,
+      "step": 1311
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 0.00015953757225433524,
+      "loss": 0.074,
+      "step": 1312
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 0.000159364161849711,
+      "loss": 0.0883,
+      "step": 1313
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 0.00015919075144508668,
+      "loss": 0.0976,
+      "step": 1314
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 0.0001590173410404624,
+      "loss": 0.1321,
+      "step": 1315
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 0.00015884393063583812,
+      "loss": 0.28,
+      "step": 1316
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 0.00015867052023121387,
+      "loss": 0.1622,
+      "step": 1317
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 0.0001584971098265896,
+      "loss": 0.1625,
+      "step": 1318
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 0.0001583236994219653,
+      "loss": 0.1282,
+      "step": 1319
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 0.000158150289017341,
+      "loss": 0.1433,
+      "step": 1320
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 0.00015797687861271675,
+      "loss": 0.118,
+      "step": 1321
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 0.00015780346820809248,
+      "loss": 0.1088,
+      "step": 1322
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 0.0001576300578034682,
+      "loss": 0.1085,
+      "step": 1323
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 0.00015745664739884392,
+      "loss": 0.1076,
+      "step": 1324
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 0.00015728323699421966,
+      "loss": 0.1046,
+      "step": 1325
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 0.00015710982658959536,
+      "loss": 0.1198,
+      "step": 1326
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 0.00015693641618497108,
+      "loss": 0.1045,
+      "step": 1327
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 0.0001567630057803468,
+      "loss": 0.0901,
+      "step": 1328
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 0.00015658959537572255,
+      "loss": 0.0972,
+      "step": 1329
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 0.00015641618497109827,
+      "loss": 0.0917,
+      "step": 1330
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 0.00015624277456647396,
+      "loss": 0.0869,
+      "step": 1331
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 0.00015606936416184968,
+      "loss": 0.1015,
+      "step": 1332
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 0.00015589595375722543,
+      "loss": 0.0788,
+      "step": 1333
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 0.00015572254335260115,
+      "loss": 0.1053,
+      "step": 1334
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 0.00015554913294797687,
+      "loss": 0.0844,
+      "step": 1335
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 0.00015537572254335256,
+      "loss": 0.0606,
+      "step": 1336
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 0.0001552023121387283,
+      "loss": 0.0745,
+      "step": 1337
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 0.00015502890173410403,
+      "loss": 0.1608,
+      "step": 1338
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 0.00015485549132947975,
+      "loss": 0.1628,
+      "step": 1339
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 0.00015468208092485547,
+      "loss": 0.1261,
+      "step": 1340
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 0.00015450867052023122,
+      "loss": 0.1063,
+      "step": 1341
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 0.00015433526011560692,
+      "loss": 0.0978,
+      "step": 1342
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 0.00015416184971098264,
+      "loss": 0.0947,
+      "step": 1343
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 0.00015398843930635836,
+      "loss": 0.1061,
+      "step": 1344
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 0.0001538150289017341,
+      "loss": 0.0974,
+      "step": 1345
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 0.00015364161849710983,
+      "loss": 0.0876,
+      "step": 1346
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 0.00015346820809248555,
+      "loss": 0.098,
+      "step": 1347
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 0.00015329479768786124,
+      "loss": 0.0721,
+      "step": 1348
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 0.000153121387283237,
+      "loss": 0.088,
+      "step": 1349
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 0.0001529479768786127,
+      "loss": 0.0813,
+      "step": 1350
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 0.00015277456647398843,
+      "loss": 0.0787,
+      "step": 1351
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 0.00015260115606936415,
+      "loss": 0.0781,
+      "step": 1352
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 0.0001524277456647399,
+      "loss": 0.0658,
+      "step": 1353
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 0.0001522543352601156,
+      "loss": 0.0721,
+      "step": 1354
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 0.0001520809248554913,
+      "loss": 0.0887,
+      "step": 1355
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 0.00015190751445086703,
+      "loss": 0.056,
+      "step": 1356
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 0.00015173410404624278,
+      "loss": 0.0834,
+      "step": 1357
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 0.0001515606936416185,
+      "loss": 0.0627,
+      "step": 1358
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 0.0001513872832369942,
+      "loss": 0.0651,
+      "step": 1359
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 0.00015121387283236992,
+      "loss": 0.0729,
+      "step": 1360
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 0.00015104046242774566,
+      "loss": 0.0604,
+      "step": 1361
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 0.00015086705202312138,
+      "loss": 0.0651,
+      "step": 1362
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 0.0001506936416184971,
+      "loss": 0.0665,
+      "step": 1363
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 0.0001505202312138728,
+      "loss": 0.0703,
+      "step": 1364
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 0.00015034682080924855,
+      "loss": 0.0708,
+      "step": 1365
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 0.00015017341040462427,
+      "loss": 0.0615,
+      "step": 1366
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 0.00015,
+      "loss": 0.0501,
+      "step": 1367
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 0.0001498265895953757,
+      "loss": 0.0749,
+      "step": 1368
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 0.00014965317919075143,
+      "loss": 0.068,
+      "step": 1369
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 0.00014947976878612715,
+      "loss": 0.0599,
+      "step": 1370
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 0.00014930635838150287,
+      "loss": 0.0688,
+      "step": 1371
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 0.0001491329479768786,
+      "loss": 0.0627,
+      "step": 1372
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 0.0001489595375722543,
+      "loss": 0.0627,
+      "step": 1373
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 0.00014878612716763003,
+      "loss": 0.0514,
+      "step": 1374
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 0.00014861271676300578,
+      "loss": 0.0587,
+      "step": 1375
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 0.00014843930635838147,
+      "loss": 0.0575,
+      "step": 1376
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 0.00014826589595375722,
+      "loss": 0.0561,
+      "step": 1377
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 0.00014809248554913291,
+      "loss": 0.074,
+      "step": 1378
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 0.00014791907514450866,
+      "loss": 0.0533,
+      "step": 1379
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 0.00014774566473988438,
+      "loss": 0.0455,
+      "step": 1380
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 0.0001475722543352601,
+      "loss": 0.0586,
+      "step": 1381
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 0.00014739884393063582,
+      "loss": 0.0423,
+      "step": 1382
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 0.00014722543352601154,
+      "loss": 0.0507,
+      "step": 1383
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 0.00014705202312138727,
+      "loss": 0.0511,
+      "step": 1384
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 0.00014687861271676299,
+      "loss": 0.0786,
+      "step": 1385
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 0.0001467052023121387,
+      "loss": 0.0446,
+      "step": 1386
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 0.00014653179190751443,
+      "loss": 0.0511,
+      "step": 1387
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 0.00014635838150289015,
+      "loss": 0.0718,
+      "step": 1388
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 0.0001461849710982659,
+      "loss": 0.1718,
+      "step": 1389
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 0.0001460115606936416,
+      "loss": 0.1058,
+      "step": 1390
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 0.00014583815028901734,
+      "loss": 0.1114,
+      "step": 1391
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 0.00014566473988439306,
+      "loss": 0.1036,
+      "step": 1392
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 0.00014549132947976878,
+      "loss": 0.092,
+      "step": 1393
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 0.0001453179190751445,
+      "loss": 0.0812,
+      "step": 1394
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 0.00014514450867052022,
+      "loss": 0.1137,
+      "step": 1395
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 0.00014497109826589594,
+      "loss": 0.0868,
+      "step": 1396
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 0.00014479768786127166,
+      "loss": 0.0821,
+      "step": 1397
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 0.00014462427745664738,
+      "loss": 0.1276,
+      "step": 1398
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 0.0001444508670520231,
+      "loss": 0.0639,
+      "step": 1399
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 0.00014427745664739882,
+      "loss": 0.0941,
+      "step": 1400
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 0.00014410404624277454,
+      "loss": 0.0772,
+      "step": 1401
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 0.00014393063583815026,
+      "loss": 0.0682,
+      "step": 1402
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 0.000143757225433526,
+      "loss": 0.0796,
+      "step": 1403
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 0.0001435838150289017,
+      "loss": 0.0795,
+      "step": 1404
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 0.00014341040462427745,
+      "loss": 0.0709,
+      "step": 1405
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 0.00014323699421965317,
+      "loss": 0.0782,
+      "step": 1406
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 0.0001430635838150289,
+      "loss": 0.0603,
+      "step": 1407
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 0.00014289017341040462,
+      "loss": 0.0712,
+      "step": 1408
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 0.00014271676300578034,
+      "loss": 0.0608,
+      "step": 1409
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 0.00014254335260115606,
+      "loss": 0.0686,
+      "step": 1410
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 0.00014236994219653178,
+      "loss": 0.0798,
+      "step": 1411
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 0.0001421965317919075,
+      "loss": 0.0663,
+      "step": 1412
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 0.00014202312138728322,
+      "loss": 0.0611,
+      "step": 1413
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 0.00014184971098265894,
+      "loss": 0.0591,
+      "step": 1414
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 0.00014167630057803466,
+      "loss": 0.0665,
+      "step": 1415
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 0.00014150289017341038,
+      "loss": 0.0568,
+      "step": 1416
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 0.00014132947976878613,
+      "loss": 0.0707,
+      "step": 1417
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 0.00014115606936416182,
+      "loss": 0.1241,
+      "step": 1418
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 0.00014098265895953757,
+      "loss": 0.0755,
+      "step": 1419
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 0.0001408092485549133,
+      "loss": 0.0778,
+      "step": 1420
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 0.000140635838150289,
+      "loss": 0.0621,
+      "step": 1421
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 0.00014046242774566473,
+      "loss": 0.062,
+      "step": 1422
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 0.00014028901734104045,
+      "loss": 0.0593,
+      "step": 1423
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 0.00014011560693641617,
+      "loss": 0.073,
+      "step": 1424
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 0.0001399421965317919,
+      "loss": 0.0475,
+      "step": 1425
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 0.00013976878612716762,
+      "loss": 0.0642,
+      "step": 1426
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 0.00013959537572254334,
