diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -10,1945 +10,1945 @@
     {
       "epoch": 0.0,
       "learning_rate": 0.0,
-      "loss": 10.1925,
+      "loss": 8.7584,
       "step": 1
     },
     {
       "epoch": 0.0,
       "learning_rate": 0.0,
-      "loss": 10.3094,
+      "loss": 8.5833,
       "step": 2
     },
     {
       "epoch": 0.01,
       "learning_rate": 6e-07,
-      "loss": 10.0606,
+      "loss": 8.5559,
       "step": 3
     },
     {
       "epoch": 0.01,
-      "learning_rate": 6e-07,
-      "loss": 10.1729,
+      "learning_rate": 1.2e-06,
+      "loss": 8.4758,
       "step": 4
     },
     {
       "epoch": 0.01,
-      "learning_rate": 1.2e-06,
-      "loss": 10.0768,
+      "learning_rate": 1.8e-06,
+      "loss": 8.6015,
       "step": 5
     },
     {
       "epoch": 0.01,
-      "learning_rate": 1.8e-06,
-      "loss": 10.0394,
+      "learning_rate": 2.4e-06,
+      "loss": 8.402,
       "step": 6
     },
     {
       "epoch": 0.02,
-      "learning_rate": 2.4e-06,
-      "loss": 10.1223,
+      "learning_rate": 2.9999999999999997e-06,
+      "loss": 8.3297,
       "step": 7
     },
     {
       "epoch": 0.02,
-      "learning_rate": 2.9999999999999997e-06,
-      "loss": 9.4767,
+      "learning_rate": 3.6e-06,
+      "loss": 7.9299,
       "step": 8
     },
     {
       "epoch": 0.02,
-      "learning_rate": 3.6e-06,
-      "loss": 9.0957,
+      "learning_rate": 4.2e-06,
+      "loss": 7.6735,
       "step": 9
     },
     {
       "epoch": 0.02,
-      "learning_rate": 4.2e-06,
-      "loss": 8.5185,
+      "learning_rate": 4.8e-06,
+      "loss": 7.5134,
       "step": 10
     },
     {
       "epoch": 0.02,
-      "learning_rate": 4.8e-06,
-      "loss": 8.4451,
+      "learning_rate": 5.399999999999999e-06,
+      "loss": 7.3091,
       "step": 11
     },
     {
       "epoch": 0.03,
-      "learning_rate": 5.399999999999999e-06,
-      "loss": 7.9992,
+      "learning_rate": 5.999999999999999e-06,
+      "loss": 7.1164,
       "step": 12
     },
     {
       "epoch": 0.03,
-      "learning_rate": 5.999999999999999e-06,
-      "loss": 7.7352,
+      "learning_rate": 6.599999999999999e-06,
+      "loss": 6.7007,
       "step": 13
     },
     {
       "epoch": 0.03,
-      "learning_rate": 6.599999999999999e-06,
-      "loss": 7.5244,
+      "learning_rate": 7.2e-06,
+      "loss": 6.6293,
       "step": 14
     },
     {
       "epoch": 0.03,
-      "learning_rate": 7.2e-06,
-      "loss": 7.2125,
+      "learning_rate": 7.799999999999998e-06,
+      "loss": 6.3926,
       "step": 15
     },
     {
       "epoch": 0.04,
-      "learning_rate": 7.799999999999998e-06,
-      "loss": 7.0344,
+      "learning_rate": 8.4e-06,
+      "loss": 6.3405,
       "step": 16
     },
     {
       "epoch": 0.04,
-      "learning_rate": 8.4e-06,
-      "loss": 6.8173,
+      "learning_rate": 8.999999999999999e-06,
+      "loss": 6.1208,
       "step": 17
     },
     {
       "epoch": 0.04,
-      "learning_rate": 8.999999999999999e-06,
-      "loss": 6.7418,
+      "learning_rate": 9.6e-06,
+      "loss": 6.0531,
       "step": 18
     },
     {
       "epoch": 0.04,
-      "learning_rate": 9.6e-06,
-      "loss": 6.3218,
+      "learning_rate": 1.02e-05,
+      "loss": 5.7463,
       "step": 19
     },
     {
       "epoch": 0.04,
-      "learning_rate": 1.02e-05,
-      "loss": 6.3455,
+      "learning_rate": 1.0799999999999998e-05,
+      "loss": 5.8868,
       "step": 20
     },
     {
       "epoch": 0.05,
-      "learning_rate": 1.0799999999999998e-05,
-      "loss": 6.1011,
+      "learning_rate": 1.14e-05,
+      "loss": 5.6696,
       "step": 21
     },
     {
       "epoch": 0.05,
-      "learning_rate": 1.14e-05,
-      "loss": 6.0982,
+      "learning_rate": 1.1999999999999999e-05,
+      "loss": 5.6137,
       "step": 22
     },
     {
       "epoch": 0.05,
-      "learning_rate": 1.1999999999999999e-05,
-      "loss": 5.9207,
+      "learning_rate": 1.26e-05,
+      "loss": 5.4833,
       "step": 23
     },
     {
       "epoch": 0.05,
-      "learning_rate": 1.26e-05,
-      "loss": 5.8474,
+      "learning_rate": 1.3199999999999997e-05,
+      "loss": 5.4312,
       "step": 24
     },
     {
       "epoch": 0.06,
-      "learning_rate": 1.3199999999999997e-05,
-      "loss": 5.5599,
+      "learning_rate": 1.3799999999999998e-05,
+      "loss": 5.1508,
       "step": 25
     },
     {
       "epoch": 0.06,
-      "learning_rate": 1.3799999999999998e-05,
-      "loss": 5.7598,
+      "learning_rate": 1.44e-05,
+      "loss": 5.4145,
       "step": 26
     },
     {
       "epoch": 0.06,
-      "learning_rate": 1.44e-05,
-      "loss": 5.5277,
+      "learning_rate": 1.4999999999999999e-05,
+      "loss": 5.1659,
       "step": 27
     },
     {
       "epoch": 0.06,
-      "learning_rate": 1.4999999999999999e-05,
-      "loss": 5.2876,
+      "learning_rate": 1.5599999999999996e-05,
+      "loss": 5.0136,
       "step": 28
     },
     {
       "epoch": 0.07,
-      "learning_rate": 1.5599999999999996e-05,
-      "loss": 5.1947,
+      "learning_rate": 1.6199999999999997e-05,
+      "loss": 4.967,
       "step": 29
     },
     {
       "epoch": 0.07,
-      "learning_rate": 1.6199999999999997e-05,
-      "loss": 5.3315,
+      "learning_rate": 1.68e-05,
+      "loss": 5.1574,
       "step": 30
     },
     {
       "epoch": 0.07,
-      "learning_rate": 1.68e-05,
-      "loss": 5.2972,
+      "learning_rate": 1.74e-05,
+      "loss": 5.1364,
       "step": 31
     },
     {
       "epoch": 0.07,
-      "learning_rate": 1.74e-05,
-      "loss": 5.059,
+      "learning_rate": 1.7999999999999997e-05,
+      "loss": 4.9412,
       "step": 32
     },
     {
       "epoch": 0.07,
-      "learning_rate": 1.7999999999999997e-05,
-      "loss": 5.1491,
+      "learning_rate": 1.8599999999999998e-05,
+      "loss": 5.039,
       "step": 33
     },
     {
       "epoch": 0.08,
-      "learning_rate": 1.8599999999999998e-05,
-      "loss": 5.0233,
+      "learning_rate": 1.92e-05,
+      "loss": 4.9206,
       "step": 34
     },
     {
       "epoch": 0.08,
-      "learning_rate": 1.92e-05,
-      "loss": 5.0224,
+      "learning_rate": 1.98e-05,
+      "loss": 4.93,
       "step": 35
     },
     {
       "epoch": 0.08,
-      "learning_rate": 1.98e-05,
-      "loss": 4.9906,
+      "learning_rate": 2.04e-05,
+      "loss": 4.8945,
       "step": 36
     },
     {
       "epoch": 0.08,
-      "learning_rate": 2.04e-05,
-      "loss": 5.091,
+      "learning_rate": 2.1e-05,
+      "loss": 5.0179,
       "step": 37
     },
     {
       "epoch": 0.09,
-      "learning_rate": 2.1e-05,
-      "loss": 4.9763,
+      "learning_rate": 2.1599999999999996e-05,
+      "loss": 4.918,
       "step": 38
     },
     {
       "epoch": 0.09,
-      "learning_rate": 2.1599999999999996e-05,
-      "loss": 5.0571,
+      "learning_rate": 2.2199999999999998e-05,
+      "loss": 5.0055,
       "step": 39
     },
     {
       "epoch": 0.09,
-      "learning_rate": 2.2199999999999998e-05,
-      "loss": 4.9133,
+      "learning_rate": 2.28e-05,
+      "loss": 4.8467,
       "step": 40
     },
     {
       "epoch": 0.09,
-      "learning_rate": 2.28e-05,
-      "loss": 4.9967,
+      "learning_rate": 2.34e-05,
+      "loss": 4.9421,
       "step": 41
     },
     {
       "epoch": 0.09,
-      "learning_rate": 2.34e-05,
-      "loss": 4.9993,
+      "learning_rate": 2.3999999999999997e-05,
+      "loss": 4.9595,
       "step": 42
     },
     {
       "epoch": 0.1,
-      "learning_rate": 2.3999999999999997e-05,
-      "loss": 4.7572,
+      "learning_rate": 2.4599999999999998e-05,
+      "loss": 4.6993,
       "step": 43
     },
     {
       "epoch": 0.1,
-      "learning_rate": 2.4599999999999998e-05,
-      "loss": 4.9231,
+      "learning_rate": 2.52e-05,
+      "loss": 4.8895,
       "step": 44
     },
     {
       "epoch": 0.1,
-      "learning_rate": 2.52e-05,
-      "loss": 5.0442,
+      "learning_rate": 2.5799999999999997e-05,
+      "loss": 4.9856,
       "step": 45
     },
     {
       "epoch": 0.1,
-      "learning_rate": 2.5799999999999997e-05,
-      "loss": 5.0736,
+      "learning_rate": 2.6399999999999995e-05,
+      "loss": 4.9902,
       "step": 46
     },
     {
       "epoch": 0.11,
-      "learning_rate": 2.6399999999999995e-05,
-      "loss": 4.978,
+      "learning_rate": 2.6999999999999996e-05,
+      "loss": 4.9278,
       "step": 47
     },
     {
       "epoch": 0.11,
-      "learning_rate": 2.6999999999999996e-05,
-      "loss": 5.0245,
+      "learning_rate": 2.7599999999999997e-05,
+      "loss": 4.9454,
       "step": 48
     },
     {
       "epoch": 0.11,
-      "learning_rate": 2.7599999999999997e-05,
-      "loss": 5.0658,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 4.9972,
       "step": 49
     },
     {
       "epoch": 0.11,
-      "learning_rate": 2.8199999999999998e-05,
-      "loss": 4.7624,
+      "learning_rate": 2.88e-05,
+      "loss": 4.6526,
       "step": 50
     },
     {
       "epoch": 0.11,
-      "learning_rate": 2.88e-05,
-      "loss": 5.9032,
+      "learning_rate": 2.94e-05,
+      "loss": 5.8451,
       "step": 51
     },
     {
       "epoch": 0.12,
-      "learning_rate": 2.94e-05,
-      "loss": 5.7495,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 5.7033,
       "step": 52
     },
     {
       "epoch": 0.12,
-      "learning_rate": 2.9999999999999997e-05,
-      "loss": 5.7386,
+      "learning_rate": 3.06e-05,
+      "loss": 5.6843,
       "step": 53
     },
     {
       "epoch": 0.12,
-      "learning_rate": 3.06e-05,
-      "loss": 5.2808,
+      "learning_rate": 3.119999999999999e-05,
+      "loss": 5.247,
       "step": 54
     },
     {
       "epoch": 0.12,
-      "learning_rate": 3.119999999999999e-05,
-      "loss": 4.8594,
+      "learning_rate": 3.1799999999999994e-05,
+      "loss": 4.7961,
       "step": 55
     },
     {
       "epoch": 0.13,
-      "learning_rate": 3.1799999999999994e-05,
-      "loss": 4.995,
+      "learning_rate": 3.2399999999999995e-05,
+      "loss": 4.9426,
       "step": 56
     },
     {
       "epoch": 0.13,
-      "learning_rate": 3.2399999999999995e-05,
-      "loss": 5.0311,
+      "learning_rate": 3.2999999999999996e-05,
+      "loss": 4.9681,
       "step": 57
     },
     {
       "epoch": 0.13,
-      "learning_rate": 3.2999999999999996e-05,
-      "loss": 5.0366,
+      "learning_rate": 3.36e-05,
+      "loss": 4.9621,
       "step": 58
     },
     {
       "epoch": 0.13,
-      "learning_rate": 3.36e-05,
-      "loss": 4.9435,
+      "learning_rate": 3.42e-05,
+      "loss": 4.9065,
       "step": 59
     },
     {
       "epoch": 0.13,
-      "learning_rate": 3.42e-05,
-      "loss": 4.8193,
+      "learning_rate": 3.48e-05,
+      "loss": 4.7891,
       "step": 60
     },
     {
       "epoch": 0.14,
-      "learning_rate": 3.48e-05,
-      "loss": 4.8357,
+      "learning_rate": 3.539999999999999e-05,
+      "loss": 4.8069,
       "step": 61
     },
     {
       "epoch": 0.14,
-      "learning_rate": 3.539999999999999e-05,
-      "loss": 4.8625,
+      "learning_rate": 3.5999999999999994e-05,
+      "loss": 4.8273,
       "step": 62
     },
     {
       "epoch": 0.14,
-      "learning_rate": 3.5999999999999994e-05,
-      "loss": 4.7677,
+      "learning_rate": 3.6599999999999995e-05,
+      "loss": 4.7276,
       "step": 63
     },
     {
       "epoch": 0.14,
-      "learning_rate": 3.6599999999999995e-05,
-      "loss": 4.8096,
+      "learning_rate": 3.7199999999999996e-05,
+      "loss": 4.7894,
       "step": 64
     },
     {
       "epoch": 0.15,
-      "learning_rate": 3.7199999999999996e-05,
-      "loss": 4.7863,
+      "learning_rate": 3.78e-05,
+      "loss": 4.7477,
       "step": 65
     },
     {
       "epoch": 0.15,
-      "learning_rate": 3.78e-05,
-      "loss": 4.6589,
+      "learning_rate": 3.84e-05,
+      "loss": 4.6233,
       "step": 66
     },
     {
       "epoch": 0.15,
-      "learning_rate": 3.84e-05,
-      "loss": 4.7568,
+      "learning_rate": 3.9e-05,
+      "loss": 4.7465,
       "step": 67
     },
     {
       "epoch": 0.15,
-      "learning_rate": 3.9e-05,
-      "loss": 4.7222,
+      "learning_rate": 3.96e-05,
+      "loss": 4.7111,
       "step": 68
     },
     {
       "epoch": 0.15,
-      "learning_rate": 3.96e-05,
-      "loss": 4.7684,
+      "learning_rate": 4.02e-05,
+      "loss": 4.7447,
       "step": 69
     },
     {
       "epoch": 0.16,
-      "learning_rate": 4.02e-05,
-      "loss": 4.6798,
+      "learning_rate": 4.08e-05,
+      "loss": 4.6527,
       "step": 70
     },
     {
       "epoch": 0.16,
-      "learning_rate": 4.08e-05,
-      "loss": 4.8219,
+      "learning_rate": 4.14e-05,
+      "loss": 4.7916,
       "step": 71
     },
     {
       "epoch": 0.16,
-      "learning_rate": 4.14e-05,
-      "loss": 4.8288,
+      "learning_rate": 4.2e-05,
+      "loss": 4.8087,
       "step": 72
     },
     {
       "epoch": 0.16,
-      "learning_rate": 4.2e-05,
-      "loss": 4.7142,
+      "learning_rate": 4.259999999999999e-05,
+      "loss": 4.6922,
       "step": 73
     },
     {
       "epoch": 0.17,
-      "learning_rate": 4.259999999999999e-05,
-      "loss": 4.7101,
+      "learning_rate": 4.319999999999999e-05,
+      "loss": 4.7028,
       "step": 74
     },
     {
       "epoch": 0.17,
-      "learning_rate": 4.319999999999999e-05,
-      "loss": 4.7489,
+      "learning_rate": 4.3799999999999994e-05,