+      "loss": 0.0514,
+      "step": 1427
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 0.00013942196531791906,
+      "loss": 0.0532,
+      "step": 1428
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 0.00013924855491329478,
+      "loss": 0.0605,
+      "step": 1429
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 0.0001390751445086705,
+      "loss": 0.0558,
+      "step": 1430
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 0.00013890173410404625,
+      "loss": 0.0475,
+      "step": 1431
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 0.00013872832369942194,
+      "loss": 0.0571,
+      "step": 1432
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 0.0001385549132947977,
+      "loss": 0.0638,
+      "step": 1433
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 0.0001383815028901734,
+      "loss": 0.0507,
+      "step": 1434
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 0.00013820809248554913,
+      "loss": 0.0645,
+      "step": 1435
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 0.00013803468208092485,
+      "loss": 0.0627,
+      "step": 1436
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 0.00013786127167630057,
+      "loss": 0.072,
+      "step": 1437
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 0.0001376878612716763,
+      "loss": 0.1086,
+      "step": 1438
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 0.000137514450867052,
+      "loss": 0.1585,
+      "step": 1439
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 0.00013734104046242773,
+      "loss": 0.2069,
+      "step": 1440
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 0.00013716763005780345,
+      "loss": 0.0923,
+      "step": 1441
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 0.00013699421965317917,
+      "loss": 0.1004,
+      "step": 1442
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 0.00013682080924855492,
+      "loss": 0.1034,
+      "step": 1443
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 0.00013664739884393061,
+      "loss": 0.0949,
+      "step": 1444
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 0.00013647398843930636,
+      "loss": 0.0756,
+      "step": 1445
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 0.00013630057803468206,
+      "loss": 0.0965,
+      "step": 1446
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 0.0001361271676300578,
+      "loss": 0.0813,
+      "step": 1447
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 0.00013595375722543352,
+      "loss": 0.0904,
+      "step": 1448
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 0.00013578034682080925,
+      "loss": 0.0747,
+      "step": 1449
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 0.00013560693641618497,
+      "loss": 0.0892,
+      "step": 1450
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 0.0001354335260115607,
+      "loss": 0.0595,
+      "step": 1451
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 0.0001352601156069364,
+      "loss": 0.0702,
+      "step": 1452
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 0.00013508670520231213,
+      "loss": 0.0778,
+      "step": 1453
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 0.00013491329479768785,
+      "loss": 0.0634,
+      "step": 1454
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 0.00013473988439306357,
+      "loss": 0.0731,
+      "step": 1455
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 0.0001345664739884393,
+      "loss": 0.076,
+      "step": 1456
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 0.00013439306358381504,
+      "loss": 0.0652,
+      "step": 1457
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 0.00013421965317919073,
+      "loss": 0.0547,
+      "step": 1458
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 0.00013404624277456648,
+      "loss": 0.0743,
+      "step": 1459
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 0.00013387283236994217,
+      "loss": 0.0592,
+      "step": 1460
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 0.00013369942196531792,
+      "loss": 0.0639,
+      "step": 1461
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 0.00013352601156069364,
+      "loss": 0.06,
+      "step": 1462
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 0.00013335260115606936,
+      "loss": 0.0712,
+      "step": 1463
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 0.00013317919075144508,
+      "loss": 0.065,
+      "step": 1464
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 0.00013300578034682078,
+      "loss": 0.0553,
+      "step": 1465
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 0.00013283236994219652,
+      "loss": 0.0654,
+      "step": 1466
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 0.00013265895953757224,
+      "loss": 0.0631,
+      "step": 1467
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 0.00013248554913294797,
+      "loss": 0.0729,
+      "step": 1468
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 0.00013231213872832369,
+      "loss": 0.0612,
+      "step": 1469
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 0.0001321387283236994,
+      "loss": 0.049,
+      "step": 1470
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 0.00013196531791907513,
+      "loss": 0.0715,
+      "step": 1471
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 0.00013179190751445085,
+      "loss": 0.0667,
+      "step": 1472
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 0.00013161849710982657,
+      "loss": 0.0601,
+      "step": 1473
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 0.0001314450867052023,
+      "loss": 0.0543,
+      "step": 1474
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 0.000131271676300578,
+      "loss": 0.0533,
+      "step": 1475
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 0.00013109826589595376,
+      "loss": 0.0508,
+      "step": 1476
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 0.00013092485549132945,
+      "loss": 0.0581,
+      "step": 1477
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 0.0001307514450867052,
+      "loss": 0.0631,
+      "step": 1478
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 0.00013057803468208092,
+      "loss": 0.0545,
+      "step": 1479
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 0.00013040462427745664,
+      "loss": 0.0621,
+      "step": 1480
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 0.00013023121387283236,
+      "loss": 0.0797,
+      "step": 1481
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 0.00013005780346820808,
+      "loss": 0.0494,
+      "step": 1482
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 0.0001298843930635838,
+      "loss": 0.063,
+      "step": 1483
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 0.00012971098265895952,
+      "loss": 0.0556,
+      "step": 1484
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 0.00012953757225433524,
+      "loss": 0.031,
+      "step": 1485
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 0.00012936416184971096,
+      "loss": 0.0532,
+      "step": 1486
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 0.00012919075144508669,
+      "loss": 0.0419,
+      "step": 1487
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 0.0001290173410404624,
+      "loss": 0.0777,
+      "step": 1488
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 0.00012884393063583813,
+      "loss": 0.1556,
+      "step": 1489
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 0.00012867052023121387,
+      "loss": 0.15,
+      "step": 1490
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 0.00012849710982658957,
+      "loss": 0.1092,
+      "step": 1491
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 0.00012832369942196532,
+      "loss": 0.0985,
+      "step": 1492
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 0.00012815028901734104,
+      "loss": 0.089,
+      "step": 1493
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 0.00012797687861271676,
+      "loss": 0.0951,
+      "step": 1494
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 0.00012780346820809248,
+      "loss": 0.0904,
+      "step": 1495
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 0.0001276300578034682,
+      "loss": 0.0874,
+      "step": 1496
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 0.00012745664739884392,
+      "loss": 0.0686,
+      "step": 1497
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 0.00012728323699421964,
+      "loss": 0.0789,
+      "step": 1498
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 0.00012710982658959536,
+      "loss": 0.0757,
+      "step": 1499
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 0.00012693641618497108,
+      "loss": 0.0886,
+      "step": 1500
+    },
+    {
+      "epoch": 6.73,
+      "eval_loss": 0.3513650596141815,
+      "eval_runtime": 571.1505,
+      "eval_samples_per_second": 4.626,
+      "eval_steps_per_second": 0.58,
+      "eval_wer": 0.10093216977389925,
+      "step": 1500
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 0.0001267630057803468,
+      "loss": 0.0772,
+      "step": 1501
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 0.00012658959537572252,
+      "loss": 0.0731,
+      "step": 1502
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 0.00012641618497109824,
+      "loss": 0.0637,
+      "step": 1503
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 0.000126242774566474,
+      "loss": 0.0737,
+      "step": 1504
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 0.00012606936416184968,
+      "loss": 0.0885,
+      "step": 1505
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 0.00012589595375722543,
+      "loss": 0.0575,
+      "step": 1506
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 0.00012572254335260115,
+      "loss": 0.0623,
+      "step": 1507
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 0.00012554913294797687,
+      "loss": 0.0807,
+      "step": 1508
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 0.0001253757225433526,
+      "loss": 0.0606,
+      "step": 1509
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 0.00012520231213872831,
+      "loss": 0.0596,
+      "step": 1510
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 0.00012502890173410404,
+      "loss": 0.066,
+      "step": 1511
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 0.00012485549132947976,
+      "loss": 0.0648,
+      "step": 1512
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 0.00012468208092485548,
+      "loss": 0.0554,
+      "step": 1513
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 0.0001245086705202312,
+      "loss": 0.0601,
+      "step": 1514
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 0.00012433526011560692,
+      "loss": 0.0526,
+      "step": 1515
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 0.00012416184971098267,
+      "loss": 0.0653,
+      "step": 1516
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 0.00012398843930635836,
+      "loss": 0.0682,
+      "step": 1517
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 0.0001238150289017341,
+      "loss": 0.0612,
+      "step": 1518
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 0.0001236416184971098,
+      "loss": 0.0616,
+      "step": 1519
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 0.00012346820809248555,
+      "loss": 0.0609,
+      "step": 1520
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 0.00012329479768786127,
+      "loss": 0.0565,
+      "step": 1521
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 0.000123121387283237,
+      "loss": 0.0499,
+      "step": 1522
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 0.0001229479768786127,
+      "loss": 0.0652,
+      "step": 1523
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 0.00012277456647398843,
+      "loss": 0.0601,
+      "step": 1524
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 0.00012260115606936415,
+      "loss": 0.0348,
+      "step": 1525
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 0.00012242774566473987,
+      "loss": 0.0554,
+      "step": 1526
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 0.0001222543352601156,
+      "loss": 0.0515,
+      "step": 1527
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 0.00012208092485549131,
+      "loss": 0.0556,
+      "step": 1528
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 0.00012190751445086703,
+      "loss": 0.1181,
+      "step": 1529
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 0.00012173410404624277,
+      "loss": 0.0656,
+      "step": 1530
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 0.00012156069364161849,
+      "loss": 0.045,
+      "step": 1531
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 0.00012138728323699421,
+      "loss": 0.0486,
+      "step": 1532
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 0.00012121387283236993,
+      "loss": 0.0517,
+      "step": 1533
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 0.00012104046242774567,
+      "loss": 0.0482,
+      "step": 1534
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 0.00012086705202312137,
+      "loss": 0.0384,
+      "step": 1535
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 0.0001206936416184971,
+      "loss": 0.1258,
+      "step": 1536
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 0.00012052023121387281,
+      "loss": 0.0393,
+      "step": 1537
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 0.00012034682080924855,
+      "loss": 0.0976,
+      "step": 1538
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 0.00012017341040462427,
+      "loss": 0.1335,
+      "step": 1539
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 0.0879,
+      "step": 1540
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 0.00011982658959537571,
+      "loss": 0.0893,
+      "step": 1541
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 0.00011965317919075144,
+      "loss": 0.0729,
+      "step": 1542
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 0.00011947976878612715,
+      "loss": 0.0855,
+      "step": 1543
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 0.00011930635838150289,
+      "loss": 0.0607,
+      "step": 1544
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 0.0001191329479768786,
+      "loss": 0.0633,
+      "step": 1545
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 0.00011895953757225433,
+      "loss": 0.0864,
+      "step": 1546
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 0.00011878612716763005,
+      "loss": 0.0667,
+      "step": 1547
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 0.00011861271676300578,
+      "loss": 0.0621,
+      "step": 1548
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 0.00011843930635838149,
+      "loss": 0.0582,
+      "step": 1549
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 0.00011826589595375722,
+      "loss": 0.0563,
+      "step": 1550
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 0.00011809248554913293,
+      "loss": 0.0595,
+      "step": 1551
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 0.00011791907514450866,
+      "loss": 0.0622,
+      "step": 1552
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 0.00011774566473988439,
+      "loss": 0.0489,
+      "step": 1553
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 0.00011757225433526012,
+      "loss": 0.07,
+      "step": 1554
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 0.00011739884393063583,
+      "loss": 0.0699,
+      "step": 1555
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 0.00011722543352601156,
+      "loss": 0.0564,
+      "step": 1556
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 0.00011705202312138727,
+      "loss": 0.0654,
+      "step": 1557
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 0.000116878612716763,
+      "loss": 0.0719,
+      "step": 1558
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 0.00011670520231213872,
+      "loss": 0.0555,
+      "step": 1559
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 0.00011653179190751443,
+      "loss": 0.0539,
+      "step": 1560
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 0.00011635838150289016,
+      "loss": 0.1026,
+      "step": 1561
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 0.00011618497109826587,
+      "loss": 0.082,
+      "step": 1562
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 0.0001160115606936416,
+      "loss": 0.0802,
+      "step": 1563
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 0.00011583815028901733,
+      "loss": 0.0734,
+      "step": 1564
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 0.00011566473988439306,
+      "loss": 0.075,
+      "step": 1565
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 0.00011549132947976877,
+      "loss": 0.074,
+      "step": 1566
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 0.0001153179190751445,
+      "loss": 0.0624,
+      "step": 1567
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 0.00011514450867052021,
+      "loss": 0.0662,
+      "step": 1568
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 0.00011497109826589594,
+      "loss": 0.0631,
+      "step": 1569
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 0.00011479768786127166,
+      "loss": 0.0493,
+      "step": 1570
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 0.00011462427745664738,
+      "loss": 0.0642,
+      "step": 1571
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 0.0001144508670520231,
+      "loss": 0.0557,
+      "step": 1572
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 0.00011427745664739884,
+      "loss": 0.0515,
+      "step": 1573
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 0.00011410404624277455,
+      "loss": 0.0615,
+      "step": 1574
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 0.00011393063583815028,
+      "loss": 0.0486,
+      "step": 1575
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 0.00011375722543352599,
+      "loss": 0.0503,
+      "step": 1576
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 0.00011358381502890172,
+      "loss": 0.0713,
+      "step": 1577
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 0.00011341040462427744,
+      "loss": 0.0545,
+      "step": 1578
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 0.00011323699421965318,
+      "loss": 0.0459,
+      "step": 1579
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 0.00011306358381502888,
+      "loss": 0.0357,
+      "step": 1580
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 0.00011289017341040462,
+      "loss": 0.0504,
+      "step": 1581
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 0.00011271676300578033,
+      "loss": 0.0427,
+      "step": 1582
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 0.00011254335260115606,
+      "loss": 0.0584,
+      "step": 1583
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 0.00011236994219653178,
+      "loss": 0.0566,
+      "step": 1584
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 0.0001121965317919075,
+      "loss": 0.0452,
+      "step": 1585
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 0.00011202312138728322,
+      "loss": 0.0555,
+      "step": 1586
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 0.00011184971098265896,
+      "loss": 0.0471,
+      "step": 1587
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 0.00011167630057803466,
+      "loss": 0.0488,
+      "step": 1588
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 0.0001115028901734104,
+      "loss": 0.0331,
+      "step": 1589
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 0.00011132947976878612,
+      "loss": 0.0503,
+      "step": 1590
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 0.00011115606936416184,
+      "loss": 0.0475,
+      "step": 1591
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 0.00011098265895953756,
+      "loss": 0.0457,
+      "step": 1592
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 0.0001108092485549133,
+      "loss": 0.0453,
+      "step": 1593
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 0.000110635838150289,
+      "loss": 0.048,
+      "step": 1594
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 0.00011046242774566474,
+      "loss": 0.043,
+      "step": 1595
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 0.00011028901734104044,
+      "loss": 0.0318,
+      "step": 1596
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 0.00011011560693641618,
+      "loss": 0.0444,
+      "step": 1597
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 0.0001099421965317919,
+      "loss": 0.0316,
+      "step": 1598
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 0.00010976878612716762,
+      "loss": 0.0392,
+      "step": 1599
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 0.00010959537572254334,
+      "loss": 0.0411,
+      "step": 1600
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 0.00010942196531791907,
+      "loss": 0.0407,
+      "step": 1601
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 0.00010924855491329478,
+      "loss": 0.0376,
+      "step": 1602
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 0.00010907514450867051,
+      "loss": 0.0372,
+      "step": 1603
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 0.00010890173410404623,
+      "loss": 0.0388,
+      "step": 1604
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 0.00010872832369942196,
+      "loss": 0.0316,
+      "step": 1605
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 0.00010855491329479768,
+      "loss": 0.0364,
+      "step": 1606
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 0.00010838150289017341,
+      "loss": 0.0496,
+      "step": 1607
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 0.00010820809248554912,
+      "loss": 0.0483,
+      "step": 1608
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 0.00010803468208092485,
+      "loss": 0.0372,
+      "step": 1609
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 0.00010786127167630056,
+      "loss": 0.0383,
+      "step": 1610
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 0.00010768786127167629,
+      "loss": 0.0558,
+      "step": 1611
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 0.00010751445086705201,
+      "loss": 0.0953,
+      "step": 1612
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 0.00010734104046242773,
+      "loss": 0.0717,
+      "step": 1613
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 0.00010716763005780346,
+      "loss": 0.0711,
+      "step": 1614
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 0.00010699421965317919,
+      "loss": 0.0735,
+      "step": 1615
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 0.0001068208092485549,
+      "loss": 0.0702,
+      "step": 1616
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 0.00010664739884393063,
+      "loss": 0.0634,
+      "step": 1617
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 0.00010647398843930635,
+      "loss": 0.0637,
+      "step": 1618
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 0.00010630057803468207,
+      "loss": 0.0591,
+      "step": 1619
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 0.00010612716763005779,
+      "loss": 0.0628,
+      "step": 1620
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 0.00010595375722543353,
+      "loss": 0.0501,
+      "step": 1621
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 0.00010578034682080923,
+      "loss": 0.0643,
+      "step": 1622
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 0.00010560693641618497,
+      "loss": 0.0555,
+      "step": 1623
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 0.00010543352601156068,
+      "loss": 0.0606,
+      "step": 1624
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 0.00010526011560693641,
+      "loss": 0.0447,
+      "step": 1625
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 0.00010508670520231213,
+      "loss": 0.0522,
+      "step": 1626
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 0.00010491329479768786,
+      "loss": 0.0526,
+      "step": 1627
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 0.00010473988439306357,
+      "loss": 0.046,
+      "step": 1628
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 0.0001045664739884393,
+      "loss": 0.0683,
+      "step": 1629
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 0.00010439306358381501,
+      "loss": 0.0516,
+      "step": 1630
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 0.00010421965317919075,
+      "loss": 0.0431,
+      "step": 1631
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 0.00010404624277456647,
+      "loss": 0.0407,
+      "step": 1632
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 0.00010387283236994219,
+      "loss": 0.0553,
+      "step": 1633
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 0.00010369942196531791,
+      "loss": 0.0441,
+      "step": 1634
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 0.00010352601156069364,
+      "loss": 0.0361,
+      "step": 1635
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 0.00010335260115606935,
+      "loss": 0.0417,
+      "step": 1636
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 0.00010317919075144509,
+      "loss": 0.0548,
+      "step": 1637
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 0.00010300578034682079,
+      "loss": 0.0483,
+      "step": 1638
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 0.00010283236994219653,
+      "loss": 0.0421,
+      "step": 1639
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 0.00010265895953757225,
+      "loss": 0.0444,
+      "step": 1640
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 0.00010248554913294798,
+      "loss": 0.037,
+      "step": 1641
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 0.00010231213872832369,
+      "loss": 0.046,
+      "step": 1642
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 0.00010213872832369942,
+      "loss": 0.0377,
+      "step": 1643
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 0.00010196531791907513,
+      "loss": 0.043,
+      "step": 1644
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 0.00010179190751445086,
+      "loss": 0.0417,
+      "step": 1645
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 0.00010161849710982658,
+      "loss": 0.0372,
+      "step": 1646
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 0.0001014450867052023,
+      "loss": 0.0432,
+      "step": 1647
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 0.00010127167630057803,
+      "loss": 0.0434,
+      "step": 1648
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 0.00010109826589595376,
+      "loss": 0.0352,
+      "step": 1649
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 0.00010092485549132947,
+      "loss": 0.0402,
+      "step": 1650
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 0.0001007514450867052,
+      "loss": 0.0381,
+      "step": 1651
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 0.00010057803468208092,
+      "loss": 0.0418,
+      "step": 1652
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 0.00010040462427745664,
+      "loss": 0.0339,
+      "step": 1653
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 0.00010023121387283236,
+      "loss": 0.0462,
+      "step": 1654
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 0.0001000578034682081,
+      "loss": 0.0406,
+      "step": 1655
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 9.98843930635838e-05,