+      "loss": 4.7345,
       "step": 75
     },
     {
       "epoch": 0.17,
-      "learning_rate": 4.3799999999999994e-05,
-      "loss": 4.6794,
+      "learning_rate": 4.4399999999999995e-05,
+      "loss": 4.6529,
       "step": 76
     },
     {
       "epoch": 0.17,
-      "learning_rate": 4.4399999999999995e-05,
-      "loss": 4.7079,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 4.6925,
       "step": 77
     },
     {
       "epoch": 0.17,
-      "learning_rate": 4.4999999999999996e-05,
-      "loss": 4.6963,
+      "learning_rate": 4.56e-05,
+      "loss": 4.6708,
       "step": 78
     },
     {
       "epoch": 0.18,
-      "learning_rate": 4.56e-05,
-      "loss": 4.6342,
+      "learning_rate": 4.62e-05,
+      "loss": 4.6112,
       "step": 79
     },
     {
       "epoch": 0.18,
-      "learning_rate": 4.62e-05,
-      "loss": 4.7477,
+      "learning_rate": 4.68e-05,
+      "loss": 4.7209,
       "step": 80
     },
     {
       "epoch": 0.18,
-      "learning_rate": 4.68e-05,
-      "loss": 4.7116,
+      "learning_rate": 4.7399999999999993e-05,
+      "loss": 4.6872,
       "step": 81
     },
     {
       "epoch": 0.18,
-      "learning_rate": 4.7399999999999993e-05,
-      "loss": 4.6465,
+      "learning_rate": 4.7999999999999994e-05,
+      "loss": 4.6316,
       "step": 82
     },
     {
       "epoch": 0.19,
-      "learning_rate": 4.7999999999999994e-05,
-      "loss": 4.7662,
+      "learning_rate": 4.8599999999999995e-05,
+      "loss": 4.7507,
       "step": 83
     },
     {
       "epoch": 0.19,
-      "learning_rate": 4.8599999999999995e-05,
-      "loss": 4.668,
+      "learning_rate": 4.9199999999999997e-05,
+      "loss": 4.6468,
       "step": 84
     },
     {
       "epoch": 0.19,
-      "learning_rate": 4.9199999999999997e-05,
-      "loss": 4.7007,
+      "learning_rate": 4.98e-05,
+      "loss": 4.6845,
       "step": 85
     },
     {
       "epoch": 0.19,
-      "learning_rate": 4.98e-05,
-      "loss": 4.6541,
+      "learning_rate": 5.04e-05,
+      "loss": 4.6201,
       "step": 86
     },
     {
       "epoch": 0.2,
-      "learning_rate": 5.04e-05,
-      "loss": 4.6354,
+      "learning_rate": 5.1e-05,
+      "loss": 4.6043,
       "step": 87
     },
     {
       "epoch": 0.2,
-      "learning_rate": 5.1e-05,
-      "loss": 4.5581,
+      "learning_rate": 5.1599999999999994e-05,
+      "loss": 4.5494,
       "step": 88
     },
     {
       "epoch": 0.2,
-      "learning_rate": 5.1599999999999994e-05,
-      "loss": 4.6933,
+      "learning_rate": 5.2199999999999995e-05,
+      "loss": 4.6675,
       "step": 89
     },
     {
       "epoch": 0.2,
-      "learning_rate": 5.2199999999999995e-05,
-      "loss": 4.7198,
+      "learning_rate": 5.279999999999999e-05,
+      "loss": 4.7288,
       "step": 90
     },
     {
       "epoch": 0.2,
-      "learning_rate": 5.279999999999999e-05,
-      "loss": 4.5918,
+      "learning_rate": 5.339999999999999e-05,
+      "loss": 4.5647,
       "step": 91
     },
     {
       "epoch": 0.21,
-      "learning_rate": 5.339999999999999e-05,
-      "loss": 4.8693,
+      "learning_rate": 5.399999999999999e-05,
+      "loss": 4.8602,
       "step": 92
     },
     {
       "epoch": 0.21,
-      "learning_rate": 5.399999999999999e-05,
-      "loss": 4.761,
+      "learning_rate": 5.459999999999999e-05,
+      "loss": 4.7511,
       "step": 93
     },
     {
       "epoch": 0.21,
-      "learning_rate": 5.459999999999999e-05,
-      "loss": 4.6738,
+      "learning_rate": 5.519999999999999e-05,
+      "loss": 4.6587,
       "step": 94
     },
     {
       "epoch": 0.21,
-      "learning_rate": 5.519999999999999e-05,
-      "loss": 4.5974,
+      "learning_rate": 5.5799999999999994e-05,
+      "loss": 4.5921,
       "step": 95
     },
     {
       "epoch": 0.22,
-      "learning_rate": 5.5799999999999994e-05,
-      "loss": 4.8575,
+      "learning_rate": 5.6399999999999995e-05,
+      "loss": 4.8161,
       "step": 96
     },
     {
       "epoch": 0.22,
-      "learning_rate": 5.6399999999999995e-05,
-      "loss": 4.7529,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 4.697,
       "step": 97
     },
     {
       "epoch": 0.22,
-      "learning_rate": 5.6999999999999996e-05,
-      "loss": 4.5701,
+      "learning_rate": 5.76e-05,
+      "loss": 4.5279,
       "step": 98
     },
     {
       "epoch": 0.22,
-      "learning_rate": 5.76e-05,
-      "loss": 4.7132,
+      "learning_rate": 5.82e-05,
+      "loss": 4.7026,
       "step": 99
     },
     {
       "epoch": 0.22,
-      "learning_rate": 5.82e-05,
-      "loss": 4.412,
+      "learning_rate": 5.88e-05,
+      "loss": 4.3868,
       "step": 100
     },
     {
       "epoch": 0.23,
       "learning_rate": 5.88e-05,
-      "loss": 6.1576,
+      "loss": 5.9542,
       "step": 101
     },
     {
       "epoch": 0.23,
       "learning_rate": 5.94e-05,
-      "loss": 5.8415,
+      "loss": 5.8248,
       "step": 102
     },
     {
       "epoch": 0.23,
       "learning_rate": 5.9999999999999995e-05,
-      "loss": 5.2825,
+      "loss": 5.5644,
       "step": 103
     },
     {
       "epoch": 0.23,
       "learning_rate": 6.0599999999999996e-05,
-      "loss": 5.1346,
+      "loss": 5.3902,
       "step": 104
     },
     {
       "epoch": 0.24,
       "learning_rate": 6.12e-05,
-      "loss": 4.858,
+      "loss": 4.9764,
       "step": 105
     },
     {
       "epoch": 0.24,
       "learning_rate": 6.18e-05,
-      "loss": 4.7614,
+      "loss": 4.8396,
       "step": 106
     },
     {
       "epoch": 0.24,
       "learning_rate": 6.239999999999999e-05,
-      "loss": 4.7545,
+      "loss": 4.7774,
       "step": 107
     },
     {
       "epoch": 0.24,
       "learning_rate": 6.299999999999999e-05,
-      "loss": 4.6803,
+      "loss": 4.7074,
       "step": 108
     },
     {
       "epoch": 0.24,
       "learning_rate": 6.359999999999999e-05,
-      "loss": 4.6851,
+      "loss": 4.6497,
       "step": 109
     },
     {
       "epoch": 0.25,
       "learning_rate": 6.419999999999999e-05,
-      "loss": 4.7716,
+      "loss": 4.7453,
       "step": 110
     },
     {
       "epoch": 0.25,
       "learning_rate": 6.479999999999999e-05,
-      "loss": 4.7469,
+      "loss": 4.7379,
       "step": 111
     },
     {
       "epoch": 0.25,
       "learning_rate": 6.539999999999999e-05,
-      "loss": 4.7064,
+      "loss": 4.7307,
       "step": 112
     },
     {
       "epoch": 0.25,
       "learning_rate": 6.599999999999999e-05,
-      "loss": 4.6541,
+      "loss": 4.6858,
       "step": 113
     },
     {
       "epoch": 0.26,
       "learning_rate": 6.659999999999999e-05,
-      "loss": 4.5913,
+      "loss": 4.5898,
       "step": 114
     },
     {
       "epoch": 0.26,
       "learning_rate": 6.72e-05,
-      "loss": 4.577,
+      "loss": 4.568,
       "step": 115
     },
     {
       "epoch": 0.26,
       "learning_rate": 6.78e-05,
-      "loss": 4.6932,
+      "loss": 4.685,
       "step": 116
     },
     {
       "epoch": 0.26,
       "learning_rate": 6.84e-05,
-      "loss": 4.6003,
+      "loss": 4.5926,
       "step": 117
     },
     {
       "epoch": 0.26,
       "learning_rate": 6.9e-05,
-      "loss": 4.6862,
+      "loss": 4.691,
       "step": 118
     },
     {
       "epoch": 0.27,
       "learning_rate": 6.96e-05,
-      "loss": 4.6215,
+      "loss": 4.6207,
       "step": 119
     },
     {
       "epoch": 0.27,
       "learning_rate": 7.02e-05,
-      "loss": 4.6836,
+      "loss": 4.6662,
       "step": 120
     },
     {
       "epoch": 0.27,
       "learning_rate": 7.079999999999999e-05,
-      "loss": 4.4829,
+      "loss": 4.4633,
       "step": 121
     },
     {
       "epoch": 0.27,
       "learning_rate": 7.139999999999999e-05,
-      "loss": 4.5853,
+      "loss": 4.5675,
       "step": 122
     },
     {
       "epoch": 0.28,
       "learning_rate": 7.199999999999999e-05,
-      "loss": 4.5679,
+      "loss": 4.5613,
       "step": 123
     },
     {
       "epoch": 0.28,
       "learning_rate": 7.259999999999999e-05,
-      "loss": 4.5366,
+      "loss": 4.5392,
       "step": 124
     },
     {
       "epoch": 0.28,
       "learning_rate": 7.319999999999999e-05,
-      "loss": 4.6412,
+      "loss": 4.6394,
       "step": 125
     },
     {
       "epoch": 0.28,
       "learning_rate": 7.379999999999999e-05,
-      "loss": 4.6462,
+      "loss": 4.6477,
       "step": 126
     },
     {
       "epoch": 0.28,
       "learning_rate": 7.439999999999999e-05,
-      "loss": 4.5981,
+      "loss": 4.5763,
       "step": 127
     },
     {
       "epoch": 0.29,
       "learning_rate": 7.5e-05,
-      "loss": 4.547,
+      "loss": 4.5374,
       "step": 128
     },
     {
       "epoch": 0.29,
       "learning_rate": 7.56e-05,
-      "loss": 4.4812,
+      "loss": 4.469,
       "step": 129
     },
     {
       "epoch": 0.29,
       "learning_rate": 7.62e-05,
-      "loss": 4.635,
+      "loss": 4.6234,
       "step": 130
     },
     {
       "epoch": 0.29,
       "learning_rate": 7.68e-05,
-      "loss": 4.4884,
+      "loss": 4.4815,
       "step": 131
     },
     {
       "epoch": 0.3,
       "learning_rate": 7.74e-05,
-      "loss": 4.5837,
+      "loss": 4.58,
       "step": 132
     },
     {
       "epoch": 0.3,
       "learning_rate": 7.8e-05,
-      "loss": 4.4777,
+      "loss": 4.4586,
       "step": 133
     },
     {
       "epoch": 0.3,
       "learning_rate": 7.86e-05,
-      "loss": 4.5206,
+      "loss": 4.512,
       "step": 134
     },
     {
       "epoch": 0.3,
       "learning_rate": 7.92e-05,
-      "loss": 4.6975,
+      "loss": 4.6734,
       "step": 135
     },
     {
       "epoch": 0.3,
       "learning_rate": 7.98e-05,
-      "loss": 4.5397,
+      "loss": 4.5342,
       "step": 136
     },
     {
       "epoch": 0.31,
       "learning_rate": 8.04e-05,
-      "loss": 4.5827,
+      "loss": 4.5749,
       "step": 137
     },
     {
       "epoch": 0.31,
       "learning_rate": 8.1e-05,
-      "loss": 4.6354,
+      "loss": 4.6223,
       "step": 138
     },
     {
       "epoch": 0.31,
       "learning_rate": 8.16e-05,
-      "loss": 4.6395,
+      "loss": 4.637,
       "step": 139
     },
     {
       "epoch": 0.31,
       "learning_rate": 8.22e-05,
-      "loss": 4.654,
+      "loss": 4.6462,
       "step": 140
     },
     {
       "epoch": 0.32,
       "learning_rate": 8.28e-05,
-      "loss": 4.7321,
+      "loss": 4.6995,
       "step": 141
     },
     {
       "epoch": 0.32,
       "learning_rate": 8.34e-05,
-      "loss": 4.6143,
+      "loss": 4.5716,
       "step": 142
     },
     {
       "epoch": 0.32,
       "learning_rate": 8.4e-05,
-      "loss": 4.5767,
+      "loss": 4.5627,
       "step": 143
     },
     {
       "epoch": 0.32,
       "learning_rate": 8.459999999999998e-05,
-      "loss": 4.6577,
+      "loss": 4.6207,
       "step": 144
     },
     {
       "epoch": 0.33,
       "learning_rate": 8.519999999999998e-05,
-      "loss": 4.5699,
+      "loss": 4.5196,
       "step": 145
     },
     {
       "epoch": 0.33,
       "learning_rate": 8.579999999999998e-05,
-      "loss": 4.6297,
+      "loss": 4.6074,
       "step": 146
     },
     {
       "epoch": 0.33,
       "learning_rate": 8.639999999999999e-05,
-      "loss": 4.785,
+      "loss": 4.7489,
       "step": 147
     },
     {
       "epoch": 0.33,
       "learning_rate": 8.699999999999999e-05,
-      "loss": 4.8082,
+      "loss": 4.7682,
       "step": 148
     },
     {
       "epoch": 0.33,
       "learning_rate": 8.759999999999999e-05,
-      "loss": 4.6153,
+      "loss": 4.5681,
       "step": 149
     },
     {
       "epoch": 0.34,
       "learning_rate": 8.819999999999999e-05,
-      "loss": 4.4768,
+      "loss": 4.469,
       "step": 150
     },
     {
       "epoch": 0.34,
       "learning_rate": 8.879999999999999e-05,
-      "loss": 5.3978,
+      "loss": 6.2583,
       "step": 151
     },
     {
       "epoch": 0.34,
       "learning_rate": 8.939999999999999e-05,
-      "loss": 5.2105,
+      "loss": 5.6164,
       "step": 152
     },
     {
       "epoch": 0.34,
       "learning_rate": 8.999999999999999e-05,
-      "loss": 5.0365,
+      "loss": 5.1907,
       "step": 153
     },
     {
       "epoch": 0.35,
       "learning_rate": 9.059999999999999e-05,
-      "loss": 4.7275,
+      "loss": 4.7733,
       "step": 154
     },
     {
       "epoch": 0.35,
       "learning_rate": 9.12e-05,
-      "loss": 4.814,
+      "loss": 4.8415,
       "step": 155
     },
     {
       "epoch": 0.35,
       "learning_rate": 9.18e-05,
-      "loss": 4.6306,
+      "loss": 4.6594,
       "step": 156
     },
     {
       "epoch": 0.35,
       "learning_rate": 9.24e-05,
-      "loss": 4.5816,
+      "loss": 4.6,
       "step": 157
     },
     {
       "epoch": 0.35,
       "learning_rate": 9.3e-05,
-      "loss": 4.6355,
+      "loss": 4.659,
       "step": 158
     },
     {
       "epoch": 0.36,
       "learning_rate": 9.36e-05,
-      "loss": 4.6574,
+      "loss": 4.6704,
       "step": 159
     },
     {
       "epoch": 0.36,
       "learning_rate": 9.419999999999999e-05,
-      "loss": 4.6793,
+      "loss": 4.7041,
       "step": 160
     },
     {
       "epoch": 0.36,
       "learning_rate": 9.479999999999999e-05,
-      "loss": 4.6908,
+      "loss": 4.7007,
       "step": 161
     },
     {
       "epoch": 0.36,
       "learning_rate": 9.539999999999999e-05,
-      "loss": 4.6475,
+      "loss": 4.6682,
       "step": 162
     },
     {