+      "loss": 0.0479,
+      "step": 1656
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 9.971098265895953e-05,
+      "loss": 0.0354,
+      "step": 1657
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 9.953757225433525e-05,
+      "loss": 0.0551,
+      "step": 1658
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 9.936416184971097e-05,
+      "loss": 0.0287,
+      "step": 1659
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 9.91907514450867e-05,
+      "loss": 0.0342,
+      "step": 1660
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 9.901734104046241e-05,
+      "loss": 0.063,
+      "step": 1661
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 9.884393063583814e-05,
+      "loss": 0.0938,
+      "step": 1662
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 9.867052023121385e-05,
+      "loss": 0.0789,
+      "step": 1663
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 9.849710982658958e-05,
+      "loss": 0.0637,
+      "step": 1664
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 9.83236994219653e-05,
+      "loss": 0.0571,
+      "step": 1665
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 9.815028901734104e-05,
+      "loss": 0.0686,
+      "step": 1666
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 9.797687861271675e-05,
+      "loss": 0.0591,
+      "step": 1667
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 9.780346820809248e-05,
+      "loss": 0.0504,
+      "step": 1668
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 9.763005780346819e-05,
+      "loss": 0.0571,
+      "step": 1669
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 9.745664739884392e-05,
+      "loss": 0.056,
+      "step": 1670
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 9.728323699421964e-05,
+      "loss": 0.0677,
+      "step": 1671
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 9.710982658959536e-05,
+      "loss": 0.0481,
+      "step": 1672
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 9.693641618497108e-05,
+      "loss": 0.0565,
+      "step": 1673
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 9.676300578034682e-05,
+      "loss": 0.052,
+      "step": 1674
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 9.658959537572253e-05,
+      "loss": 0.0489,
+      "step": 1675
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 9.641618497109826e-05,
+      "loss": 0.0433,
+      "step": 1676
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 9.624277456647398e-05,
+      "loss": 0.0491,
+      "step": 1677
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 9.60693641618497e-05,
+      "loss": 0.0445,
+      "step": 1678
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 9.589595375722542e-05,
+      "loss": 0.0593,
+      "step": 1679
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 9.572254335260116e-05,
+      "loss": 0.0491,
+      "step": 1680
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 9.554913294797686e-05,
+      "loss": 0.0614,
+      "step": 1681
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 9.53757225433526e-05,
+      "loss": 0.0562,
+      "step": 1682
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 9.52023121387283e-05,
+      "loss": 0.0352,
+      "step": 1683
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 9.502890173410404e-05,
+      "loss": 0.0484,
+      "step": 1684
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 9.485549132947976e-05,
+      "loss": 0.0393,
+      "step": 1685
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 9.468208092485548e-05,
+      "loss": 0.0397,
+      "step": 1686
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 9.45086705202312e-05,
+      "loss": 0.0423,
+      "step": 1687
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 9.433526011560693e-05,
+      "loss": 0.0478,
+      "step": 1688
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 9.416184971098264e-05,
+      "loss": 0.0464,
+      "step": 1689
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 9.398843930635838e-05,
+      "loss": 0.0414,
+      "step": 1690
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 9.38150289017341e-05,
+      "loss": 0.051,
+      "step": 1691
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 9.364161849710982e-05,
+      "loss": 0.0405,
+      "step": 1692
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 9.346820809248554e-05,
+      "loss": 0.0464,
+      "step": 1693
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 9.329479768786127e-05,
+      "loss": 0.0312,
+      "step": 1694
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 9.312138728323698e-05,
+      "loss": 0.0429,
+      "step": 1695
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 9.294797687861271e-05,
+      "loss": 0.0376,
+      "step": 1696
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 9.277456647398842e-05,
+      "loss": 0.0374,
+      "step": 1697
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 9.260115606936415e-05,
+      "loss": 0.0412,
+      "step": 1698
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 9.242774566473988e-05,
+      "loss": 0.029,
+      "step": 1699
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 9.22543352601156e-05,
+      "loss": 0.0462,
+      "step": 1700
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 9.208092485549132e-05,
+      "loss": 0.0403,
+      "step": 1701
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 9.190751445086705e-05,
+      "loss": 0.0455,
+      "step": 1702
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 9.173410404624276e-05,
+      "loss": 0.044,
+      "step": 1703
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 9.156069364161849e-05,
+      "loss": 0.0309,
+      "step": 1704
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 9.138728323699421e-05,
+      "loss": 0.0503,
+      "step": 1705
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 9.121387283236993e-05,
+      "loss": 0.0443,
+      "step": 1706
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 9.104046242774565e-05,
+      "loss": 0.0415,
+      "step": 1707
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 9.086705202312139e-05,
+      "loss": 0.0499,
+      "step": 1708
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 9.06936416184971e-05,
+      "loss": 0.0416,
+      "step": 1709
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 9.052023121387283e-05,
+      "loss": 0.0572,
+      "step": 1710
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 9.034682080924854e-05,
+      "loss": 0.0469,
+      "step": 1711
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 9.017341040462427e-05,
+      "loss": 0.0813,
+      "step": 1712
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 0.0663,
+      "step": 1713
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 8.982658959537573e-05,
+      "loss": 0.0592,
+      "step": 1714
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 8.965317919075143e-05,
+      "loss": 0.0604,
+      "step": 1715
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 8.947976878612717e-05,
+      "loss": 0.0669,
+      "step": 1716
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 8.930635838150287e-05,
+      "loss": 0.0608,
+      "step": 1717
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 8.913294797687861e-05,
+      "loss": 0.0554,
+      "step": 1718
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 8.895953757225433e-05,
+      "loss": 0.0624,
+      "step": 1719
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 8.878612716763005e-05,
+      "loss": 0.0649,
+      "step": 1720
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 8.861271676300577e-05,
+      "loss": 0.057,
+      "step": 1721
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 8.84393063583815e-05,
+      "loss": 0.0445,
+      "step": 1722
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 8.826589595375721e-05,
+      "loss": 0.0504,
+      "step": 1723
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 8.809248554913295e-05,
+      "loss": 0.0529,
+      "step": 1724
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 8.791907514450865e-05,
+      "loss": 0.0628,
+      "step": 1725
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 8.774566473988439e-05,
+      "loss": 0.047,
+      "step": 1726
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 8.757225433526011e-05,
+      "loss": 0.0438,
+      "step": 1727
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 8.739884393063584e-05,
+      "loss": 0.0491,
+      "step": 1728
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 8.722543352601155e-05,
+      "loss": 0.0572,
+      "step": 1729
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 8.705202312138728e-05,
+      "loss": 0.0363,
+      "step": 1730
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 8.687861271676299e-05,
+      "loss": 0.0537,
+      "step": 1731
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 8.670520231213873e-05,
+      "loss": 0.0313,
+      "step": 1732
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 8.653179190751445e-05,
+      "loss": 0.0508,
+      "step": 1733
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 8.635838150289017e-05,
+      "loss": 0.043,
+      "step": 1734
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 8.618497109826589e-05,
+      "loss": 0.0354,
+      "step": 1735
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 8.601156069364162e-05,
+      "loss": 0.0435,
+      "step": 1736
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 8.583815028901733e-05,
+      "loss": 0.0427,
+      "step": 1737
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 8.566473988439306e-05,
+      "loss": 0.0411,
+      "step": 1738
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 8.549132947976878e-05,
+      "loss": 0.0423,
+      "step": 1739
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 8.53179190751445e-05,
+      "loss": 0.0406,
+      "step": 1740
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 8.514450867052023e-05,
+      "loss": 0.0399,
+      "step": 1741
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 8.497109826589596e-05,
+      "loss": 0.0419,
+      "step": 1742
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 8.479768786127167e-05,
+      "loss": 0.0363,
+      "step": 1743
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 8.46242774566474e-05,
+      "loss": 0.0392,
+      "step": 1744
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 8.445086705202311e-05,
+      "loss": 0.0461,
+      "step": 1745
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 8.427745664739884e-05,
+      "loss": 0.0458,
+      "step": 1746
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 8.410404624277456e-05,
+      "loss": 0.0416,
+      "step": 1747
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 8.393063583815028e-05,
+      "loss": 0.036,
+      "step": 1748
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 8.3757225433526e-05,
+      "loss": 0.0414,
+      "step": 1749
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 8.358381502890174e-05,
+      "loss": 0.0326,
+      "step": 1750
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 8.341040462427745e-05,
+      "loss": 0.0508,
+      "step": 1751
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 8.323699421965317e-05,
+      "loss": 0.0486,
+      "step": 1752
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 8.30635838150289e-05,
+      "loss": 0.0363,
+      "step": 1753
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 8.289017341040461e-05,
+      "loss": 0.0366,
+      "step": 1754
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 8.271676300578034e-05,
+      "loss": 0.0399,
+      "step": 1755
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 8.254335260115605e-05,
+      "loss": 0.0408,
+      "step": 1756
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 8.236994219653178e-05,
+      "loss": 0.0324,
+      "step": 1757
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 8.21965317919075e-05,
+      "loss": 0.0344,
+      "step": 1758
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 8.202312138728322e-05,
+      "loss": 0.0545,
+      "step": 1759
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 8.184971098265895e-05,
+      "loss": 0.0309,
+      "step": 1760
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 8.167630057803468e-05,
+      "loss": 0.1026,
+      "step": 1761
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 8.150289017341039e-05,
+      "loss": 0.075,
+      "step": 1762
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 8.132947976878612e-05,
+      "loss": 0.0526,
+      "step": 1763
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 8.115606936416184e-05,
+      "loss": 0.064,
+      "step": 1764
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 8.098265895953756e-05,
+      "loss": 0.0562,
+      "step": 1765
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 8.080924855491328e-05,
+      "loss": 0.0543,
+      "step": 1766
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 8.063583815028902e-05,
+      "loss": 0.0638,
+      "step": 1767
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 8.046242774566472e-05,
+      "loss": 0.0364,
+      "step": 1768
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 8.028901734104046e-05,
+      "loss": 0.0621,
+      "step": 1769
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 8.011560693641617e-05,
+      "loss": 0.0567,
+      "step": 1770
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 7.99421965317919e-05,
+      "loss": 0.0482,
+      "step": 1771
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 7.976878612716762e-05,
+      "loss": 0.0518,
+      "step": 1772
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 7.959537572254334e-05,
+      "loss": 0.0447,
+      "step": 1773
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 7.942196531791906e-05,
+      "loss": 0.0483,
+      "step": 1774
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 7.92485549132948e-05,
+      "loss": 0.0337,
+      "step": 1775
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 7.90751445086705e-05,
+      "loss": 0.0454,
+      "step": 1776
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 7.890173410404624e-05,
+      "loss": 0.0472,
+      "step": 1777
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 7.872832369942196e-05,
+      "loss": 0.0442,
+      "step": 1778
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 7.855491329479768e-05,
+      "loss": 0.0333,
+      "step": 1779
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 7.83815028901734e-05,
+      "loss": 0.0467,
+      "step": 1780
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 7.820809248554913e-05,
+      "loss": 0.0412,
+      "step": 1781
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 7.803468208092484e-05,
+      "loss": 0.0464,
+      "step": 1782
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 7.786127167630058e-05,
+      "loss": 0.0307,
+      "step": 1783
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 7.768786127167628e-05,
+      "loss": 0.058,
+      "step": 1784
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 7.751445086705202e-05,
+      "loss": 0.0672,
+      "step": 1785
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 7.734104046242774e-05,
+      "loss": 0.0566,
+      "step": 1786
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 7.716763005780346e-05,
+      "loss": 0.0577,
+      "step": 1787
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 7.699421965317918e-05,
+      "loss": 0.0516,
+      "step": 1788
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 7.682080924855491e-05,
+      "loss": 0.1426,
+      "step": 1789
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 7.664739884393062e-05,
+      "loss": 0.0358,
+      "step": 1790
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 7.647398843930635e-05,
+      "loss": 0.0846,
+      "step": 1791
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 7.630057803468207e-05,
+      "loss": 0.0666,
+      "step": 1792
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 7.61271676300578e-05,
+      "loss": 0.0505,
+      "step": 1793
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 7.595375722543352e-05,
+      "loss": 0.0439,
+      "step": 1794
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 7.578034682080925e-05,
+      "loss": 0.0355,
+      "step": 1795
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 7.560693641618496e-05,
+      "loss": 0.0469,
+      "step": 1796
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 7.543352601156069e-05,
+      "loss": 0.0565,
+      "step": 1797
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 7.52601156069364e-05,
+      "loss": 0.0368,
+      "step": 1798
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 7.508670520231213e-05,
+      "loss": 0.0524,
+      "step": 1799
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 7.491329479768785e-05,
+      "loss": 0.0369,
+      "step": 1800
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 7.473988439306357e-05,
+      "loss": 0.0475,
+      "step": 1801
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 7.45664739884393e-05,
+      "loss": 0.049,
+      "step": 1802
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 7.439306358381502e-05,
+      "loss": 0.0408,
+      "step": 1803
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 7.421965317919074e-05,
+      "loss": 0.0297,
+      "step": 1804
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 7.404624277456646e-05,
+      "loss": 0.099,
+      "step": 1805
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 7.387283236994219e-05,
+      "loss": 0.0438,
+      "step": 1806
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 7.369942196531791e-05,
+      "loss": 0.1455,
+      "step": 1807
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 7.352601156069363e-05,
+      "loss": 0.0296,
+      "step": 1808
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 7.335260115606935e-05,
+      "loss": 0.0297,
+      "step": 1809
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 7.317919075144507e-05,
+      "loss": 0.0351,
+      "step": 1810
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 7.30057803468208e-05,
+      "loss": 0.1393,
+      "step": 1811
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 7.30057803468208e-05,
+      "loss": 0.1093,
+      "step": 1812
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 7.30057803468208e-05,
+      "loss": 0.08,
+      "step": 1813
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 7.283236994219653e-05,
+      "loss": 0.0995,
+      "step": 1814
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 7.265895953757225e-05,
+      "loss": 0.0839,
+      "step": 1815
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 7.248554913294797e-05,
+      "loss": 0.2315,
+      "step": 1816
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 7.231213872832369e-05,
+      "loss": 0.2594,
+      "step": 1817
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 7.213872832369941e-05,
+      "loss": 0.1639,
+      "step": 1818
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 7.196531791907513e-05,
+      "loss": 0.1329,
+      "step": 1819
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 7.179190751445085e-05,
+      "loss": 0.1424,
+      "step": 1820
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 7.161849710982659e-05,
+      "loss": 0.3705,
+      "step": 1821
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 7.144508670520231e-05,
+      "loss": 0.4238,
+      "step": 1822
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 7.127167630057803e-05,
+      "loss": 0.3596,
+      "step": 1823
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 7.109826589595375e-05,
+      "loss": 0.201,
+      "step": 1824
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 7.092485549132947e-05,
+      "loss": 0.184,
+      "step": 1825
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 7.075144508670519e-05,
+      "loss": 0.219,
+      "step": 1826
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 7.057803468208091e-05,
+      "loss": 0.2169,
+      "step": 1827
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 7.040462427745665e-05,
+      "loss": 0.0858,
+      "step": 1828
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 7.023121387283237e-05,
+      "loss": 0.1802,
+      "step": 1829
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 7.005780346820809e-05,
+      "loss": 0.0855,
+      "step": 1830
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 6.988439306358381e-05,
+      "loss": 0.1098,
+      "step": 1831
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 6.971098265895953e-05,
+      "loss": 0.1523,
+      "step": 1832
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 6.953757225433525e-05,
+      "loss": 0.1435,
+      "step": 1833
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 6.936416184971097e-05,
+      "loss": 0.1203,
+      "step": 1834
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 6.91907514450867e-05,
+      "loss": 0.2379,
+      "step": 1835
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 6.901734104046242e-05,
+      "loss": 0.1833,
+      "step": 1836
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 6.884393063583815e-05,
+      "loss": 0.1643,
+      "step": 1837
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 6.867052023121387e-05,
+      "loss": 0.1469,
+      "step": 1838
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 6.849710982658959e-05,
+      "loss": 0.0913,
+      "step": 1839
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 6.832369942196531e-05,
+      "loss": 0.0403,
+      "step": 1840
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 6.815028901734103e-05,
+      "loss": 0.096,
+      "step": 1841
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 6.797687861271676e-05,
+      "loss": 0.1088,
+      "step": 1842
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 6.780346820809248e-05,
+      "loss": 0.2121,
+      "step": 1843
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 6.76300578034682e-05,
+      "loss": 0.1506,
+      "step": 1844
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 6.745664739884392e-05,
+      "loss": 0.0472,
+      "step": 1845
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 6.728323699421964e-05,
+      "loss": 0.0873,
+      "step": 1846
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 6.710982658959537e-05,
+      "loss": 0.1552,
+      "step": 1847
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 6.693641618497109e-05,
+      "loss": 0.047,
+      "step": 1848
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 6.676300578034682e-05,
+      "loss": 0.0831,
+      "step": 1849
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 6.658959537572254e-05,
+      "loss": 0.0891,
+      "step": 1850
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 6.641618497109826e-05,
+      "loss": 0.0929,
+      "step": 1851
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 6.624277456647398e-05,
+      "loss": 0.0883,
+      "step": 1852
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 6.60693641618497e-05,
+      "loss": 0.0795,
+      "step": 1853
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 6.589595375722542e-05,
+      "loss": 0.1435,
+      "step": 1854
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 6.572254335260114e-05,
+      "loss": 0.1741,
+      "step": 1855
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 6.554913294797688e-05,
+      "loss": 0.0961,
+      "step": 1856
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 6.53757225433526e-05,
+      "loss": 0.1966,
+      "step": 1857
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 6.520231213872832e-05,
+      "loss": 0.1273,
+      "step": 1858
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 6.502890173410404e-05,
+      "loss": 0.0822,
+      "step": 1859
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 6.485549132947976e-05,
+      "loss": 0.0902,
+      "step": 1860
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 6.468208092485548e-05,
+      "loss": 0.0709,
+      "step": 1861
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 6.45086705202312e-05,
+      "loss": 0.0338,
+      "step": 1862
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 6.433526011560694e-05,
+      "loss": 0.0312,
+      "step": 1863
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 6.416184971098266e-05,
+      "loss": 0.0409,
+      "step": 1864
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 6.398843930635838e-05,
+      "loss": 0.0308,
+      "step": 1865
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 6.38150289017341e-05,
+      "loss": 0.0323,
+      "step": 1866
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 6.364161849710982e-05,
+      "loss": 0.0841,
+      "step": 1867
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 6.346820809248554e-05,
+      "loss": 0.086,
+      "step": 1868
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 6.329479768786126e-05,
+      "loss": 0.0813,
+      "step": 1869
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 6.3121387283237e-05,
+      "loss": 0.0751,
+      "step": 1870
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 6.294797687861272e-05,
+      "loss": 0.0999,
+      "step": 1871
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 6.277456647398844e-05,
+      "loss": 0.0301,
+      "step": 1872
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 6.260115606936416e-05,
+      "loss": 0.038,
+      "step": 1873
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 6.242774566473988e-05,
+      "loss": 0.0383,
+      "step": 1874
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 6.22543352601156e-05,
+      "loss": 0.0304,
+      "step": 1875
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 6.208092485549133e-05,
+      "loss": 0.0909,
+      "step": 1876
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 6.190751445086705e-05,
+      "loss": 0.0432,
+      "step": 1877
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 6.173410404624277e-05,
+      "loss": 0.0992,
+      "step": 1878
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 6.15606936416185e-05,
+      "loss": 0.0197,
+      "step": 1879
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 6.138728323699422e-05,
+      "loss": 0.0453,
+      "step": 1880
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 6.121387283236994e-05,
+      "loss": 0.0355,
+      "step": 1881
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 6.104046242774566e-05,
+      "loss": 0.0304,
+      "step": 1882
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 6.0867052023121385e-05,
+      "loss": 0.0871,
+      "step": 1883
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 6.0693641618497105e-05,
+      "loss": 0.0741,
+      "step": 1884
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 6.052023121387283e-05,
+      "loss": 0.0567,
+      "step": 1885
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 6.034682080924855e-05,
+      "loss": 0.1103,
+      "step": 1886
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 6.0173410404624274e-05,
+      "loss": 0.0536,
+      "step": 1887
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 0.0413,
+      "step": 1888
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 5.982658959537572e-05,