       "epoch": 0.37,
       "learning_rate": 9.599999999999999e-05,
-      "loss": 4.5892,
+      "loss": 4.6197,
       "step": 163
     },
     {
       "epoch": 0.37,
       "learning_rate": 9.659999999999999e-05,
-      "loss": 4.6243,
+      "loss": 4.62,
       "step": 164
     },
     {
       "epoch": 0.37,
       "learning_rate": 9.719999999999999e-05,
-      "loss": 4.4789,
+      "loss": 4.4891,
       "step": 165
     },
     {
       "epoch": 0.37,
       "learning_rate": 9.779999999999999e-05,
-      "loss": 4.5289,
+      "loss": 4.5451,
       "step": 166
     },
     {
       "epoch": 0.37,
       "learning_rate": 9.839999999999999e-05,
-      "loss": 4.4315,
+      "loss": 4.4452,
       "step": 167
     },
     {
       "epoch": 0.38,
       "learning_rate": 9.9e-05,
-      "loss": 4.505,
+      "loss": 4.5057,
       "step": 168
     },
     {
       "epoch": 0.38,
       "learning_rate": 9.96e-05,
-      "loss": 4.5744,
+      "loss": 4.5834,
       "step": 169
     },
     {
       "epoch": 0.38,
       "learning_rate": 0.0001002,
-      "loss": 4.5639,
+      "loss": 4.5699,
       "step": 170
     },
     {
       "epoch": 0.38,
       "learning_rate": 0.0001008,
-      "loss": 4.563,
+      "loss": 4.5678,
       "step": 171
     },
     {
       "epoch": 0.39,
       "learning_rate": 0.0001014,
-      "loss": 4.6649,
+      "loss": 4.6897,
       "step": 172
     },
     {
       "epoch": 0.39,
       "learning_rate": 0.000102,
-      "loss": 4.4817,
+      "loss": 4.4966,
       "step": 173
     },
     {
       "epoch": 0.39,
       "learning_rate": 0.0001026,
-      "loss": 4.5898,
+      "loss": 4.6041,
       "step": 174
     },
     {
       "epoch": 0.39,
       "learning_rate": 0.00010319999999999999,
-      "loss": 4.6714,
+      "loss": 4.6913,
       "step": 175
     },
     {
       "epoch": 0.39,
       "learning_rate": 0.00010379999999999999,
-      "loss": 4.4916,
+      "loss": 4.4995,
       "step": 176
     },
     {
       "epoch": 0.4,
       "learning_rate": 0.00010439999999999999,
-      "loss": 4.4783,
+      "loss": 4.4945,
       "step": 177
     },
     {
       "epoch": 0.4,
       "learning_rate": 0.00010499999999999999,
-      "loss": 4.549,
+      "loss": 4.5747,
       "step": 178
     },
     {
       "epoch": 0.4,
       "learning_rate": 0.00010559999999999998,
-      "loss": 4.6129,
+      "loss": 4.631,
       "step": 179
     },
     {
       "epoch": 0.4,
       "learning_rate": 0.00010619999999999998,
-      "loss": 4.6554,
+      "loss": 4.6599,
       "step": 180
     },
     {
       "epoch": 0.41,
       "learning_rate": 0.00010679999999999998,
-      "loss": 4.5494,
+      "loss": 4.5648,
       "step": 181
     },
     {
       "epoch": 0.41,
       "learning_rate": 0.00010739999999999998,
-      "loss": 4.5649,
+      "loss": 4.5608,
       "step": 182
     },
     {
       "epoch": 0.41,
       "learning_rate": 0.00010799999999999998,
-      "loss": 4.5621,
+      "loss": 4.5623,
       "step": 183
     },
     {
       "epoch": 0.41,
       "learning_rate": 0.00010859999999999998,
-      "loss": 4.5943,
+      "loss": 4.6001,
       "step": 184
     },
     {
       "epoch": 0.41,
       "learning_rate": 0.00010919999999999998,
-      "loss": 4.5343,
+      "loss": 4.5398,
       "step": 185
     },
     {
       "epoch": 0.42,
       "learning_rate": 0.00010979999999999999,
-      "loss": 4.4241,
+      "loss": 4.4157,
       "step": 186
     },
     {
       "epoch": 0.42,
       "learning_rate": 0.00011039999999999999,
-      "loss": 4.4747,
+      "loss": 4.4919,
       "step": 187
     },
     {
       "epoch": 0.42,
       "learning_rate": 0.00011099999999999999,
-      "loss": 4.4546,
+      "loss": 4.4482,
       "step": 188
     },
     {
       "epoch": 0.42,
       "learning_rate": 0.00011159999999999999,
-      "loss": 4.5061,
+      "loss": 4.5045,
       "step": 189
     },
     {
       "epoch": 0.43,
       "learning_rate": 0.00011219999999999999,
-      "loss": 4.6313,
+      "loss": 4.618,
       "step": 190
     },
     {
       "epoch": 0.43,
       "learning_rate": 0.00011279999999999999,
-      "loss": 4.5184,
+      "loss": 4.5325,
       "step": 191
     },
     {
       "epoch": 0.43,
       "learning_rate": 0.00011339999999999999,
-      "loss": 4.5427,
+      "loss": 4.5477,
       "step": 192
     },
     {
       "epoch": 0.43,
       "learning_rate": 0.00011399999999999999,
-      "loss": 4.4119,
+      "loss": 4.4201,
       "step": 193
     },
     {
       "epoch": 0.43,
       "learning_rate": 0.0001146,
-      "loss": 4.5766,
+      "loss": 4.5925,
       "step": 194
     },
     {
       "epoch": 0.44,
       "learning_rate": 0.0001152,
-      "loss": 4.6803,
+      "loss": 4.6806,
       "step": 195
     },
     {
       "epoch": 0.44,
       "learning_rate": 0.0001158,
-      "loss": 4.5398,
+      "loss": 4.5368,
       "step": 196
     },
     {
       "epoch": 0.44,
       "learning_rate": 0.0001164,
-      "loss": 4.4595,
+      "loss": 4.4654,
       "step": 197
     },
     {
       "epoch": 0.44,
       "learning_rate": 0.000117,
-      "loss": 4.5649,
+      "loss": 4.5484,
       "step": 198
     },
     {
       "epoch": 0.45,
       "learning_rate": 0.0001176,
-      "loss": 4.4631,
+      "loss": 4.4713,
       "step": 199
     },
     {
       "epoch": 0.45,
       "learning_rate": 0.0001182,
-      "loss": 4.4779,
+      "loss": 4.5144,
       "step": 200
     },
     {
       "epoch": 0.45,
       "learning_rate": 0.0001188,
-      "loss": 5.7726,
+      "loss": 5.4254,
       "step": 201
     },
     {
       "epoch": 0.45,
       "learning_rate": 0.0001194,
-      "loss": 5.6596,
+      "loss": 5.4006,
       "step": 202
     },
     {
       "epoch": 0.46,
       "learning_rate": 0.00011999999999999999,
-      "loss": 5.3307,
+      "loss": 5.2168,
       "step": 203
     },
     {
       "epoch": 0.46,
       "learning_rate": 0.00012059999999999999,
-      "loss": 5.0387,
+      "loss": 4.9626,
       "step": 204
     },
     {
       "epoch": 0.46,
       "learning_rate": 0.00012119999999999999,
-      "loss": 4.969,
+      "loss": 4.8992,
       "step": 205
     },
     {
       "epoch": 0.46,
       "learning_rate": 0.00012179999999999999,
-      "loss": 4.8809,
+      "loss": 4.811,
       "step": 206
     },
     {
       "epoch": 0.46,
       "learning_rate": 0.0001224,
-      "loss": 4.6725,
+      "loss": 4.6223,
       "step": 207
     },
     {
       "epoch": 0.47,
       "learning_rate": 0.00012299999999999998,
-      "loss": 4.7561,
+      "loss": 4.7224,
       "step": 208
     },
     {
       "epoch": 0.47,
       "learning_rate": 0.0001236,
-      "loss": 4.7825,
+      "loss": 4.7679,
       "step": 209
     },
     {
       "epoch": 0.47,
       "learning_rate": 0.00012419999999999998,
-      "loss": 4.7099,
+      "loss": 4.6853,
       "step": 210
     },
     {
       "epoch": 0.47,
       "learning_rate": 0.00012479999999999997,
-      "loss": 4.5271,
+      "loss": 4.5118,
       "step": 211
     },
     {
       "epoch": 0.48,
       "learning_rate": 0.00012539999999999999,
-      "loss": 4.642,
+      "loss": 4.6325,
       "step": 212
     },
     {
       "epoch": 0.48,
       "learning_rate": 0.00012599999999999997,
-      "loss": 4.4652,
+      "loss": 4.4748,
       "step": 213
     },
     {
       "epoch": 0.48,
       "learning_rate": 0.0001266,
-      "loss": 4.5103,
+      "loss": 4.513,
       "step": 214
     },
     {
       "epoch": 0.48,
       "learning_rate": 0.00012719999999999997,
-      "loss": 4.5699,
+      "loss": 4.5576,
       "step": 215
     },
     {
       "epoch": 0.48,
       "learning_rate": 0.0001278,
-      "loss": 4.5555,
+      "loss": 4.555,
       "step": 216
     },
     {
       "epoch": 0.49,
       "learning_rate": 0.00012839999999999998,
-      "loss": 4.5327,
+      "loss": 4.5305,
       "step": 217
     },
     {
       "epoch": 0.49,
       "learning_rate": 0.000129,
-      "loss": 4.5224,
+      "loss": 4.5359,
       "step": 218
     },
     {
       "epoch": 0.49,
       "learning_rate": 0.00012959999999999998,
-      "loss": 4.4599,
+      "loss": 4.4294,
       "step": 219
     },
     {
       "epoch": 0.49,
       "learning_rate": 0.0001302,
-      "loss": 4.5868,
+      "loss": 4.5847,
       "step": 220
     },
     {
       "epoch": 0.5,
       "learning_rate": 0.00013079999999999998,
-      "loss": 4.4898,
+      "loss": 4.4666,
       "step": 221
     },
     {
       "epoch": 0.5,
       "learning_rate": 0.0001314,
-      "loss": 4.5415,
+      "loss": 4.5346,
       "step": 222
     },
     {
       "epoch": 0.5,
       "learning_rate": 0.00013199999999999998,
-      "loss": 4.5053,
+      "loss": 4.4938,
       "step": 223
     },
     {
       "epoch": 0.5,
       "learning_rate": 0.0001326,
-      "loss": 4.4692,
+      "loss": 4.4626,
       "step": 224
     },
     {
       "epoch": 0.5,
       "learning_rate": 0.00013319999999999999,
-      "loss": 4.6647,
+      "loss": 4.668,
       "step": 225
     },
     {
       "epoch": 0.51,
       "learning_rate": 0.0001338,
-      "loss": 4.5009,
+      "loss": 4.4876,
       "step": 226
     },
     {
       "epoch": 0.51,
       "learning_rate": 0.0001344,
-      "loss": 4.6426,
+      "loss": 4.6419,
       "step": 227
     },
     {
       "epoch": 0.51,
       "learning_rate": 0.000135,
-      "loss": 4.5966,
+      "loss": 4.5957,
       "step": 228
     },
     {
       "epoch": 0.51,
       "learning_rate": 0.0001356,
-      "loss": 4.5119,
+      "loss": 4.5052,
       "step": 229
     },
     {
       "epoch": 0.52,
       "learning_rate": 0.0001362,
-      "loss": 4.528,
+      "loss": 4.5292,
       "step": 230
     },
     {
       "epoch": 0.52,
       "learning_rate": 0.0001368,
-      "loss": 4.6159,
+      "loss": 4.5961,
       "step": 231
     },
     {
       "epoch": 0.52,
       "learning_rate": 0.0001374,
-      "loss": 4.4335,
+      "loss": 4.4206,
       "step": 232
     },
     {
       "epoch": 0.52,
       "learning_rate": 0.000138,
-      "loss": 4.452,
+      "loss": 4.4587,
       "step": 233
     },
     {
       "epoch": 0.52,
       "learning_rate": 0.0001386,
-      "loss": 4.5153,
+      "loss": 4.51,
       "step": 234
     },
     {
       "epoch": 0.53,
       "learning_rate": 0.0001392,
-      "loss": 4.5963,
+      "loss": 4.5874,
       "step": 235
     },
     {
       "epoch": 0.53,
       "learning_rate": 0.00013979999999999998,
-      "loss": 4.4854,
+      "loss": 4.4853,
       "step": 236
     },
     {
       "epoch": 0.53,
       "learning_rate": 0.0001404,
-      "loss": 4.5933,
+      "loss": 4.5932,
       "step": 237
     },
     {
       "epoch": 0.53,
       "learning_rate": 0.00014099999999999998,
-      "loss": 4.4843,
+      "loss": 4.4827,
       "step": 238
     },
     {
       "epoch": 0.54,
       "learning_rate": 0.00014159999999999997,
-      "loss": 4.4572,
+      "loss": 4.4535,
       "step": 239
     },
     {
       "epoch": 0.54,
       "learning_rate": 0.0001422,
-      "loss": 4.528,
+      "loss": 4.532,
       "step": 240
     },
     {
       "epoch": 0.54,
       "learning_rate": 0.00014279999999999997,
-      "loss": 4.5125,
+      "loss": 4.5145,
       "step": 241
     },
     {
       "epoch": 0.54,
       "learning_rate": 0.0001434,
-      "loss": 4.5634,
+      "loss": 4.5681,
       "step": 242
     },
     {
       "epoch": 0.54,
       "learning_rate": 0.00014399999999999998,
-      "loss": 4.7007,
+      "loss": 4.7017,
       "step": 243
     },
     {
       "epoch": 0.55,
       "learning_rate": 0.0001446,
-      "loss": 4.5203,
+      "loss": 4.5186,
       "step": 244
     },
     {
       "epoch": 0.55,
       "learning_rate": 0.00014519999999999998,
-      "loss": 4.5344,
+      "loss": 4.5254,
       "step": 245
     },
     {
       "epoch": 0.55,
       "learning_rate": 0.0001458,
-      "loss": 4.4341,
+      "loss": 4.4415,
       "step": 246
     },
     {
       "epoch": 0.55,
       "learning_rate": 0.00014639999999999998,
-      "loss": 4.3632,
+      "loss": 4.3858,
       "step": 247
     },
     {
       "epoch": 0.56,
       "learning_rate": 0.000147,
-      "loss": 4.6095,
+      "loss": 4.6209,
       "step": 248
     },
     {
       "epoch": 0.56,
       "learning_rate": 0.00014759999999999998,
-      "loss": 4.5632,
+      "loss": 4.5217,
       "step": 249
     },
     {
       "epoch": 0.56,
       "learning_rate": 0.0001482,
-      "loss": 4.3235,
+      "loss": 4.3601,
       "step": 250
     },
     {
       "epoch": 0.56,
       "learning_rate": 0.00014879999999999998,
-      "loss": 5.7329,
+      "loss": 5.5394,
       "step": 251
     },
     {
       "epoch": 0.57,
       "learning_rate": 0.0001494,
-      "loss": 5.6422,
+      "loss": 5.4343,
       "step": 252
     },
     {
       "epoch": 0.57,
       "learning_rate": 0.00015,
-      "loss": 5.3057,
+      "loss": 5.2117,
       "step": 253
     },
     {
       "epoch": 0.57,
       "learning_rate": 0.00015059999999999997,
-      "loss": 4.933,
+      "loss": 4.8667,
       "step": 254
     },
     {
       "epoch": 0.57,
       "learning_rate": 0.0001512,
-      "loss": 4.8009,
+      "loss": 4.7689,
       "step": 255
     },
     {
       "epoch": 0.57,
       "learning_rate": 0.00015179999999999998,
-      "loss": 4.7696,
+      "loss": 4.7223,
       "step": 256
     },
     {
       "epoch": 0.58,