+      "loss": 0.0408,
+      "step": 1889
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 5.965317919075144e-05,
+      "loss": 0.0403,
+      "step": 1890
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 5.9479768786127164e-05,
+      "loss": 0.042,
+      "step": 1891
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 5.930635838150289e-05,
+      "loss": 0.0374,
+      "step": 1892
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 5.913294797687861e-05,
+      "loss": 0.0483,
+      "step": 1893
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 5.895953757225433e-05,
+      "loss": 0.0782,
+      "step": 1894
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 5.878612716763006e-05,
+      "loss": 0.087,
+      "step": 1895
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 5.861271676300578e-05,
+      "loss": 0.0324,
+      "step": 1896
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 5.84393063583815e-05,
+      "loss": 0.0401,
+      "step": 1897
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 5.8265895953757215e-05,
+      "loss": 0.0318,
+      "step": 1898
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 5.8092485549132936e-05,
+      "loss": 0.0416,
+      "step": 1899
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 5.791907514450866e-05,
+      "loss": 0.0305,
+      "step": 1900
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 5.7745664739884384e-05,
+      "loss": 0.0404,
+      "step": 1901
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 5.7572254335260105e-05,
+      "loss": 0.0756,
+      "step": 1902
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 5.739884393063583e-05,
+      "loss": 0.0329,
+      "step": 1903
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 5.722543352601155e-05,
+      "loss": 0.0344,
+      "step": 1904
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 5.705202312138727e-05,
+      "loss": 0.0259,
+      "step": 1905
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 5.6878612716762994e-05,
+      "loss": 0.0342,
+      "step": 1906
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 5.670520231213872e-05,
+      "loss": 0.0425,
+      "step": 1907
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 5.653179190751444e-05,
+      "loss": 0.0411,
+      "step": 1908
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 5.635838150289016e-05,
+      "loss": 0.0291,
+      "step": 1909
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 5.618497109826589e-05,
+      "loss": 0.0399,
+      "step": 1910
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 5.601156069364161e-05,
+      "loss": 0.0319,
+      "step": 1911
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 5.583815028901733e-05,
+      "loss": 0.0314,
+      "step": 1912
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 5.566473988439306e-05,
+      "loss": 0.0389,
+      "step": 1913
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 5.549132947976878e-05,
+      "loss": 0.034,
+      "step": 1914
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 5.53179190751445e-05,
+      "loss": 0.0278,
+      "step": 1915
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 5.514450867052022e-05,
+      "loss": 0.0305,
+      "step": 1916
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 5.497109826589595e-05,
+      "loss": 0.0461,
+      "step": 1917
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 5.479768786127167e-05,
+      "loss": 0.0264,
+      "step": 1918
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 5.462427745664739e-05,
+      "loss": 0.0278,
+      "step": 1919
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 5.445086705202312e-05,
+      "loss": 0.0276,
+      "step": 1920
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 5.427745664739884e-05,
+      "loss": 0.0338,
+      "step": 1921
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 5.410404624277456e-05,
+      "loss": 0.028,
+      "step": 1922
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 5.393063583815028e-05,
+      "loss": 0.0304,
+      "step": 1923
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 5.375722543352601e-05,
+      "loss": 0.0405,
+      "step": 1924
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 5.358381502890173e-05,
+      "loss": 0.0265,
+      "step": 1925
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 5.341040462427745e-05,
+      "loss": 0.0289,
+      "step": 1926
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 5.3236994219653176e-05,
+      "loss": 0.0272,
+      "step": 1927
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 5.3063583815028896e-05,
+      "loss": 0.0278,
+      "step": 1928
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 5.289017341040462e-05,
+      "loss": 0.0364,
+      "step": 1929
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 5.271676300578034e-05,
+      "loss": 0.0183,
+      "step": 1930
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 5.2543352601156065e-05,
+      "loss": 0.03,
+      "step": 1931
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 5.2369942196531786e-05,
+      "loss": 0.0345,
+      "step": 1932
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 5.2196531791907507e-05,
+      "loss": 0.0346,
+      "step": 1933
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 5.2023121387283234e-05,
+      "loss": 0.0374,
+      "step": 1934
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 5.1849710982658955e-05,
+      "loss": 0.0673,
+      "step": 1935
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 5.1676300578034675e-05,
+      "loss": 0.0524,
+      "step": 1936
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 5.1502890173410396e-05,
+      "loss": 0.0417,
+      "step": 1937
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 5.1329479768786124e-05,
+      "loss": 0.0507,
+      "step": 1938
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 5.1156069364161844e-05,
+      "loss": 0.0441,
+      "step": 1939
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 5.0982658959537565e-05,
+      "loss": 0.0386,
+      "step": 1940
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 5.080924855491329e-05,
+      "loss": 0.0389,
+      "step": 1941
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 5.063583815028901e-05,
+      "loss": 0.0461,
+      "step": 1942
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 5.0462427745664734e-05,
+      "loss": 0.0384,
+      "step": 1943
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 5.028901734104046e-05,
+      "loss": 0.034,
+      "step": 1944
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 5.011560693641618e-05,
+      "loss": 0.029,
+      "step": 1945
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 4.99421965317919e-05,
+      "loss": 0.0438,
+      "step": 1946
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 4.976878612716762e-05,
+      "loss": 0.0402,
+      "step": 1947
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 4.959537572254335e-05,
+      "loss": 0.034,
+      "step": 1948
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 4.942196531791907e-05,
+      "loss": 0.0376,
+      "step": 1949
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 4.924855491329479e-05,
+      "loss": 0.03,
+      "step": 1950
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 4.907514450867052e-05,
+      "loss": 0.0375,
+      "step": 1951
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 4.890173410404624e-05,
+      "loss": 0.044,
+      "step": 1952
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 4.872832369942196e-05,
+      "loss": 0.0438,
+      "step": 1953
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 4.855491329479768e-05,
+      "loss": 0.0375,
+      "step": 1954
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 4.838150289017341e-05,
+      "loss": 0.0526,
+      "step": 1955
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 4.820809248554913e-05,
+      "loss": 0.0449,
+      "step": 1956
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 4.803468208092485e-05,
+      "loss": 0.0414,
+      "step": 1957
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 4.786127167630058e-05,
+      "loss": 0.031,
+      "step": 1958
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 4.76878612716763e-05,
+      "loss": 0.0402,
+      "step": 1959
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 4.751445086705202e-05,
+      "loss": 0.0268,
+      "step": 1960
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 4.734104046242774e-05,
+      "loss": 0.036,
+      "step": 1961
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 4.716763005780347e-05,
+      "loss": 0.0352,
+      "step": 1962
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 4.699421965317919e-05,
+      "loss": 0.0335,
+      "step": 1963
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 4.682080924855491e-05,
+      "loss": 0.0351,
+      "step": 1964
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 4.6647398843930636e-05,
+      "loss": 0.0366,
+      "step": 1965
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 4.647398843930636e-05,
+      "loss": 0.0244,
+      "step": 1966
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 4.630057803468208e-05,
+      "loss": 0.0418,
+      "step": 1967
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 4.61271676300578e-05,
+      "loss": 0.0453,
+      "step": 1968
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 4.5953757225433526e-05,
+      "loss": 0.0253,
+      "step": 1969
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 4.5780346820809246e-05,
+      "loss": 0.0302,
+      "step": 1970
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 4.560693641618497e-05,
+      "loss": 0.0309,
+      "step": 1971
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 4.5433526011560694e-05,
+      "loss": 0.0245,
+      "step": 1972
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 4.5260115606936415e-05,
+      "loss": 0.0167,
+      "step": 1973
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 4.5086705202312136e-05,
+      "loss": 0.0278,
+      "step": 1974
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 4.491329479768786e-05,
+      "loss": 0.0301,
+      "step": 1975
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 4.4739884393063584e-05,
+      "loss": 0.0281,
+      "step": 1976
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 4.4566473988439305e-05,
+      "loss": 0.0392,
+      "step": 1977
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 4.4393063583815025e-05,
+      "loss": 0.0225,
+      "step": 1978
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 4.421965317919075e-05,
+      "loss": 0.0263,
+      "step": 1979
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 4.404624277456647e-05,
+      "loss": 0.0203,
+      "step": 1980
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 4.3872832369942194e-05,
+      "loss": 0.0267,
+      "step": 1981
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 4.369942196531792e-05,
+      "loss": 0.032,
+      "step": 1982
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 4.352601156069364e-05,
+      "loss": 0.0291,
+      "step": 1983
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 4.335260115606936e-05,
+      "loss": 0.0374,
+      "step": 1984
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 4.3179190751445084e-05,
+      "loss": 0.0435,
+      "step": 1985
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 4.300578034682081e-05,
+      "loss": 0.0417,
+      "step": 1986
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 4.283236994219653e-05,
+      "loss": 0.05,
+      "step": 1987
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 4.265895953757225e-05,
+      "loss": 0.0402,
+      "step": 1988
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 4.248554913294798e-05,
+      "loss": 0.0487,
+      "step": 1989
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 4.23121387283237e-05,
+      "loss": 0.0366,
+      "step": 1990
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 4.213872832369942e-05,
+      "loss": 0.0349,
+      "step": 1991
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 4.196531791907514e-05,
+      "loss": 0.0305,
+      "step": 1992
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 4.179190751445087e-05,
+      "loss": 0.0443,
+      "step": 1993
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 4.161849710982658e-05,
+      "loss": 0.0418,
+      "step": 1994
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 4.1445086705202304e-05,
+      "loss": 0.03,
+      "step": 1995
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 4.1271676300578025e-05,
+      "loss": 0.0308,
+      "step": 1996
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 4.109826589595375e-05,
+      "loss": 0.0293,
+      "step": 1997
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 4.092485549132947e-05,
+      "loss": 0.0419,
+      "step": 1998
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 4.075144508670519e-05,
+      "loss": 0.0258,
+      "step": 1999
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 4.057803468208092e-05,
+      "loss": 0.0291,
+      "step": 2000
+    },
+    {
+      "epoch": 8.97,
+      "eval_loss": 0.3589690625667572,
+      "eval_runtime": 570.6282,
+      "eval_samples_per_second": 4.63,
+      "eval_steps_per_second": 0.58,
+      "eval_wer": 0.09641015470051567,
+      "step": 2000
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 4.040462427745664e-05,
+      "loss": 0.0345,
+      "step": 2001
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 4.023121387283236e-05,
+      "loss": 0.0343,
+      "step": 2002
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 4.005780346820808e-05,
+      "loss": 0.0374,
+      "step": 2003
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 3.988439306358381e-05,
+      "loss": 0.0244,
+      "step": 2004
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 3.971098265895953e-05,
+      "loss": 0.0371,
+      "step": 2005
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 3.953757225433525e-05,
+      "loss": 0.0275,
+      "step": 2006
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 3.936416184971098e-05,
+      "loss": 0.0245,
+      "step": 2007
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 3.91907514450867e-05,
+      "loss": 0.057,
+      "step": 2008
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 3.901734104046242e-05,
+      "loss": 0.0298,
+      "step": 2009
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 3.884393063583814e-05,
+      "loss": 0.0394,
+      "step": 2010
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 3.867052023121387e-05,
+      "loss": 0.0377,
+      "step": 2011
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 3.849710982658959e-05,
+      "loss": 0.0317,
+      "step": 2012
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 3.832369942196531e-05,
+      "loss": 0.0363,
+      "step": 2013
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 3.815028901734104e-05,
+      "loss": 0.0391,
+      "step": 2014
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 3.797687861271676e-05,
+      "loss": 0.0324,
+      "step": 2015
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 3.780346820809248e-05,
+      "loss": 0.042,
+      "step": 2016
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 3.76300578034682e-05,
+      "loss": 0.029,
+      "step": 2017
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 3.745664739884393e-05,
+      "loss": 0.0427,
+      "step": 2018
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 3.728323699421965e-05,
+      "loss": 0.0366,
+      "step": 2019
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 3.710982658959537e-05,
+      "loss": 0.0306,
+      "step": 2020
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 3.6936416184971096e-05,
+      "loss": 0.0419,
+      "step": 2021
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 3.6763005780346816e-05,
+      "loss": 0.0289,
+      "step": 2022
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 3.658959537572254e-05,
+      "loss": 0.033,
+      "step": 2023
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 3.6416184971098265e-05,
+      "loss": 0.0274,
+      "step": 2024
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 3.6242774566473985e-05,
+      "loss": 0.2688,
+      "step": 2025
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 3.6069364161849706e-05,
+      "loss": 0.0241,
+      "step": 2026
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 3.5895953757225427e-05,
+      "loss": 0.0307,
+      "step": 2027
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 3.5722543352601154e-05,
+      "loss": 0.0317,
+      "step": 2028
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 3.5549132947976875e-05,
+      "loss": 0.0328,
+      "step": 2029
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 3.5375722543352595e-05,
+      "loss": 0.0291,
+      "step": 2030
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 3.520231213872832e-05,
+      "loss": 0.0224,
+      "step": 2031
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 3.5028901734104043e-05,
+      "loss": 0.0383,
+      "step": 2032
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 3.4855491329479764e-05,
+      "loss": 0.0319,
+      "step": 2033
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 3.4682080924855485e-05,
+      "loss": 0.0283,
+      "step": 2034
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 3.450867052023121e-05,
+      "loss": 0.0184,
+      "step": 2035
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 3.433526011560693e-05,
+      "loss": 0.0273,
+      "step": 2036
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 3.4161849710982654e-05,
+      "loss": 0.0288,
+      "step": 2037
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 3.398843930635838e-05,
+      "loss": 0.0342,
+      "step": 2038
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 3.38150289017341e-05,
+      "loss": 0.0355,
+      "step": 2039
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 3.364161849710982e-05,
+      "loss": 0.0316,
+      "step": 2040
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 3.346820809248554e-05,
+      "loss": 0.0263,
+      "step": 2041
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 3.329479768786127e-05,
+      "loss": 0.0373,
+      "step": 2042
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 3.312138728323699e-05,
+      "loss": 0.0224,
+      "step": 2043
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 3.294797687861271e-05,
+      "loss": 0.029,
+      "step": 2044
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 3.277456647398844e-05,
+      "loss": 0.0298,
+      "step": 2045
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 3.260115606936416e-05,
+      "loss": 0.0223,
+      "step": 2046
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 3.242774566473988e-05,
+      "loss": 0.0351,
+      "step": 2047
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 3.22543352601156e-05,
+      "loss": 0.023,
+      "step": 2048
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 3.208092485549133e-05,
+      "loss": 0.0231,
+      "step": 2049
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 3.190751445086705e-05,
+      "loss": 0.022,
+      "step": 2050
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 3.173410404624277e-05,
+      "loss": 0.0329,
+      "step": 2051
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 3.15606936416185e-05,
+      "loss": 0.0294,
+      "step": 2052
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 3.138728323699422e-05,
+      "loss": 0.0151,
+      "step": 2053
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 3.121387283236994e-05,
+      "loss": 0.0167,
+      "step": 2054
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 3.1040462427745667e-05,
+      "loss": 0.0182,
+      "step": 2055
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 3.086705202312139e-05,
+      "loss": 0.0431,
+      "step": 2056
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 3.069364161849711e-05,
+      "loss": 0.0711,
+      "step": 2057
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 3.052023121387283e-05,
+      "loss": 0.0487,
+      "step": 2058
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 3.0346820809248553e-05,
+      "loss": 0.0368,
+      "step": 2059
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 3.0173410404624277e-05,
+      "loss": 0.0333,
+      "step": 2060
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 0.0409,
+      "step": 2061
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 2.982658959537572e-05,
+      "loss": 0.038,
+      "step": 2062
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 2.9653179190751446e-05,
+      "loss": 0.0301,
+      "step": 2063
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 2.9479768786127166e-05,
+      "loss": 0.0407,
+      "step": 2064
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 2.930635838150289e-05,
+      "loss": 0.0388,
+      "step": 2065
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 2.9132947976878608e-05,
+      "loss": 0.0413,
+      "step": 2066
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 2.895953757225433e-05,
+      "loss": 0.0317,
+      "step": 2067
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 2.8786127167630052e-05,
+      "loss": 0.0397,
+      "step": 2068
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 2.8612716763005776e-05,
+      "loss": 0.0456,
+      "step": 2069
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 2.8439306358381497e-05,
+      "loss": 0.0348,
+      "step": 2070
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 2.826589595375722e-05,
+      "loss": 0.0273,
+      "step": 2071
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 2.8092485549132945e-05,
+      "loss": 0.0381,
+      "step": 2072
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 2.7919075144508666e-05,
+      "loss": 0.0338,
+      "step": 2073
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 2.774566473988439e-05,
+      "loss": 0.0308,
+      "step": 2074
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 2.757225433526011e-05,
+      "loss": 0.0325,
+      "step": 2075
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 2.7398843930635835e-05,
+      "loss": 0.0207,
+      "step": 2076
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 2.722543352601156e-05,
+      "loss": 0.0315,
+      "step": 2077
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 2.705202312138728e-05,
+      "loss": 0.0329,
+      "step": 2078
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 2.6878612716763003e-05,
+      "loss": 0.0331,
+      "step": 2079
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 2.6705202312138724e-05,
+      "loss": 0.0251,
+      "step": 2080
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 2.6531791907514448e-05,
+      "loss": 0.0311,
+      "step": 2081
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 2.635838150289017e-05,
+      "loss": 0.0378,
+      "step": 2082
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 2.6184971098265893e-05,
+      "loss": 0.0282,
+      "step": 2083
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 2.6011560693641617e-05,
+      "loss": 0.0314,
+      "step": 2084
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 2.5838150289017338e-05,
+      "loss": 0.0311,
+      "step": 2085
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 2.5664739884393062e-05,
+      "loss": 0.034,
+      "step": 2086
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 2.5491329479768782e-05,
+      "loss": 0.0324,
+      "step": 2087
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 2.5317919075144507e-05,
+      "loss": 0.0358,
+      "step": 2088
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 2.514450867052023e-05,
+      "loss": 0.0237,
+      "step": 2089
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 2.497109826589595e-05,
+      "loss": 0.0342,
+      "step": 2090
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 2.4797687861271675e-05,
+      "loss": 0.0208,
+      "step": 2091
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 2.4624277456647396e-05,
+      "loss": 0.0299,
+      "step": 2092
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 2.445086705202312e-05,
+      "loss": 0.0341,
+      "step": 2093
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 2.427745664739884e-05,
+      "loss": 0.0236,
+      "step": 2094
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 2.4104046242774565e-05,
+      "loss": 0.0257,
+      "step": 2095
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 2.393063583815029e-05,
+      "loss": 0.0215,
+      "step": 2096
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 2.375722543352601e-05,
+      "loss": 0.0304,
+      "step": 2097
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 2.3583815028901734e-05,
+      "loss": 0.0249,
+      "step": 2098
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 2.3410404624277454e-05,
+      "loss": 0.0361,
+      "step": 2099
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 2.323699421965318e-05,
+      "loss": 0.0242,
+      "step": 2100
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 2.30635838150289e-05,
+      "loss": 0.0198,
+      "step": 2101
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 2.2890173410404623e-05,
+      "loss": 0.0237,
+      "step": 2102
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 2.2716763005780347e-05,
+      "loss": 0.0177,
+      "step": 2103
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 2.2543352601156068e-05,
+      "loss": 0.0389,
+      "step": 2104
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 2.2369942196531792e-05,
+      "loss": 0.0212,
+      "step": 2105
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 2.2196531791907513e-05,
+      "loss": 0.0244,
+      "step": 2106
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 2.2023121387283237e-05,
+      "loss": 0.0884,
+      "step": 2107
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 2.184971098265896e-05,
+      "loss": 0.0482,
+      "step": 2108
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 2.167630057803468e-05,
+      "loss": 0.0422,
+      "step": 2109
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 2.1502890173410405e-05,
+      "loss": 0.0355,
+      "step": 2110
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 2.1329479768786126e-05,
+      "loss": 0.0358,
+      "step": 2111
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 2.115606936416185e-05,
+      "loss": 0.0419,
+      "step": 2112
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 2.098265895953757e-05,
+      "loss": 0.0411,
+      "step": 2113
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 2.080924855491329e-05,
+      "loss": 0.0319,
+      "step": 2114
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 2.0635838150289012e-05,
+      "loss": 0.0288,
+      "step": 2115
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 2.0462427745664736e-05,
+      "loss": 0.0269,
+      "step": 2116
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 2.028901734104046e-05,
+      "loss": 0.0355,
+      "step": 2117
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 2.011560693641618e-05,
+      "loss": 0.032,
+      "step": 2118
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 1.9942196531791905e-05,