       "learning_rate": 0.0001524,
-      "loss": 4.6338,
+      "loss": 4.6127,
       "step": 257
     },
     {
       "epoch": 0.58,
       "learning_rate": 0.00015299999999999998,
-      "loss": 4.5604,
+      "loss": 4.5517,
       "step": 258
     },
     {
       "epoch": 0.58,
       "learning_rate": 0.0001536,
-      "loss": 4.6174,
+      "loss": 4.6104,
       "step": 259
     },
     {
       "epoch": 0.58,
       "learning_rate": 0.00015419999999999998,
-      "loss": 4.5476,
+      "loss": 4.5236,
       "step": 260
     },
     {
       "epoch": 0.59,
       "learning_rate": 0.0001548,
-      "loss": 4.6667,
+      "loss": 4.6442,
       "step": 261
     },
     {
       "epoch": 0.59,
       "learning_rate": 0.00015539999999999998,
-      "loss": 4.5453,
+      "loss": 4.5383,
       "step": 262
     },
     {
       "epoch": 0.59,
       "learning_rate": 0.000156,
-      "loss": 4.684,
+      "loss": 4.698,
       "step": 263
     },
     {
       "epoch": 0.59,
       "learning_rate": 0.00015659999999999998,
-      "loss": 4.5324,
+      "loss": 4.5208,
       "step": 264
     },
     {
       "epoch": 0.59,
       "learning_rate": 0.0001572,
-      "loss": 4.5577,
+      "loss": 4.5523,
       "step": 265
     },
     {
       "epoch": 0.6,
       "learning_rate": 0.0001578,
-      "loss": 4.492,
+      "loss": 4.5186,
       "step": 266
     },
     {
       "epoch": 0.6,
       "learning_rate": 0.0001584,
-      "loss": 4.6238,
+      "loss": 4.6301,
       "step": 267
     },
     {
       "epoch": 0.6,
       "learning_rate": 0.000159,
-      "loss": 4.4791,
+      "loss": 4.4806,
       "step": 268
     },
     {
       "epoch": 0.6,
       "learning_rate": 0.0001596,
-      "loss": 4.6115,
+      "loss": 4.5985,
       "step": 269
     },
     {
       "epoch": 0.61,
       "learning_rate": 0.0001602,
-      "loss": 4.5068,
+      "loss": 4.5172,
       "step": 270
     },
     {
       "epoch": 0.61,
       "learning_rate": 0.0001608,
-      "loss": 4.4667,
+      "loss": 4.4582,
       "step": 271
     },
     {
       "epoch": 0.61,
       "learning_rate": 0.0001614,
-      "loss": 4.5432,
+      "loss": 4.5442,
       "step": 272
     },
     {
       "epoch": 0.61,
       "learning_rate": 0.000162,
-      "loss": 4.3894,
+      "loss": 4.3896,
       "step": 273
     },
     {
       "epoch": 0.61,
       "learning_rate": 0.0001626,
-      "loss": 4.448,
+      "loss": 4.4235,
       "step": 274
     },
     {
       "epoch": 0.62,
       "learning_rate": 0.0001632,
-      "loss": 4.5522,
+      "loss": 4.5386,
       "step": 275
     },
     {
       "epoch": 0.62,
       "learning_rate": 0.0001638,
-      "loss": 4.4934,
+      "loss": 4.4787,
       "step": 276
     },
     {
       "epoch": 0.62,
       "learning_rate": 0.0001644,
-      "loss": 4.4612,
+      "loss": 4.4511,
       "step": 277
     },
     {
       "epoch": 0.62,
       "learning_rate": 0.000165,
-      "loss": 4.4943,
+      "loss": 4.5093,
       "step": 278
     },
     {
       "epoch": 0.63,
       "learning_rate": 0.0001656,
-      "loss": 4.5346,
+      "loss": 4.5233,
       "step": 279
     },
     {
       "epoch": 0.63,
       "learning_rate": 0.0001662,
-      "loss": 4.6029,
+      "loss": 4.5999,
       "step": 280
     },
     {
       "epoch": 0.63,
       "learning_rate": 0.0001668,
-      "loss": 4.5479,
+      "loss": 4.5561,
       "step": 281
     },
     {
       "epoch": 0.63,
       "learning_rate": 0.0001674,
-      "loss": 4.3756,
+      "loss": 4.3713,
       "step": 282
     },
     {
       "epoch": 0.63,
       "learning_rate": 0.000168,
-      "loss": 4.3436,
+      "loss": 4.3426,
       "step": 283
     },
     {
       "epoch": 0.64,
       "learning_rate": 0.0001686,
-      "loss": 4.4843,
+      "loss": 4.5017,
       "step": 284
     },
     {
       "epoch": 0.64,
       "learning_rate": 0.00016919999999999997,
-      "loss": 4.5724,
+      "loss": 4.5855,
       "step": 285
     },
     {
       "epoch": 0.64,
       "learning_rate": 0.00016979999999999998,
-      "loss": 4.476,
+      "loss": 4.5,
       "step": 286
     },
     {
       "epoch": 0.64,
       "learning_rate": 0.00017039999999999997,
-      "loss": 4.5815,
+      "loss": 4.5955,
       "step": 287
     },
     {
       "epoch": 0.65,
       "learning_rate": 0.00017099999999999998,
-      "loss": 4.496,
+      "loss": 4.4935,
       "step": 288
     },
     {
       "epoch": 0.65,
       "learning_rate": 0.00017159999999999997,
-      "loss": 4.4438,
+      "loss": 4.4523,
       "step": 289
     },
     {
       "epoch": 0.65,
       "learning_rate": 0.00017219999999999998,
-      "loss": 4.4612,
+      "loss": 4.4598,
       "step": 290
     },
     {
       "epoch": 0.65,
       "learning_rate": 0.00017279999999999997,
-      "loss": 4.6847,
+      "loss": 4.6806,
       "step": 291
     },
     {
       "epoch": 0.65,
       "learning_rate": 0.00017339999999999996,
-      "loss": 4.5454,
+      "loss": 4.5561,
       "step": 292
     },
     {
       "epoch": 0.66,
       "learning_rate": 0.00017399999999999997,
-      "loss": 4.5934,
+      "loss": 4.5914,
       "step": 293
     },
     {
       "epoch": 0.66,
       "learning_rate": 0.00017459999999999996,
-      "loss": 4.5504,
+      "loss": 4.537,
       "step": 294
     },
     {
       "epoch": 0.66,
       "learning_rate": 0.00017519999999999998,
-      "loss": 4.4558,
+      "loss": 4.4617,
       "step": 295
     },
     {
       "epoch": 0.66,
       "learning_rate": 0.00017579999999999996,
-      "loss": 4.4989,
+      "loss": 4.4864,
       "step": 296
     },
     {
       "epoch": 0.67,
       "learning_rate": 0.00017639999999999998,
-      "loss": 4.4146,
+      "loss": 4.4368,
       "step": 297
     },
     {
       "epoch": 0.67,
       "learning_rate": 0.00017699999999999997,
-      "loss": 4.6314,
+      "loss": 4.6713,
       "step": 298
     },
     {
       "epoch": 0.67,
       "learning_rate": 0.00017759999999999998,
-      "loss": 4.2778,
+      "loss": 4.2836,
       "step": 299
     },
     {
       "epoch": 0.67,
       "learning_rate": 0.00017819999999999997,
-      "loss": 4.4758,
+      "loss": 4.4759,
       "step": 300
     },
     {
       "epoch": 0.67,
       "learning_rate": 0.00017879999999999998,
-      "loss": 5.5599,
+      "loss": 5.7464,
       "step": 301
     },
     {
       "epoch": 0.68,
       "learning_rate": 0.00017939999999999997,
-      "loss": 5.3478,
+      "loss": 5.5533,
       "step": 302
     },
     {
       "epoch": 0.68,
       "learning_rate": 0.00017999999999999998,
-      "loss": 5.0825,
+      "loss": 5.2607,
       "step": 303
     },
     {
       "epoch": 0.68,
       "learning_rate": 0.00018059999999999997,
-      "loss": 4.996,
+      "loss": 5.0842,
       "step": 304
     },
     {
       "epoch": 0.68,
       "learning_rate": 0.00018119999999999999,
-      "loss": 4.8184,
+      "loss": 4.9463,
       "step": 305
     },
     {
       "epoch": 0.69,
       "learning_rate": 0.00018179999999999997,
-      "loss": 4.6625,
+      "loss": 4.7063,
       "step": 306
     },
     {
       "epoch": 0.69,
       "learning_rate": 0.0001824,
-      "loss": 4.666,
+      "loss": 4.6942,
       "step": 307
     },
     {
       "epoch": 0.69,
       "learning_rate": 0.00018299999999999998,
-      "loss": 4.6589,
+      "loss": 4.6681,
       "step": 308
     },
     {
       "epoch": 0.69,
       "learning_rate": 0.0001836,
-      "loss": 4.641,
+      "loss": 4.6439,
       "step": 309
     },
     {
       "epoch": 0.7,
       "learning_rate": 0.00018419999999999998,
-      "loss": 4.7132,
+      "loss": 4.7031,
       "step": 310
     },
     {
       "epoch": 0.7,
       "learning_rate": 0.0001848,
-      "loss": 4.4362,
+      "loss": 4.4416,
       "step": 311
     },
     {
       "epoch": 0.7,
       "learning_rate": 0.00018539999999999998,
-      "loss": 4.5677,
+      "loss": 4.5832,
       "step": 312
     },
     {
       "epoch": 0.7,
       "learning_rate": 0.000186,
-      "loss": 4.5749,
+      "loss": 4.5866,
       "step": 313
     },
     {
       "epoch": 0.7,
       "learning_rate": 0.00018659999999999998,
-      "loss": 4.4865,
+      "loss": 4.4809,
       "step": 314
     },
     {
       "epoch": 0.71,
       "learning_rate": 0.0001872,
-      "loss": 4.5187,
+      "loss": 4.5242,
       "step": 315
     },
     {
       "epoch": 0.71,
       "learning_rate": 0.00018779999999999998,
-      "loss": 4.4829,
+      "loss": 4.4804,
       "step": 316
     },
     {
       "epoch": 0.71,
       "learning_rate": 0.00018839999999999997,
-      "loss": 4.4654,
+      "loss": 4.4799,
       "step": 317
     },
     {
       "epoch": 0.71,
       "learning_rate": 0.00018899999999999999,
-      "loss": 4.4546,
+      "loss": 4.4712,
       "step": 318
     },
     {
       "epoch": 0.72,
       "learning_rate": 0.00018959999999999997,
-      "loss": 4.471,
+      "loss": 4.4968,
       "step": 319
     },
     {
       "epoch": 0.72,
       "learning_rate": 0.0001902,
-      "loss": 4.5822,
+      "loss": 4.5755,
       "step": 320
     },
     {
       "epoch": 0.72,
       "learning_rate": 0.00019079999999999998,
-      "loss": 4.4716,
+      "loss": 4.4656,
       "step": 321
     },
     {
       "epoch": 0.72,
       "learning_rate": 0.0001914,
-      "loss": 4.3796,
+      "loss": 4.3633,
       "step": 322
     },
     {
       "epoch": 0.72,
       "learning_rate": 0.00019199999999999998,
-      "loss": 4.477,
+      "loss": 4.458,
       "step": 323
     },
     {
       "epoch": 0.73,
       "learning_rate": 0.0001926,
-      "loss": 4.3998,
+      "loss": 4.3913,
       "step": 324
     },
     {
@@ -1960,7768 +1960,7768 @@
     {
       "epoch": 0.73,
       "learning_rate": 0.0001938,
-      "loss": 4.5013,
+      "loss": 4.4925,
       "step": 326
     },
     {
       "epoch": 0.73,
       "learning_rate": 0.00019439999999999998,
-      "loss": 4.4171,
+      "loss": 4.4414,
       "step": 327
     },
     {
       "epoch": 0.74,
       "learning_rate": 0.000195,
-      "loss": 4.3593,
+      "loss": 4.3735,
       "step": 328
     },
     {
       "epoch": 0.74,
       "learning_rate": 0.00019559999999999998,
-      "loss": 4.4115,
+      "loss": 4.4079,
       "step": 329
     },
     {
       "epoch": 0.74,
       "learning_rate": 0.0001962,
-      "loss": 4.4946,
+      "loss": 4.5079,
       "step": 330
     },
     {
       "epoch": 0.74,
       "learning_rate": 0.00019679999999999999,
-      "loss": 4.5211,
+      "loss": 4.515,
       "step": 331
     },
     {
       "epoch": 0.74,
       "learning_rate": 0.0001974,
-      "loss": 4.3388,
+      "loss": 4.3635,
       "step": 332
     },
     {
       "epoch": 0.75,
       "learning_rate": 0.000198,
-      "loss": 4.5557,
+      "loss": 4.5648,
       "step": 333
     },
     {
       "epoch": 0.75,
       "learning_rate": 0.0001986,
-      "loss": 4.4958,
+      "loss": 4.4682,
       "step": 334
     },
     {
       "epoch": 0.75,
       "learning_rate": 0.0001992,
-      "loss": 4.5496,
+      "loss": 4.5643,
       "step": 335
     },
     {
       "epoch": 0.75,
       "learning_rate": 0.0001998,
-      "loss": 4.4278,
+      "loss": 4.4503,
       "step": 336
     },
     {
       "epoch": 0.76,
       "learning_rate": 0.0002004,
-      "loss": 4.4312,
+      "loss": 4.4666,
       "step": 337
     },
     {
       "epoch": 0.76,
       "learning_rate": 0.000201,
-      "loss": 4.4803,
+      "loss": 4.4885,
       "step": 338
     },
     {
       "epoch": 0.76,
       "learning_rate": 0.0002016,
-      "loss": 4.6174,
+      "loss": 4.5766,
       "step": 339
     },
     {
       "epoch": 0.76,
       "learning_rate": 0.0002022,
-      "loss": 4.5558,
+      "loss": 4.6036,
       "step": 340
     },
     {
       "epoch": 0.76,
       "learning_rate": 0.0002028,
-      "loss": 4.4443,
+      "loss": 4.4403,
       "step": 341
     },
     {
       "epoch": 0.77,
       "learning_rate": 0.00020339999999999998,
-      "loss": 4.5098,
+      "loss": 4.5286,
       "step": 342
     },
     {
       "epoch": 0.77,
       "learning_rate": 0.000204,
-      "loss": 4.4478,
+      "loss": 4.4455,
       "step": 343
     },
     {
       "epoch": 0.77,
       "learning_rate": 0.00020459999999999999,
-      "loss": 4.5555,
+      "loss": 4.5612,
       "step": 344
     },
     {
       "epoch": 0.77,
       "learning_rate": 0.0002052,
-      "loss": 4.2998,
+      "loss": 4.2895,
       "step": 345
     },
     {
       "epoch": 0.78,
       "learning_rate": 0.0002058,
-      "loss": 4.4556,
+      "loss": 4.4651,
       "step": 346
     },
     {
       "epoch": 0.78,
       "learning_rate": 0.00020639999999999998,
-      "loss": 4.3504,
+      "loss": 4.3408,
       "step": 347
     },
     {
       "epoch": 0.78,
       "learning_rate": 0.00020699999999999996,
-      "loss": 4.2708,
+      "loss": 4.1872,
       "step": 348
     },
     {
       "epoch": 0.78,
       "learning_rate": 0.00020759999999999998,
-      "loss": 4.3868,
+      "loss": 4.3745,
       "step": 349
     },
     {
       "epoch": 0.78,
       "learning_rate": 0.00020819999999999996,
-      "loss": 4.0596,
+      "loss": 4.0009,
       "step": 350
     },
     {
       "epoch": 0.79,
       "learning_rate": 0.00020879999999999998,
-      "loss": 5.457,
+      "loss": 5.993,
       "step": 351
     },
     {
       "epoch": 0.79,
       "learning_rate": 0.00020939999999999997,