+      "loss": 0.0251,
+      "step": 2119
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 1.9768786127167626e-05,
+      "loss": 0.0295,
+      "step": 2120
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 1.959537572254335e-05,
+      "loss": 0.0405,
+      "step": 2121
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 1.942196531791907e-05,
+      "loss": 0.0289,
+      "step": 2122
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 1.9248554913294795e-05,
+      "loss": 0.0337,
+      "step": 2123
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 1.907514450867052e-05,
+      "loss": 0.035,
+      "step": 2124
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 1.890173410404624e-05,
+      "loss": 0.029,
+      "step": 2125
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 1.8728323699421963e-05,
+      "loss": 0.03,
+      "step": 2126
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 1.8554913294797684e-05,
+      "loss": 0.0287,
+      "step": 2127
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 1.8381502890173408e-05,
+      "loss": 0.0354,
+      "step": 2128
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 1.8208092485549132e-05,
+      "loss": 0.0262,
+      "step": 2129
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 1.8034682080924853e-05,
+      "loss": 0.0225,
+      "step": 2130
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 1.7861271676300577e-05,
+      "loss": 0.0288,
+      "step": 2131
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 1.7687861271676298e-05,
+      "loss": 0.0241,
+      "step": 2132
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 1.7514450867052022e-05,
+      "loss": 0.0277,
+      "step": 2133
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 1.7341040462427742e-05,
+      "loss": 0.0298,
+      "step": 2134
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 1.7167630057803466e-05,
+      "loss": 0.0278,
+      "step": 2135
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 1.699421965317919e-05,
+      "loss": 0.0253,
+      "step": 2136
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 1.682080924855491e-05,
+      "loss": 0.0331,
+      "step": 2137
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 1.6647398843930635e-05,
+      "loss": 0.0272,
+      "step": 2138
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 1.6473988439306356e-05,
+      "loss": 0.0237,
+      "step": 2139
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 1.630057803468208e-05,
+      "loss": 0.027,
+      "step": 2140
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 1.61271676300578e-05,
+      "loss": 0.026,
+      "step": 2141
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 1.5953757225433525e-05,
+      "loss": 0.0281,
+      "step": 2142
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 1.578034682080925e-05,
+      "loss": 0.0283,
+      "step": 2143
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 1.560693641618497e-05,
+      "loss": 0.0188,
+      "step": 2144
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 1.5433526011560694e-05,
+      "loss": 0.0235,
+      "step": 2145
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 1.5260115606936414e-05,
+      "loss": 0.0257,
+      "step": 2146
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 1.5086705202312138e-05,
+      "loss": 0.0233,
+      "step": 2147
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 1.491329479768786e-05,
+      "loss": 0.0331,
+      "step": 2148
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 1.4739884393063583e-05,
+      "loss": 0.0364,
+      "step": 2149
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 1.4566473988439304e-05,
+      "loss": 0.0312,
+      "step": 2150
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 1.4393063583815026e-05,
+      "loss": 0.022,
+      "step": 2151
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 1.4219653179190749e-05,
+      "loss": 0.0338,
+      "step": 2152
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 1.4046242774566473e-05,
+      "loss": 0.0279,
+      "step": 2153
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 1.3872832369942195e-05,
+      "loss": 0.0309,
+      "step": 2154
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 1.3699421965317917e-05,
+      "loss": 0.056,
+      "step": 2155
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 1.352601156069364e-05,
+      "loss": 0.0299,
+      "step": 2156
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 1.3352601156069362e-05,
+      "loss": 0.056,
+      "step": 2157
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 1.3179190751445084e-05,
+      "loss": 0.0458,
+      "step": 2158
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 1.3005780346820809e-05,
+      "loss": 0.0395,
+      "step": 2159
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 1.2832369942196531e-05,
+      "loss": 0.0401,
+      "step": 2160
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 1.2658959537572253e-05,
+      "loss": 0.0302,
+      "step": 2161
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 1.2485549132947976e-05,
+      "loss": 0.037,
+      "step": 2162
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 1.2312138728323698e-05,
+      "loss": 0.0276,
+      "step": 2163
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 1.213872832369942e-05,
+      "loss": 0.0316,
+      "step": 2164
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 1.1965317919075144e-05,
+      "loss": 0.0344,
+      "step": 2165
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 1.1791907514450867e-05,
+      "loss": 0.0365,
+      "step": 2166
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 1.161849710982659e-05,
+      "loss": 0.0248,
+      "step": 2167
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 1.1445086705202312e-05,
+      "loss": 0.027,
+      "step": 2168
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 1.1271676300578034e-05,
+      "loss": 0.0345,
+      "step": 2169
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 1.1098265895953756e-05,
+      "loss": 0.0264,
+      "step": 2170
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 1.092485549132948e-05,
+      "loss": 0.0279,
+      "step": 2171
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 1.0751445086705203e-05,
+      "loss": 0.0295,
+      "step": 2172
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 1.0578034682080925e-05,
+      "loss": 0.034,
+      "step": 2173
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 1.0404624277456646e-05,
+      "loss": 0.0232,
+      "step": 2174
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 1.0231213872832368e-05,
+      "loss": 0.028,
+      "step": 2175
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 1.005780346820809e-05,
+      "loss": 0.0351,
+      "step": 2176
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 9.884393063583813e-06,
+      "loss": 0.0379,
+      "step": 2177
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 9.710982658959535e-06,
+      "loss": 0.0293,
+      "step": 2178
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 9.53757225433526e-06,
+      "loss": 0.0337,
+      "step": 2179
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 9.364161849710982e-06,
+      "loss": 0.0362,
+      "step": 2180
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 9.190751445086704e-06,
+      "loss": 0.0274,
+      "step": 2181
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 9.017341040462426e-06,
+      "loss": 0.0325,
+      "step": 2182
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 8.843930635838149e-06,
+      "loss": 0.0346,
+      "step": 2183
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 8.670520231213871e-06,
+      "loss": 0.0263,
+      "step": 2184
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 8.497109826589595e-06,
+      "loss": 0.035,
+      "step": 2185
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 8.323699421965318e-06,
+      "loss": 0.0327,
+      "step": 2186
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 8.15028901734104e-06,
+      "loss": 0.0339,
+      "step": 2187
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 7.976878612716762e-06,
+      "loss": 0.0233,
+      "step": 2188
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 7.803468208092485e-06,
+      "loss": 0.028,
+      "step": 2189
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 7.630057803468207e-06,
+      "loss": 0.0294,
+      "step": 2190
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 7.45664739884393e-06,
+      "loss": 0.0224,
+      "step": 2191
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 7.283236994219652e-06,
+      "loss": 0.0323,
+      "step": 2192
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 7.109826589595374e-06,
+      "loss": 0.0337,
+      "step": 2193
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 6.9364161849710975e-06,
+      "loss": 0.0313,
+      "step": 2194
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 6.76300578034682e-06,
+      "loss": 0.013,
+      "step": 2195
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 6.589595375722542e-06,
+      "loss": 0.0371,
+      "step": 2196
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 6.4161849710982654e-06,
+      "loss": 0.0245,
+      "step": 2197
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 6.242774566473988e-06,
+      "loss": 0.0334,
+      "step": 2198
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 6.06936416184971e-06,
+      "loss": 0.022,
+      "step": 2199
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 5.895953757225433e-06,
+      "loss": 0.0241,
+      "step": 2200
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 5.722543352601156e-06,
+      "loss": 0.0273,
+      "step": 2201
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 5.549132947976878e-06,
+      "loss": 0.0196,
+      "step": 2202
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 5.375722543352601e-06,
+      "loss": 0.026,
+      "step": 2203
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 5.202312138728323e-06,
+      "loss": 0.0247,
+      "step": 2204
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 5.028901734104045e-06,
+      "loss": 0.0185,
+      "step": 2205
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 4.855491329479768e-06,
+      "loss": 0.0272,
+      "step": 2206
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 4.682080924855491e-06,
+      "loss": 0.0398,
+      "step": 2207
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 4.508670520231213e-06,
+      "loss": 0.0464,
+      "step": 2208
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 4.335260115606936e-06,
+      "loss": 0.0335,
+      "step": 2209
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 4.161849710982659e-06,
+      "loss": 0.0381,
+      "step": 2210
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 3.988439306358381e-06,
+      "loss": 0.042,
+      "step": 2211
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 3.8150289017341036e-06,
+      "loss": 0.035,
+      "step": 2212
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 3.641618497109826e-06,
+      "loss": 0.0309,
+      "step": 2213
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 3.4682080924855487e-06,
+      "loss": 0.0347,
+      "step": 2214
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 3.294797687861271e-06,
+      "loss": 0.0296,
+      "step": 2215
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 3.121387283236994e-06,
+      "loss": 0.0353,
+      "step": 2216
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 2.9479768786127167e-06,
+      "loss": 0.0381,
+      "step": 2217
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 2.774566473988439e-06,
+      "loss": 0.0363,
+      "step": 2218
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 2.6011560693641614e-06,
+      "loss": 0.0239,
+      "step": 2219
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 2.427745664739884e-06,
+      "loss": 0.0394,
+      "step": 2220
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 2.2543352601156066e-06,
+      "loss": 0.0335,
+      "step": 2221
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 2.0809248554913294e-06,
+      "loss": 0.0213,
+      "step": 2222
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 1.9075144508670518e-06,
+      "loss": 0.0287,
+      "step": 2223
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 1.7341040462427744e-06,
+      "loss": 0.034,
+      "step": 2224
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 1.560693641618497e-06,
+      "loss": 0.0296,
+      "step": 2225
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 1.3872832369942195e-06,
+      "loss": 0.0192,
+      "step": 2226
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 1.213872832369942e-06,
+      "loss": 0.0272,
+      "step": 2227
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 1.0404624277456647e-06,
+      "loss": 0.0126,
+      "step": 2228
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 8.670520231213872e-07,
+      "loss": 0.0346,
+      "step": 2229
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 6.936416184971098e-07,
+      "loss": 0.0134,
+      "step": 2230
+    },
+    {
+      "epoch": 10.0,
+      "step": 2230,
       "total_flos": 0.0,
-      "train_loss": 2.498668275362574,
-      "train_runtime": 25980.5507,
-      "train_samples_per_second": 5.492,
+      "train_loss": 1.290548439615828,
+      "train_runtime": 51756.4072,
+      "train_samples_per_second": 5.514,
       "train_steps_per_second": 0.043
     }
   ],
-  "max_steps": 1115,
-  "num_train_epochs": 5,
+  "max_steps": 2230,
+  "num_train_epochs": 10,
   "total_flos": 0.0,
   "trial_name": null,
   "trial_params": null