-      "loss": 5.1618,
+      "loss": 5.8259,
       "step": 352
     },
     {
       "epoch": 0.79,
       "learning_rate": 0.00020999999999999998,
-      "loss": 4.8537,
+      "loss": 5.3393,
       "step": 353
     },
     {
       "epoch": 0.79,
       "learning_rate": 0.00021059999999999997,
-      "loss": 4.6417,
+      "loss": 4.9743,
       "step": 354
     },
     {
       "epoch": 0.8,
       "learning_rate": 0.00021119999999999996,
-      "loss": 4.7513,
+      "loss": 4.9793,
       "step": 355
     },
     {
       "epoch": 0.8,
       "learning_rate": 0.00021179999999999997,
-      "loss": 4.7933,
+      "loss": 4.9003,
       "step": 356
     },
     {
       "epoch": 0.8,
       "learning_rate": 0.00021239999999999996,
-      "loss": 4.6334,
+      "loss": 4.7375,
       "step": 357
     },
     {
       "epoch": 0.8,
       "learning_rate": 0.00021299999999999997,
-      "loss": 4.6693,
+      "loss": 4.7576,
       "step": 358
     },
     {
       "epoch": 0.8,
       "learning_rate": 0.00021359999999999996,
-      "loss": 4.5336,
+      "loss": 4.6035,
       "step": 359
     },
     {
       "epoch": 0.81,
       "learning_rate": 0.00021419999999999998,
-      "loss": 4.5268,
+      "loss": 4.5928,
       "step": 360
     },
     {
       "epoch": 0.81,
       "learning_rate": 0.00021479999999999996,
-      "loss": 4.4647,
+      "loss": 4.5001,
       "step": 361
     },
     {
       "epoch": 0.81,
       "learning_rate": 0.00021539999999999998,
-      "loss": 4.4749,
+      "loss": 4.5366,
       "step": 362
     },
     {
       "epoch": 0.81,
       "learning_rate": 0.00021599999999999996,
-      "loss": 4.4175,
+      "loss": 4.4527,
       "step": 363
     },
     {
       "epoch": 0.82,
       "learning_rate": 0.00021659999999999998,
-      "loss": 4.4526,
+      "loss": 4.5041,
       "step": 364
     },
     {
       "epoch": 0.82,
       "learning_rate": 0.00021719999999999997,
-      "loss": 4.5824,
+      "loss": 4.6164,
       "step": 365
     },
     {
       "epoch": 0.82,
       "learning_rate": 0.00021779999999999998,
-      "loss": 4.3921,
+      "loss": 4.4303,
       "step": 366
     },
     {
       "epoch": 0.82,
       "learning_rate": 0.00021839999999999997,
-      "loss": 4.5027,
+      "loss": 4.5242,
       "step": 367
     },
     {
       "epoch": 0.83,
       "learning_rate": 0.00021899999999999998,
-      "loss": 4.4884,
+      "loss": 4.5089,
       "step": 368
     },
     {
       "epoch": 0.83,
       "learning_rate": 0.00021959999999999997,
-      "loss": 4.4069,
+      "loss": 4.4234,
       "step": 369
     },
     {
       "epoch": 0.83,
       "learning_rate": 0.00022019999999999999,
-      "loss": 4.3835,
+      "loss": 4.4205,
       "step": 370
     },
     {
       "epoch": 0.83,
       "learning_rate": 0.00022079999999999997,
-      "loss": 4.3501,
+      "loss": 4.3864,
       "step": 371
     },
     {
       "epoch": 0.83,
       "learning_rate": 0.0002214,
-      "loss": 4.4292,
+      "loss": 4.4607,
       "step": 372
     },
     {
       "epoch": 0.84,
       "learning_rate": 0.00022199999999999998,
-      "loss": 4.3984,
+      "loss": 4.4477,
       "step": 373
     },
     {
       "epoch": 0.84,
       "learning_rate": 0.0002226,
-      "loss": 4.2254,
+      "loss": 4.2569,
       "step": 374
     },
     {
       "epoch": 0.84,
       "learning_rate": 0.00022319999999999998,
-      "loss": 4.4513,
+      "loss": 4.4793,
       "step": 375
     },
     {
       "epoch": 0.84,
       "learning_rate": 0.0002238,
-      "loss": 4.3402,
+      "loss": 4.3588,
       "step": 376
     },
     {
       "epoch": 0.85,
       "learning_rate": 0.00022439999999999998,
-      "loss": 4.4062,
+      "loss": 4.4029,
       "step": 377
     },
     {
       "epoch": 0.85,
       "learning_rate": 0.000225,
-      "loss": 4.3849,
+      "loss": 4.3932,
       "step": 378
     },
     {
       "epoch": 0.85,
       "learning_rate": 0.00022559999999999998,
-      "loss": 4.2675,
+      "loss": 4.2944,
       "step": 379
     },
     {
       "epoch": 0.85,
       "learning_rate": 0.00022619999999999997,
-      "loss": 4.3187,
+      "loss": 4.3351,
       "step": 380
     },
     {
       "epoch": 0.85,
       "learning_rate": 0.00022679999999999998,
-      "loss": 4.3882,
+      "loss": 4.4359,
       "step": 381
     },
     {
       "epoch": 0.86,
       "learning_rate": 0.00022739999999999997,
-      "loss": 4.426,
+      "loss": 4.4558,
       "step": 382
     },
     {
       "epoch": 0.86,
       "learning_rate": 0.00022799999999999999,
-      "loss": 4.3767,
+      "loss": 4.3975,
       "step": 383
     },
     {
       "epoch": 0.86,
       "learning_rate": 0.00022859999999999997,
-      "loss": 4.3028,
+      "loss": 4.308,
       "step": 384
     },
     {
       "epoch": 0.86,
       "learning_rate": 0.0002292,
-      "loss": 4.3851,
+      "loss": 4.4021,
       "step": 385
     },
     {
       "epoch": 0.87,
       "learning_rate": 0.00022979999999999997,
-      "loss": 4.2933,
+      "loss": 4.2904,
       "step": 386
     },
     {
       "epoch": 0.87,
       "learning_rate": 0.0002304,
-      "loss": 4.4037,
+      "loss": 4.4339,
       "step": 387
     },
     {
       "epoch": 0.87,
       "learning_rate": 0.00023099999999999998,
-      "loss": 4.4482,
+      "loss": 4.4791,
       "step": 388
     },
     {
       "epoch": 0.87,
       "learning_rate": 0.0002316,
-      "loss": 4.3396,
+      "loss": 4.3777,
       "step": 389
     },
     {
       "epoch": 0.87,
       "learning_rate": 0.00023219999999999998,
-      "loss": 4.3461,
+      "loss": 4.3603,
       "step": 390
     },
     {
       "epoch": 0.88,
       "learning_rate": 0.0002328,
-      "loss": 4.4515,
+      "loss": 4.478,
       "step": 391
     },
     {
       "epoch": 0.88,
       "learning_rate": 0.00023339999999999998,
-      "loss": 4.2404,
+      "loss": 4.2619,
       "step": 392
     },
     {
       "epoch": 0.88,
       "learning_rate": 0.000234,
-      "loss": 4.1869,
+      "loss": 4.2223,
       "step": 393
     },
     {
       "epoch": 0.88,
       "learning_rate": 0.00023459999999999998,
-      "loss": 4.2223,
+      "loss": 4.2092,
       "step": 394
     },
     {
       "epoch": 0.89,
       "learning_rate": 0.0002352,
-      "loss": 4.3675,
+      "loss": 4.3659,
       "step": 395
     },
     {
       "epoch": 0.89,
       "learning_rate": 0.00023579999999999999,
-      "loss": 4.2826,
+      "loss": 4.2694,
       "step": 396
     },
     {
       "epoch": 0.89,
       "learning_rate": 0.0002364,
-      "loss": 4.4047,
+      "loss": 4.3929,
       "step": 397
     },
     {
       "epoch": 0.89,
       "learning_rate": 0.000237,
-      "loss": 4.1577,
+      "loss": 4.1923,
       "step": 398
     },
     {
       "epoch": 0.89,
       "learning_rate": 0.0002376,
-      "loss": 4.2136,
+      "loss": 4.1854,
       "step": 399
     },
     {
       "epoch": 0.9,
       "learning_rate": 0.0002382,
-      "loss": 4.0011,
+      "loss": 4.0039,
       "step": 400
     },
     {
       "epoch": 0.9,
       "learning_rate": 0.0002388,
-      "loss": 5.3932,
+      "loss": 5.8865,
       "step": 401
     },
     {
       "epoch": 0.9,
       "learning_rate": 0.0002394,
-      "loss": 4.8733,
+      "loss": 5.6627,
       "step": 402
     },
     {
       "epoch": 0.9,
       "learning_rate": 0.00023999999999999998,
-      "loss": 4.7643,
+      "loss": 5.4797,
       "step": 403
     },
     {
       "epoch": 0.91,
       "learning_rate": 0.0002406,
-      "loss": 4.5921,
+      "loss": 5.0418,
       "step": 404
     },
     {
       "epoch": 0.91,
       "learning_rate": 0.00024119999999999998,
-      "loss": 4.5268,
+      "loss": 4.8755,
       "step": 405
     },
     {
       "epoch": 0.91,
       "learning_rate": 0.0002418,
-      "loss": 4.5454,
+      "loss": 4.7933,
       "step": 406
     },
     {
       "epoch": 0.91,
       "learning_rate": 0.00024239999999999998,
-      "loss": 4.5719,
+      "loss": 4.7055,
       "step": 407
     },
     {
       "epoch": 0.91,
       "learning_rate": 0.000243,
-      "loss": 4.5135,
+      "loss": 4.6617,
       "step": 408
     },
     {
       "epoch": 0.92,
       "learning_rate": 0.00024359999999999999,
-      "loss": 4.4509,
+      "loss": 4.5539,
       "step": 409
     },
     {
       "epoch": 0.92,
       "learning_rate": 0.00024419999999999997,
-      "loss": 4.395,
+      "loss": 4.492,
       "step": 410
     },
     {
       "epoch": 0.92,
       "learning_rate": 0.0002448,
-      "loss": 4.5676,
+      "loss": 4.6706,
       "step": 411
     },
     {
       "epoch": 0.92,
       "learning_rate": 0.00024539999999999995,
-      "loss": 4.3981,
+      "loss": 4.4853,
       "step": 412
     },
     {
       "epoch": 0.93,
       "learning_rate": 0.00024599999999999996,
-      "loss": 4.3338,
+      "loss": 4.4369,
       "step": 413
     },
     {
       "epoch": 0.93,
       "learning_rate": 0.0002466,
-      "loss": 4.4346,
+      "loss": 4.4848,
       "step": 414
     },
     {
       "epoch": 0.93,
       "learning_rate": 0.0002472,
-      "loss": 4.4811,
+      "loss": 4.5455,
       "step": 415
     },
     {
       "epoch": 0.93,
       "learning_rate": 0.00024779999999999995,
-      "loss": 4.3513,
+      "loss": 4.4175,
       "step": 416
     },
     {
       "epoch": 0.93,
       "learning_rate": 0.00024839999999999997,
-      "loss": 4.3441,
+      "loss": 4.4364,
       "step": 417
     },
     {
       "epoch": 0.94,
       "learning_rate": 0.000249,
-      "loss": 4.2663,
+      "loss": 4.3047,
       "step": 418
     },
     {
       "epoch": 0.94,
       "learning_rate": 0.00024959999999999994,
-      "loss": 4.3528,
+      "loss": 4.4146,
       "step": 419
     },
     {
       "epoch": 0.94,
       "learning_rate": 0.00025019999999999996,
-      "loss": 4.2688,
+      "loss": 4.3201,
       "step": 420
     },
     {
       "epoch": 0.94,
       "learning_rate": 0.00025079999999999997,
-      "loss": 4.3758,
+      "loss": 4.4562,
       "step": 421
     },
     {
       "epoch": 0.95,
       "learning_rate": 0.0002514,
-      "loss": 4.2928,
+      "loss": 4.3312,
       "step": 422
     },
     {
       "epoch": 0.95,
       "learning_rate": 0.00025199999999999995,
-      "loss": 4.3609,
+      "loss": 4.4033,
       "step": 423
     },
     {
       "epoch": 0.95,
       "learning_rate": 0.00025259999999999996,
-      "loss": 4.4505,
+      "loss": 4.4837,
       "step": 424
     },
     {
       "epoch": 0.95,
       "learning_rate": 0.0002532,
-      "loss": 4.3456,
+      "loss": 4.3916,
       "step": 425
     },
     {
       "epoch": 0.96,
       "learning_rate": 0.0002538,
-      "loss": 4.407,
+      "loss": 4.405,
       "step": 426
     },
     {
       "epoch": 0.96,
       "learning_rate": 0.00025439999999999995,
-      "loss": 4.2921,
+      "loss": 4.3168,
       "step": 427
     },
     {
       "epoch": 0.96,
       "learning_rate": 0.00025499999999999996,
-      "loss": 4.2807,
+      "loss": 4.3303,
       "step": 428
     },
     {
       "epoch": 0.96,
       "learning_rate": 0.0002556,
-      "loss": 4.1697,
+      "loss": 4.2059,
       "step": 429
     },
     {
       "epoch": 0.96,
       "learning_rate": 0.0002562,
-      "loss": 4.3221,
+      "loss": 4.3448,
       "step": 430
     },
     {
       "epoch": 0.97,
       "learning_rate": 0.00025679999999999995,
-      "loss": 4.2951,
+      "loss": 4.3357,
       "step": 431
     },
     {
       "epoch": 0.97,
       "learning_rate": 0.00025739999999999997,
-      "loss": 4.4834,
+      "loss": 4.5314,
       "step": 432
     },
     {
       "epoch": 0.97,
       "learning_rate": 0.000258,
-      "loss": 4.3382,
+      "loss": 4.3478,
       "step": 433
     },
     {
       "epoch": 0.97,
       "learning_rate": 0.0002586,
-      "loss": 4.2568,
+      "loss": 4.2702,
       "step": 434
     },
     {
       "epoch": 0.98,
       "learning_rate": 0.00025919999999999996,
-      "loss": 4.3072,
+      "loss": 4.3382,
       "step": 435
     },
     {
       "epoch": 0.98,
       "learning_rate": 0.00025979999999999997,
-      "loss": 4.2,
+      "loss": 4.2302,
       "step": 436
     },
     {
       "epoch": 0.98,
       "learning_rate": 0.0002604,
-      "loss": 4.1902,
+      "loss": 4.2964,
       "step": 437
     },
     {
       "epoch": 0.98,
       "learning_rate": 0.000261,
-      "loss": 4.2883,
+      "loss": 4.3367,
       "step": 438
     },
     {
       "epoch": 0.98,
       "learning_rate": 0.00026159999999999996,
-      "loss": 4.2476,
+      "loss": 4.3053,
       "step": 439
     },
     {
       "epoch": 0.99,
       "learning_rate": 0.0002622,
-      "loss": 4.2958,
+      "loss": 4.3341,
       "step": 440
     },
     {
       "epoch": 0.99,
       "learning_rate": 0.0002628,
-      "loss": 4.1187,
+      "loss": 4.2008,
       "step": 441
     },
     {
       "epoch": 0.99,
       "learning_rate": 0.00026339999999999995,
-      "loss": 4.3498,
+      "loss": 4.3948,
       "step": 442
     },
     {
       "epoch": 0.99,
       "learning_rate": 0.00026399999999999997,
-      "loss": 4.0483,
+      "loss": 4.0869,
       "step": 443
     },
     {
       "epoch": 1.0,
       "learning_rate": 0.0002646,
-      "loss": 3.9359,
+      "loss": 3.9334,
       "step": 444
     },
     {
       "epoch": 1.0,
       "learning_rate": 0.0002652,
-      "loss": 4.0019,
+      "loss": 4.0116,
       "step": 445
     },
     {
       "epoch": 1.0,
       "learning_rate": 0.00026579999999999996,
-      "loss": 3.5722,
+      "loss": 3.6733,
       "step": 446
     },
     {
       "epoch": 1.0,
       "learning_rate": 0.00026639999999999997,
-      "loss": 5.523,
+      "loss": 5.8448,
       "step": 447
     },
     {
       "epoch": 1.0,
       "learning_rate": 0.000267,
-      "loss": 5.164,
+      "loss": 5.7358,
       "step": 448
     },
     {
       "epoch": 1.01,
       "learning_rate": 0.0002676,
-      "loss": 4.8554,
+      "loss": 5.3335,
       "step": 449
     },
     {
       "epoch": 1.01,
       "learning_rate": 0.00026819999999999996,
-      "loss": 4.6654,
+      "loss": 4.9919,
       "step": 450
     },
     {
       "epoch": 1.01,
       "learning_rate": 0.0002688,
-      "loss": 4.5437,
+      "loss": 4.7456,
       "step": 451
     },
     {
       "epoch": 1.01,
       "learning_rate": 0.0002694,
-      "loss": 4.4698,
+      "loss": 4.6112,
       "step": 452
     },
     {
       "epoch": 1.02,
       "learning_rate": 0.00027,
-      "loss": 4.2867,
+      "loss": 4.4133,
       "step": 453
     },
     {
       "epoch": 1.02,
       "learning_rate": 0.00027059999999999996,
-      "loss": 4.3267,
+      "loss": 4.3851,
       "step": 454
     },
     {
       "epoch": 1.02,
       "learning_rate": 0.0002712,
-      "loss": 4.2144,
+      "loss": 4.2591,
       "step": 455
     },
     {
       "epoch": 1.02,
       "learning_rate": 0.0002718,
-      "loss": 4.2471,
+      "loss": 4.3699,
       "step": 456
     },
     {
       "epoch": 1.02,
       "learning_rate": 0.0002724,
-      "loss": 4.0365,
+      "loss": 4.1498,
       "step": 457
     },
     {
       "epoch": 1.03,
       "learning_rate": 0.00027299999999999997,
-      "loss": 3.9685,
+      "loss": 4.1313,
       "step": 458
     },
     {
       "epoch": 1.03,
       "learning_rate": 0.0002736,
-      "loss": 4.1301,
+      "loss": 4.2658,
       "step": 459
     },
     {
       "epoch": 1.03,
       "learning_rate": 0.0002742,
-      "loss": 4.063,
+      "loss": 4.1512,
       "step": 460
     },
     {
       "epoch": 1.03,
       "learning_rate": 0.0002748,
-      "loss": 4.067,
+      "loss": 4.1147,
       "step": 461
     },
     {
       "epoch": 1.04,
       "learning_rate": 0.00027539999999999997,
-      "loss": 3.9844,
+      "loss": 4.0338,
       "step": 462
     },
     {
       "epoch": 1.04,
       "learning_rate": 0.000276,
-      "loss": 4.0705,
+      "loss": 4.0853,
       "step": 463
     },
     {
       "epoch": 1.04,
       "learning_rate": 0.0002766,
-      "loss": 3.9171,
+      "loss": 4.0146,
       "step": 464
     },
     {
       "epoch": 1.04,
       "learning_rate": 0.0002772,
-      "loss": 3.9141,
+      "loss": 3.9486,
       "step": 465
     },
     {
       "epoch": 1.04,
       "learning_rate": 0.0002778,
-      "loss": 3.9032,
+      "loss": 3.9637,
       "step": 466
     },
     {
       "epoch": 1.05,
       "learning_rate": 0.0002784,
-      "loss": 4.0287,
+      "loss": 4.0797,
       "step": 467
     },
     {
       "epoch": 1.05,
       "learning_rate": 0.000279,
-      "loss": 3.8513,
+      "loss": 3.9126,
       "step": 468
     },
     {
       "epoch": 1.05,
       "learning_rate": 0.00027959999999999997,
-      "loss": 3.8698,
+      "loss": 3.9294,
       "step": 469
     },
     {
       "epoch": 1.05,
       "learning_rate": 0.0002802,
-      "loss": 4.0594,
+      "loss": 4.0885,
       "step": 470
     },
     {
       "epoch": 1.06,
       "learning_rate": 0.0002808,
-      "loss": 3.9681,
+      "loss": 4.0577,
       "step": 471
     },
     {
       "epoch": 1.06,
       "learning_rate": 0.00028139999999999996,
-      "loss": 3.9133,
+      "loss": 3.9314,
       "step": 472
     },
     {
       "epoch": 1.06,
       "learning_rate": 0.00028199999999999997,
-      "loss": 3.8527,
+      "loss": 3.9632,
       "step": 473
     },
     {
       "epoch": 1.06,
       "learning_rate": 0.0002826,
-      "loss": 3.7028,
+      "loss": 3.7662,
       "step": 474
     },
     {
       "epoch": 1.07,
       "learning_rate": 0.00028319999999999994,
-      "loss": 3.678,
+      "loss": 3.7393,
       "step": 475
     },
     {
       "epoch": 1.07,
       "learning_rate": 0.00028379999999999996,
-      "loss": 3.8978,
+      "loss": 3.9623,
       "step": 476
     },
     {
       "epoch": 1.07,
       "learning_rate": 0.0002844,
-      "loss": 3.7861,
+      "loss": 3.8472,
       "step": 477
     },
     {
       "epoch": 1.07,
       "learning_rate": 0.000285,
-      "loss": 3.8226,
+      "loss": 3.8946,
       "step": 478
     },
     {
       "epoch": 1.07,
       "learning_rate": 0.00028559999999999995,
-      "loss": 3.8174,
+      "loss": 3.9246,
       "step": 479
     },
     {
       "epoch": 1.08,
       "learning_rate": 0.00028619999999999996,
-      "loss": 3.8163,
+      "loss": 3.9423,
       "step": 480
     },
     {
       "epoch": 1.08,
       "learning_rate": 0.0002868,
-      "loss": 3.9948,
+      "loss": 4.0871,
       "step": 481
     },
     {
       "epoch": 1.08,
       "learning_rate": 0.00028739999999999994,
-      "loss": 3.752,
+      "loss": 3.7827,
       "step": 482
     },
     {
       "epoch": 1.08,
       "learning_rate": 0.00028799999999999995,
-      "loss": 3.8112,
+      "loss": 3.8545,
       "step": 483
     },
     {
       "epoch": 1.09,
       "learning_rate": 0.00028859999999999997,
-      "loss": 3.7391,
+      "loss": 3.8231,
       "step": 484
     },
     {
       "epoch": 1.09,
       "learning_rate": 0.0002892,
-      "loss": 3.7673,
+      "loss": 3.8769,
       "step": 485
     },
     {
       "epoch": 1.09,
       "learning_rate": 0.00028979999999999994,
-      "loss": 3.6326,
+      "loss": 3.6917,
       "step": 486
     },
     {
       "epoch": 1.09,
       "learning_rate": 0.00029039999999999996,
-      "loss": 3.8127,
+      "loss": 3.971,
       "step": 487
     },
     {
       "epoch": 1.09,
       "learning_rate": 0.00029099999999999997,
-      "loss": 3.6941,
+      "loss": 3.7305,
       "step": 488
     },
     {
       "epoch": 1.1,
       "learning_rate": 0.0002916,
-      "loss": 3.8133,
+      "loss": 3.9215,
       "step": 489
     },
     {
       "epoch": 1.1,
       "learning_rate": 0.00029219999999999995,
-      "loss": 3.5072,
+      "loss": 3.6148,
       "step": 490
     },
     {
       "epoch": 1.1,
       "learning_rate": 0.00029279999999999996,
-      "loss": 3.5734,
+      "loss": 3.634,
       "step": 491
     },
     {
       "epoch": 1.1,
       "learning_rate": 0.0002934,
-      "loss": 3.438,
+      "loss": 3.4109,
       "step": 492
     },
     {
       "epoch": 1.11,
       "learning_rate": 0.000294,
-      "loss": 3.3375,
+      "loss": 3.3743,
       "step": 493
     },
     {
       "epoch": 1.11,
       "learning_rate": 0.00029459999999999995,
-      "loss": 3.4771,
+      "loss": 3.5371,
       "step": 494
     },
     {
       "epoch": 1.11,
       "learning_rate": 0.00029519999999999997,
-      "loss": 3.0575,
+      "loss": 3.1645,
       "step": 495
     },
     {
       "epoch": 1.11,
       "learning_rate": 0.0002958,
-      "loss": 2.8942,
+      "loss": 2.9923,
       "step": 496
     },
     {
       "epoch": 1.11,
       "learning_rate": 0.0002964,
-      "loss": 5.571,
+      "loss": 6.0733,
       "step": 497
     },
     {
       "epoch": 1.12,
       "learning_rate": 0.00029699999999999996,
-      "loss": 5.1212,
+      "loss": 5.432,
       "step": 498
     },
     {
       "epoch": 1.12,
       "learning_rate": 0.00029759999999999997,
-      "loss": 4.784,
+      "loss": 4.9446,
       "step": 499
     },
     {
       "epoch": 1.12,
       "learning_rate": 0.0002982,
-      "loss": 4.4678,
+      "loss": 4.6521,
       "step": 500
     },
     {
       "epoch": 1.12,
-      "eval_loss": 4.352957248687744,
-      "eval_runtime": 602.6395,
-      "eval_samples_per_second": 4.384,
-      "eval_steps_per_second": 0.549,
-      "eval_wer": 1.7716977389924633,
+      "eval_loss": 4.444121360778809,
+      "eval_runtime": 583.2794,
+      "eval_samples_per_second": 4.53,
+      "eval_steps_per_second": 0.567,
+      "eval_wer": 1.7011503371677905,
       "step": 500
     },
     {
       "epoch": 1.12,
       "learning_rate": 0.0002988,
-      "loss": 4.2885,
+      "loss": 4.414,
       "step": 501
     },
     {
       "epoch": 1.13,
       "learning_rate": 0.00029939999999999996,
-      "loss": 4.3183,
+      "loss": 4.4024,
       "step": 502
     },
     {
       "epoch": 1.13,
       "learning_rate": 0.0003,
-      "loss": 4.2899,
+      "loss": 4.3597,
       "step": 503
     },
     {
       "epoch": 1.13,
       "learning_rate": 0.00029982658959537567,
-      "loss": 4.0863,
+      "loss": 4.1466,
       "step": 504
     },
     {
       "epoch": 1.13,
       "learning_rate": 0.0002996531791907514,
-      "loss": 4.1456,
+      "loss": 4.204,
       "step": 505
     },
     {
       "epoch": 1.13,
       "learning_rate": 0.00029947976878612716,
-      "loss": 4.2061,
+      "loss": 4.296,
       "step": 506
     },
     {
       "epoch": 1.14,
       "learning_rate": 0.00029930635838150286,
-      "loss": 4.041,
+      "loss": 4.1418,
       "step": 507
     },
     {
       "epoch": 1.14,
       "learning_rate": 0.0002991329479768786,
-      "loss": 4.2261,
+      "loss": 4.3272,
       "step": 508
     },
     {
       "epoch": 1.14,
       "learning_rate": 0.0002989595375722543,
-      "loss": 4.072,
+      "loss": 4.1266,
       "step": 509
     },
     {
       "epoch": 1.14,
       "learning_rate": 0.00029878612716763005,
-      "loss": 4.0255,
+      "loss": 4.1107,
       "step": 510
     },
     {
       "epoch": 1.15,
       "learning_rate": 0.00029861271676300574,
-      "loss": 3.9969,
+      "loss": 4.0438,
       "step": 511
     },
     {
       "epoch": 1.15,
       "learning_rate": 0.0002984393063583815,
-      "loss": 4.0715,
+      "loss": 4.0974,
       "step": 512
     },
     {
       "epoch": 1.15,
       "learning_rate": 0.0002982658959537572,
-      "loss": 4.0107,
+      "loss": 4.07,
       "step": 513
     },
     {
       "epoch": 1.15,
       "learning_rate": 0.00029809248554913293,
-      "loss": 3.8789,
+      "loss": 3.9892,
       "step": 514
     },
     {
       "epoch": 1.15,
       "learning_rate": 0.0002979190751445086,
-      "loss": 4.014,
+      "loss": 4.0635,
       "step": 515
     },
     {
       "epoch": 1.16,
       "learning_rate": 0.00029774566473988437,
-      "loss": 4.005,
+      "loss": 4.0718,
       "step": 516
     },
     {
       "epoch": 1.16,
       "learning_rate": 0.00029757225433526006,
-      "loss": 3.9039,
+      "loss": 3.99,
       "step": 517
     },
     {
       "epoch": 1.16,
       "learning_rate": 0.0002973988439306358,
-      "loss": 3.9395,
+      "loss": 4.0126,
       "step": 518
     },
     {
       "epoch": 1.16,
       "learning_rate": 0.00029722543352601156,
-      "loss": 3.9881,
+      "loss": 4.0171,
       "step": 519
     },
     {
       "epoch": 1.17,
       "learning_rate": 0.00029705202312138725,
-      "loss": 3.9593,
+      "loss": 4.0175,
       "step": 520
     },
     {
       "epoch": 1.17,
       "learning_rate": 0.00029687861271676295,
-      "loss": 3.8773,
+      "loss": 3.9543,
       "step": 521
     },
     {
       "epoch": 1.17,
       "learning_rate": 0.0002967052023121387,
-      "loss": 3.8301,
+      "loss": 3.9275,
       "step": 522
     },
     {
       "epoch": 1.17,
       "learning_rate": 0.00029653179190751444,
-      "loss": 3.8361,
+      "loss": 3.97,
       "step": 523
     },
     {
       "epoch": 1.17,
       "learning_rate": 0.00029635838150289014,
-      "loss": 3.9224,
+      "loss": 4.0399,
       "step": 524
     },
     {
       "epoch": 1.18,
       "learning_rate": 0.00029618497109826583,
-      "loss": 3.8038,
+      "loss": 3.8694,
       "step": 525
     },
     {
       "epoch": 1.18,
       "learning_rate": 0.0002960115606936416,
-      "loss": 3.765,
+      "loss": 3.8722,
       "step": 526
     },
     {
       "epoch": 1.18,
       "learning_rate": 0.0002958381502890173,
-      "loss": 3.7414,
+      "loss": 3.8752,
       "step": 527
     },
     {
       "epoch": 1.18,
       "learning_rate": 0.000295664739884393,
-      "loss": 3.8548,
+      "loss": 3.9737,
       "step": 528
     },
     {
       "epoch": 1.19,
       "learning_rate": 0.00029549132947976877,
-      "loss": 3.6642,
+      "loss": 3.7855,
       "step": 529
     },
     {
       "epoch": 1.19,
       "learning_rate": 0.0002953179190751445,
-      "loss": 3.7058,
+      "loss": 3.8993,
       "step": 530
     },
     {
       "epoch": 1.19,
       "learning_rate": 0.0002951445086705202,
-      "loss": 3.7377,
+      "loss": 3.9065,
       "step": 531
     },
     {
       "epoch": 1.19,
       "learning_rate": 0.0002949710982658959,
-      "loss": 3.809,
+      "loss": 4.0046,
       "step": 532
     },
     {
       "epoch": 1.2,
       "learning_rate": 0.00029479768786127165,
-      "loss": 3.7202,
+      "loss": 3.8726,
       "step": 533
     },
     {
       "epoch": 1.2,
       "learning_rate": 0.0002946242774566474,
-      "loss": 3.779,
+      "loss": 3.9889,
       "step": 534
     },
     {
       "epoch": 1.2,
       "learning_rate": 0.0002944508670520231,
-      "loss": 3.6233,
+      "loss": 3.8137,
       "step": 535
     },
     {
       "epoch": 1.2,
       "learning_rate": 0.00029427745664739884,
-      "loss": 3.5324,
+      "loss": 3.712,
       "step": 536
     },
     {
       "epoch": 1.2,
       "learning_rate": 0.00029410404624277453,
-      "loss": 3.6339,
+      "loss": 3.8098,
       "step": 537
     },
     {
       "epoch": 1.21,
       "learning_rate": 0.0002939306358381503,
-      "loss": 3.4023,
+      "loss": 3.5976,
       "step": 538
     },
     {
       "epoch": 1.21,
       "learning_rate": 0.00029375722543352597,
-      "loss": 3.4368,
+      "loss": 3.6913,
       "step": 539
     },
     {
       "epoch": 1.21,
       "learning_rate": 0.0002935838150289017,
-      "loss": 3.4439,
+      "loss": 3.7072,
       "step": 540
     },
     {
       "epoch": 1.21,
       "learning_rate": 0.0002934104046242774,
-      "loss": 3.3336,
+      "loss": 3.6075,
       "step": 541
     },
     {
       "epoch": 1.22,
       "learning_rate": 0.00029323699421965316,
-      "loss": 3.2701,
+      "loss": 3.5426,
       "step": 542
     },
     {
       "epoch": 1.22,
       "learning_rate": 0.00029306358381502886,
-      "loss": 3.1287,
+      "loss": 3.4171,
       "step": 543
     },
     {
       "epoch": 1.22,
       "learning_rate": 0.0002928901734104046,
-      "loss": 2.9568,
+      "loss": 3.1822,
       "step": 544
     },
     {
       "epoch": 1.22,
       "learning_rate": 0.0002927167630057803,
-      "loss": 2.9493,
+      "loss": 3.3456,
       "step": 545
     },
     {
       "epoch": 1.22,
       "learning_rate": 0.00029254335260115604,
-      "loss": 2.5995,
+      "loss": 2.927,
       "step": 546
     },
     {
       "epoch": 1.23,
       "learning_rate": 0.0002923699421965318,
-      "loss": 6.2992,
+      "loss": 5.8096,
       "step": 547
     },
     {
       "epoch": 1.23,
       "learning_rate": 0.0002921965317919075,
-      "loss": 5.2557,
+      "loss": 5.1263,
       "step": 548
     },
     {
       "epoch": 1.23,
       "learning_rate": 0.0002920231213872832,
-      "loss": 4.6328,
+      "loss": 4.7905,
       "step": 549
     },
     {
       "epoch": 1.23,
       "learning_rate": 0.0002918497109826589,
-      "loss": 4.4602,
+      "loss": 4.6619,
       "step": 550
     },
     {
       "epoch": 1.24,
       "learning_rate": 0.0002916763005780347,
-      "loss": 4.3019,
+      "loss": 4.4531,
       "step": 551
     },
     {
       "epoch": 1.24,
       "learning_rate": 0.00029150289017341037,
-      "loss": 4.1883,
+      "loss": 4.3329,
       "step": 552
     },
     {
       "epoch": 1.24,
       "learning_rate": 0.0002913294797687861,
-      "loss": 4.2302,
+      "loss": 4.4172,
       "step": 553
     },
     {
       "epoch": 1.24,
       "learning_rate": 0.00029115606936416186,
-      "loss": 4.0509,
+      "loss": 4.2514,
       "step": 554
     },
     {
       "epoch": 1.24,
       "learning_rate": 0.00029098265895953756,
-      "loss": 4.0135,
+      "loss": 4.2204,
       "step": 555
     },
     {
       "epoch": 1.25,
       "learning_rate": 0.00029080924855491325,
-      "loss": 3.8072,
+      "loss": 4.0057,
       "step": 556
     },
     {
       "epoch": 1.25,
       "learning_rate": 0.000290635838150289,
-      "loss": 3.8271,
+      "loss": 4.0476,
       "step": 557
     },
     {
       "epoch": 1.25,
       "learning_rate": 0.00029046242774566475,
-      "loss": 3.8352,
+      "loss": 4.0568,
       "step": 558
     },
     {
       "epoch": 1.25,
       "learning_rate": 0.00029028901734104044,
-      "loss": 3.7885,
+      "loss": 4.0171,
       "step": 559
     },
     {
       "epoch": 1.26,
       "learning_rate": 0.00029011560693641613,
-      "loss": 3.8345,
+      "loss": 4.0169,
       "step": 560
     },
     {
       "epoch": 1.26,
       "learning_rate": 0.0002899421965317919,
-      "loss": 3.7898,
+      "loss": 4.0429,
       "step": 561
     },
     {
       "epoch": 1.26,
       "learning_rate": 0.00028976878612716763,
-      "loss": 3.6793,
+      "loss": 3.8805,
       "step": 562
     },
     {
       "epoch": 1.26,
       "learning_rate": 0.0002895953757225433,
-      "loss": 3.7274,
+      "loss": 4.0099,
       "step": 563
     },
     {
       "epoch": 1.26,
       "learning_rate": 0.00028942196531791907,
-      "loss": 3.5994,
+      "loss": 3.8666,
       "step": 564
     },
     {
       "epoch": 1.27,
       "learning_rate": 0.00028924855491329476,
-      "loss": 3.5573,
+      "loss": 3.8508,
       "step": 565
     },
     {
       "epoch": 1.27,
       "learning_rate": 0.0002890751445086705,
-      "loss": 3.5576,
+      "loss": 3.8488,
       "step": 566
     },
     {
       "epoch": 1.27,
       "learning_rate": 0.0002889017341040462,
-      "loss": 3.577,
+      "loss": 3.8655,
       "step": 567
     },
     {
       "epoch": 1.27,
       "learning_rate": 0.00028872832369942195,
-      "loss": 3.6755,
+      "loss": 4.0298,
       "step": 568
     },
     {
       "epoch": 1.28,
       "learning_rate": 0.00028855491329479765,
-      "loss": 3.39,
+      "loss": 3.8001,
       "step": 569
     },
     {
       "epoch": 1.28,
       "learning_rate": 0.0002883815028901734,
-      "loss": 3.5657,
+      "loss": 3.8585,
       "step": 570
     },
     {
       "epoch": 1.28,
       "learning_rate": 0.0002882080924855491,
-      "loss": 3.4084,
+      "loss": 3.7756,
       "step": 571
     },
     {
       "epoch": 1.28,
       "learning_rate": 0.00028803468208092484,
-      "loss": 3.5125,
+      "loss": 3.9276,
       "step": 572
     },
     {
       "epoch": 1.28,
       "learning_rate": 0.00028786127167630053,
-      "loss": 3.4383,
+      "loss": 3.8585,
       "step": 573
     },
     {
       "epoch": 1.29,
       "learning_rate": 0.0002876878612716763,
-      "loss": 3.3082,
+      "loss": 3.7575,
       "step": 574
     },
     {
       "epoch": 1.29,
       "learning_rate": 0.000287514450867052,
-      "loss": 3.299,
+      "loss": 3.7872,
       "step": 575
     },
     {
       "epoch": 1.29,
       "learning_rate": 0.0002873410404624277,
-      "loss": 3.323,
+      "loss": 3.8801,
       "step": 576
     },
     {
       "epoch": 1.29,
       "learning_rate": 0.0002871676300578034,
-      "loss": 3.3031,
+      "loss": 3.9001,
       "step": 577
     },
     {
       "epoch": 1.3,
       "learning_rate": 0.00028699421965317916,
-      "loss": 3.1504,
+      "loss": 3.7872,
       "step": 578
     },
     {
       "epoch": 1.3,
       "learning_rate": 0.0002868208092485549,
-      "loss": 3.2823,
+      "loss": 3.7781,
       "step": 579
     },
     {
       "epoch": 1.3,
       "learning_rate": 0.0002866473988439306,
-      "loss": 3.182,
+      "loss": 3.8367,
       "step": 580
     },
     {
       "epoch": 1.3,
       "learning_rate": 0.00028647398843930635,
-      "loss": 3.3284,
+      "loss": 3.966,
       "step": 581
     },
     {
       "epoch": 1.3,
       "learning_rate": 0.0002863005780346821,
-      "loss": 3.1441,
+      "loss": 3.8266,
       "step": 582
     },
     {
       "epoch": 1.31,
       "learning_rate": 0.0002861271676300578,
-      "loss": 3.0156,
+      "loss": 3.8546,
       "step": 583
     },
     {
       "epoch": 1.31,
       "learning_rate": 0.0002859537572254335,
-      "loss": 2.8212,
+      "loss": 3.6754,
       "step": 584
     },
     {
       "epoch": 1.31,
       "learning_rate": 0.00028578034682080923,
-      "loss": 2.73,
+      "loss": 3.5941,
       "step": 585
     },
     {
       "epoch": 1.31,
       "learning_rate": 0.000285606936416185,
-      "loss": 2.8376,
+      "loss": 3.7293,
       "step": 586
     },
     {
       "epoch": 1.32,
       "learning_rate": 0.0002854335260115607,
-      "loss": 2.8082,
+      "loss": 3.6958,
       "step": 587
     },
     {
       "epoch": 1.32,
       "learning_rate": 0.00028526011560693637,
-      "loss": 2.5886,
+      "loss": 3.5926,
       "step": 588
     },
     {
       "epoch": 1.32,
       "learning_rate": 0.0002850867052023121,
-      "loss": 2.344,
+      "loss": 3.3374,
       "step": 589
     },
     {
       "epoch": 1.32,
       "learning_rate": 0.00028491329479768786,
-      "loss": 2.2516,
+      "loss": 3.3626,
       "step": 590
     },
     {
       "epoch": 1.33,
       "learning_rate": 0.00028473988439306356,
-      "loss": 2.307,
+      "loss": 3.4774,
       "step": 591
     },
     {
       "epoch": 1.33,
       "learning_rate": 0.0002845664739884393,
-      "loss": 2.0804,
+      "loss": 3.2165,
       "step": 592
     },
     {
       "epoch": 1.33,
       "learning_rate": 0.000284393063583815,
-      "loss": 1.7886,
+      "loss": 2.9479,
       "step": 593
     },
     {
       "epoch": 1.33,
       "learning_rate": 0.00028421965317919075,
-      "loss": 1.9756,
+      "loss": 2.984,
       "step": 594
     },
     {
       "epoch": 1.33,
       "learning_rate": 0.00028404624277456644,
-      "loss": 2.025,
+      "loss": 3.0605,
       "step": 595
     },
     {
       "epoch": 1.34,
       "learning_rate": 0.0002838728323699422,
-      "loss": 1.888,
+      "loss": 2.7119,
       "step": 596
     },
     {
       "epoch": 1.34,
       "learning_rate": 0.0002836994219653179,
-      "loss": 6.5668,
+      "loss": 5.6219,
       "step": 597
     },
     {
       "epoch": 1.34,
       "learning_rate": 0.00028352601156069363,
-      "loss": 5.3153,
+      "loss": 5.3126,
       "step": 598
     },
     {
       "epoch": 1.34,
       "learning_rate": 0.0002833526011560693,
-      "loss": 4.1101,
+      "loss": 4.943,
       "step": 599
     },
     {
       "epoch": 1.35,
       "learning_rate": 0.00028317919075144507,
-      "loss": 3.7808,
+      "loss": 4.7358,
       "step": 600
     },
     {
       "epoch": 1.35,
       "learning_rate": 0.00028300578034682076,
-      "loss": 3.4887,
+      "loss": 4.3826,
       "step": 601
     },
     {
       "epoch": 1.35,
       "learning_rate": 0.0002828323699421965,
-      "loss": 3.5204,
+      "loss": 4.3389,
       "step": 602
     },
     {
       "epoch": 1.35,
       "learning_rate": 0.00028265895953757226,
-      "loss": 3.3659,
+      "loss": 4.2211,
       "step": 603
     },
     {
       "epoch": 1.35,
       "learning_rate": 0.00028248554913294795,
-      "loss": 3.2103,
+      "loss": 4.1059,
       "step": 604
     },
     {
       "epoch": 1.36,
       "learning_rate": 0.00028231213872832365,
-      "loss": 3.0861,
+      "loss": 4.0504,
       "step": 605
     },
     {
       "epoch": 1.36,
       "learning_rate": 0.0002821387283236994,
-      "loss": 2.8952,
+      "loss": 3.958,
       "step": 606
     },
     {
       "epoch": 1.36,
       "learning_rate": 0.00028196531791907514,
-      "loss": 2.9341,
+      "loss": 4.0532,
       "step": 607
     },
     {
       "epoch": 1.36,
       "learning_rate": 0.00028179190751445083,
-      "loss": 2.7669,
+      "loss": 3.9602,
       "step": 608
     },
     {
       "epoch": 1.37,
       "learning_rate": 0.0002816184971098266,
-      "loss": 2.687,
+      "loss": 3.979,
       "step": 609
     },
     {
       "epoch": 1.37,
       "learning_rate": 0.0002814450867052023,
-      "loss": 2.5708,
+      "loss": 3.8953,
       "step": 610
     },
     {
       "epoch": 1.37,
       "learning_rate": 0.000281271676300578,
-      "loss": 2.3895,
+      "loss": 3.8389,
       "step": 611
     },
     {
       "epoch": 1.37,
       "learning_rate": 0.0002810982658959537,
-      "loss": 2.38,
+      "loss": 3.8168,
       "step": 612
     },
     {
       "epoch": 1.37,
       "learning_rate": 0.00028092485549132947,
-      "loss": 2.4185,
+      "loss": 3.8634,
       "step": 613
     },
     {
       "epoch": 1.38,
       "learning_rate": 0.00028075144508670516,
-      "loss": 2.3059,
+      "loss": 3.8456,
       "step": 614
     },
     {
       "epoch": 1.38,
       "learning_rate": 0.0002805780346820809,
-      "loss": 2.2793,
+      "loss": 3.9215,
       "step": 615
     },
     {
       "epoch": 1.38,
       "learning_rate": 0.0002804046242774566,
-      "loss": 2.1804,
+      "loss": 3.7437,
       "step": 616
     },
     {
       "epoch": 1.38,
       "learning_rate": 0.00028023121387283235,
-      "loss": 2.1364,
+      "loss": 3.7972,
       "step": 617
     },
     {
       "epoch": 1.39,
       "learning_rate": 0.00028005780346820804,
-      "loss": 2.0321,
+      "loss": 3.6682,
       "step": 618
     },
     {
       "epoch": 1.39,
       "learning_rate": 0.0002798843930635838,
-      "loss": 2.03,
+      "loss": 3.7301,
       "step": 619
     },
     {
       "epoch": 1.39,
       "learning_rate": 0.00027971098265895954,
-      "loss": 1.8484,
+      "loss": 3.663,
       "step": 620
     },
     {
       "epoch": 1.39,
       "learning_rate": 0.00027953757225433523,
-      "loss": 2.004,
+      "loss": 3.7823,
       "step": 621
     },
     {
       "epoch": 1.39,
       "learning_rate": 0.0002793641618497109,
-      "loss": 1.9594,
+      "loss": 3.763,
       "step": 622
     },
     {
       "epoch": 1.4,
       "learning_rate": 0.00027919075144508667,
-      "loss": 1.8258,
+      "loss": 3.696,
       "step": 623
     },
     {
       "epoch": 1.4,
       "learning_rate": 0.0002790173410404624,
-      "loss": 1.8189,
+      "loss": 3.6394,
       "step": 624
     },
     {
       "epoch": 1.4,
       "learning_rate": 0.0002788439306358381,
-      "loss": 1.7799,
+      "loss": 3.6472,
       "step": 625
     },
     {
       "epoch": 1.4,
       "learning_rate": 0.00027867052023121386,
-      "loss": 1.6976,
+      "loss": 3.6325,
       "step": 626
     },
     {
       "epoch": 1.41,
       "learning_rate": 0.00027849710982658955,
-      "loss": 1.6733,
+      "loss": 3.5473,
       "step": 627
     },
     {
       "epoch": 1.41,
       "learning_rate": 0.0002783236994219653,
-      "loss": 1.703,
+      "loss": 3.6429,
       "step": 628
     },
     {
       "epoch": 1.41,
       "learning_rate": 0.000278150289017341,
-      "loss": 1.7001,
+      "loss": 3.6473,
       "step": 629
     },
     {
       "epoch": 1.41,
       "learning_rate": 0.00027797687861271674,
-      "loss": 1.608,
+      "loss": 3.5185,
       "step": 630
     },
     {
       "epoch": 1.41,
       "learning_rate": 0.0002778034682080925,
-      "loss": 1.7678,
+      "loss": 3.7047,
       "step": 631
     },
     {
       "epoch": 1.42,
       "learning_rate": 0.0002776300578034682,
-      "loss": 1.4661,
+      "loss": 3.467,
       "step": 632
     },
     {
       "epoch": 1.42,
       "learning_rate": 0.0002774566473988439,
-      "loss": 1.4454,
+      "loss": 3.4672,
       "step": 633
     },
     {
       "epoch": 1.42,
       "learning_rate": 0.0002772832369942196,
-      "loss": 1.2762,
+      "loss": 3.366,
       "step": 634
     },
     {
       "epoch": 1.42,
       "learning_rate": 0.0002771098265895954,
-      "loss": 1.283,
+      "loss": 3.3636,
       "step": 635
     },
     {
       "epoch": 1.43,
       "learning_rate": 0.00027693641618497107,
-      "loss": 1.2842,
+      "loss": 3.2793,
       "step": 636
     },
     {
       "epoch": 1.43,
       "learning_rate": 0.0002767630057803468,
-      "loss": 1.5555,
+      "loss": 3.5051,
       "step": 637
     },
     {
       "epoch": 1.43,
       "learning_rate": 0.0002765895953757225,
-      "loss": 1.2237,
+      "loss": 3.1726,
       "step": 638
     },
     {
       "epoch": 1.43,
       "learning_rate": 0.00027641618497109826,
-      "loss": 1.402,
+      "loss": 3.3622,
       "step": 639
     },
     {
       "epoch": 1.43,
       "learning_rate": 0.00027624277456647395,
-      "loss": 1.2479,
+      "loss": 3.1272,
       "step": 640
     },
     {
       "epoch": 1.44,
       "learning_rate": 0.0002760693641618497,
-      "loss": 1.0123,
+      "loss": 2.7958,
       "step": 641
     },
     {
       "epoch": 1.44,
       "learning_rate": 0.0002758959537572254,
-      "loss": 1.2752,
+      "loss": 3.0075,
       "step": 642
     },
     {
       "epoch": 1.44,
       "learning_rate": 0.00027572254335260114,
-      "loss": 1.1875,
+      "loss": 2.8361,
       "step": 643
     },
     {
       "epoch": 1.44,
       "learning_rate": 0.00027554913294797683,
-      "loss": 1.1334,
+      "loss": 2.4531,
       "step": 644
     },
     {
       "epoch": 1.45,
       "learning_rate": 0.0002753757225433526,
-      "loss": 1.1448,
+      "loss": 2.4185,
       "step": 645
     },
     {
       "epoch": 1.45,
       "learning_rate": 0.0002752023121387283,
-      "loss": 1.412,
+      "loss": 2.4573,
       "step": 646
     },
     {
       "epoch": 1.45,
       "learning_rate": 0.000275028901734104,
-      "loss": 6.0175,
+      "loss": 5.8762,
       "step": 647
     },
     {
       "epoch": 1.45,
       "learning_rate": 0.00027485549132947977,
-      "loss": 4.3887,
+      "loss": 5.5451,
       "step": 648
     },
     {
       "epoch": 1.46,
       "learning_rate": 0.00027468208092485546,
-      "loss": 3.5137,
+      "loss": 5.1378,
       "step": 649
     },
     {
       "epoch": 1.46,
       "learning_rate": 0.00027450867052023116,
-      "loss": 2.9133,
+      "loss": 4.6691,
       "step": 650
     },
     {
       "epoch": 1.46,
       "learning_rate": 0.0002743352601156069,
-      "loss": 2.6317,
+      "loss": 4.1925,
       "step": 651
     },
     {
       "epoch": 1.46,
       "learning_rate": 0.00027416184971098265,
-      "loss": 2.4746,
+      "loss": 3.9882,
       "step": 652
     },
     {
       "epoch": 1.46,
       "learning_rate": 0.00027398843930635835,
-      "loss": 2.2373,
+      "loss": 3.879,
       "step": 653
     },
     {
       "epoch": 1.47,
       "learning_rate": 0.0002738150289017341,
-      "loss": 2.0943,
+      "loss": 3.895,
       "step": 654
     },
     {
       "epoch": 1.47,
       "learning_rate": 0.00027364161849710984,
-      "loss": 2.0145,
+      "loss": 3.8319,
       "step": 655
     },
     {
       "epoch": 1.47,
       "learning_rate": 0.00027346820809248554,
-      "loss": 1.7654,
+      "loss": 3.7363,
       "step": 656
     },
     {
       "epoch": 1.47,
       "learning_rate": 0.00027329479768786123,
-      "loss": 1.7211,
+      "loss": 3.6597,
       "step": 657
     },
     {
       "epoch": 1.48,
       "learning_rate": 0.000273121387283237,
-      "loss": 1.5969,
+      "loss": 3.6444,
       "step": 658
     },
     {
       "epoch": 1.48,
       "learning_rate": 0.0002729479768786127,
-      "loss": 1.5845,
+      "loss": 3.5363,
       "step": 659
     },
     {
       "epoch": 1.48,
       "learning_rate": 0.0002727745664739884,
-      "loss": 1.4751,
+      "loss": 3.4726,
       "step": 660
     },
     {
       "epoch": 1.48,
       "learning_rate": 0.0002726011560693641,
-      "loss": 1.3363,
+      "loss": 3.3294,
       "step": 661
     },
     {
       "epoch": 1.48,
       "learning_rate": 0.00027242774566473986,
-      "loss": 1.4412,
+      "loss": 3.4765,
       "step": 662
     },
     {
       "epoch": 1.49,
       "learning_rate": 0.0002722543352601156,
-      "loss": 1.2796,
+      "loss": 3.3104,
       "step": 663
     },
     {
       "epoch": 1.49,
       "learning_rate": 0.0002720809248554913,
-      "loss": 1.2538,
+      "loss": 3.3259,
       "step": 664
     },
     {
       "epoch": 1.49,
       "learning_rate": 0.00027190751445086705,
-      "loss": 1.1078,
+      "loss": 3.1058,
       "step": 665
     },
     {
       "epoch": 1.49,
       "learning_rate": 0.00027173410404624274,
-      "loss": 1.2755,
+      "loss": 3.2575,
       "step": 666
     },
     {
       "epoch": 1.5,
       "learning_rate": 0.0002715606936416185,
-      "loss": 1.0332,
+      "loss": 3.0172,
       "step": 667
     },
     {
       "epoch": 1.5,
       "learning_rate": 0.0002713872832369942,
-      "loss": 1.0823,
+      "loss": 3.0809,
       "step": 668
     },
     {
       "epoch": 1.5,
       "learning_rate": 0.00027121387283236993,
-      "loss": 1.1291,
+      "loss": 3.1742,
       "step": 669
     },
     {
       "epoch": 1.5,
       "learning_rate": 0.0002710404624277456,
-      "loss": 1.0924,
+      "loss": 3.0458,
       "step": 670
     },
     {
       "epoch": 1.5,
       "learning_rate": 0.0002708670520231214,
-      "loss": 1.1536,
+      "loss": 2.9033,
       "step": 671
     },
     {
       "epoch": 1.51,
       "learning_rate": 0.00027069364161849707,
-      "loss": 1.1519,
+      "loss": 2.9879,
       "step": 672
     },
     {
       "epoch": 1.51,
       "learning_rate": 0.0002705202312138728,
-      "loss": 1.0834,
+      "loss": 2.8564,
       "step": 673
     },
     {
       "epoch": 1.51,
       "learning_rate": 0.0002703468208092485,
-      "loss": 1.0253,
+      "loss": 2.786,
       "step": 674
     },
     {
       "epoch": 1.51,
       "learning_rate": 0.00027017341040462426,
-      "loss": 1.0866,
+      "loss": 2.9336,
       "step": 675
     },
     {
       "epoch": 1.52,
       "learning_rate": 0.00027,
-      "loss": 0.9893,
+      "loss": 2.7786,
       "step": 676
     },
     {
       "epoch": 1.52,
       "learning_rate": 0.0002698265895953757,
-      "loss": 0.9927,
+      "loss": 2.664,
       "step": 677
     },
     {
       "epoch": 1.52,
       "learning_rate": 0.0002696531791907514,
-      "loss": 0.9538,
+      "loss": 2.65,
       "step": 678
     },
     {
       "epoch": 1.52,
       "learning_rate": 0.00026947976878612714,
-      "loss": 0.9511,
+      "loss": 2.6503,
       "step": 679
     },
     {
       "epoch": 1.52,
       "learning_rate": 0.0002693063583815029,
-      "loss": 0.988,
+      "loss": 2.7218,
       "step": 680
     },
     {
       "epoch": 1.53,
       "learning_rate": 0.0002691329479768786,
-      "loss": 0.903,
+      "loss": 2.6982,
       "step": 681
     },
     {
       "epoch": 1.53,
       "learning_rate": 0.00026895953757225433,
-      "loss": 0.9232,
+      "loss": 2.5694,
       "step": 682
     },
     {
       "epoch": 1.53,
       "learning_rate": 0.0002687861271676301,
-      "loss": 1.0025,
+      "loss": 2.5128,
       "step": 683
     },
     {
       "epoch": 1.53,
-      "learning_rate": 0.00026861271676300577,
-      "loss": 0.8771,
+      "learning_rate": 0.0002687861271676301,
+      "loss": 2.7005,
       "step": 684
     },
     {
       "epoch": 1.54,
-      "learning_rate": 0.00026843930635838146,
-      "loss": 0.8433,
+      "learning_rate": 0.00026861271676300577,
+      "loss": 2.325,
       "step": 685
     },
     {
       "epoch": 1.54,
-      "learning_rate": 0.0002682658959537572,
-      "loss": 0.9884,
+      "learning_rate": 0.00026843930635838146,
+      "loss": 2.5138,
       "step": 686
     },
     {
       "epoch": 1.54,
-      "learning_rate": 0.00026809248554913296,
-      "loss": 0.8276,
+      "learning_rate": 0.0002682658959537572,
+      "loss": 2.348,
       "step": 687
     },
     {
       "epoch": 1.54,
-      "learning_rate": 0.00026791907514450865,
-      "loss": 0.8752,
+      "learning_rate": 0.00026809248554913296,
+      "loss": 2.2269,
       "step": 688
     },
     {
       "epoch": 1.54,
-      "learning_rate": 0.00026774566473988435,
-      "loss": 0.9028,
+      "learning_rate": 0.00026791907514450865,
+      "loss": 2.2502,
       "step": 689
     },
     {
       "epoch": 1.55,
-      "learning_rate": 0.0002675722543352601,
-      "loss": 0.828,
+      "learning_rate": 0.00026774566473988435,
+      "loss": 1.9571,
       "step": 690
     },
     {
       "epoch": 1.55,
-      "learning_rate": 0.00026739884393063584,
-      "loss": 0.9169,
+      "learning_rate": 0.0002675722543352601,
+      "loss": 2.003,
       "step": 691
     },
     {
       "epoch": 1.55,
-      "learning_rate": 0.00026722543352601153,
-      "loss": 0.9081,
+      "learning_rate": 0.00026739884393063584,
+      "loss": 1.8658,
       "step": 692
     },
     {
       "epoch": 1.55,
-      "learning_rate": 0.0002670520231213873,
-      "loss": 0.8513,
+      "learning_rate": 0.00026722543352601153,
+      "loss": 1.5905,
       "step": 693
     },
     {
       "epoch": 1.56,
-      "learning_rate": 0.000266878612716763,
-      "loss": 0.8775,
+      "learning_rate": 0.0002670520231213873,
+      "loss": 1.7254,
       "step": 694
     },
     {
       "epoch": 1.56,
-      "learning_rate": 0.0002667052023121387,
-      "loss": 0.7978,
+      "learning_rate": 0.000266878612716763,
+      "loss": 1.4685,
       "step": 695
     },
     {
       "epoch": 1.56,
-      "learning_rate": 0.0002665317919075144,
-      "loss": 0.7886,
+      "learning_rate": 0.0002667052023121387,
+      "loss": 1.1921,
       "step": 696
     },
     {
       "epoch": 1.56,
-      "learning_rate": 0.00026635838150289016,
-      "loss": 4.5421,
+      "learning_rate": 0.0002665317919075144,
+      "loss": 8.7366,
       "step": 697
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.00026618497109826586,
-      "loss": 3.028,
+      "learning_rate": 0.00026635838150289016,
+      "loss": 6.141,
       "step": 698
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.00026601156069364155,
-      "loss": 1.9085,
+      "learning_rate": 0.00026618497109826586,
+      "loss": 4.6943,
       "step": 699
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.0002658381502890173,
-      "loss": 1.8521,
+      "learning_rate": 0.00026601156069364155,
+      "loss": 3.7024,
       "step": 700
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.00026566473988439305,
-      "loss": 2.0182,
+      "learning_rate": 0.0002658381502890173,
+      "loss": 3.4388,
       "step": 701
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.00026549132947976874,
-      "loss": 1.5691,
+      "learning_rate": 0.00026566473988439305,
+      "loss": 3.0568,
       "step": 702
     },
     {
       "epoch": 1.58,
-      "learning_rate": 0.0002653179190751445,
-      "loss": 1.5199,
+      "learning_rate": 0.00026549132947976874,
+      "loss": 3.0959,
       "step": 703
     },
     {
       "epoch": 1.58,
-      "learning_rate": 0.00026514450867052024,
-      "loss": 1.4381,
+      "learning_rate": 0.0002653179190751445,
+      "loss": 2.9189,
       "step": 704
     },
     {
       "epoch": 1.58,
-      "learning_rate": 0.00026497109826589593,
-      "loss": 1.2407,
+      "learning_rate": 0.00026514450867052024,
+      "loss": 2.7027,
       "step": 705
     },
     {
       "epoch": 1.58,
-      "learning_rate": 0.0002647976878612716,
-      "loss": 1.288,
+      "learning_rate": 0.00026497109826589593,
+      "loss": 2.7317,
       "step": 706
     },
     {
       "epoch": 1.59,
-      "learning_rate": 0.00026462427745664737,
-      "loss": 1.1379,
+      "learning_rate": 0.0002647976878612716,
+      "loss": 2.5358,
       "step": 707
     },
     {
       "epoch": 1.59,
-      "learning_rate": 0.0002644508670520231,
-      "loss": 1.1389,
+      "learning_rate": 0.00026462427745664737,
+      "loss": 2.3571,
       "step": 708
     },
     {
       "epoch": 1.59,
-      "learning_rate": 0.0002642774566473988,
-      "loss": 1.1698,
+      "learning_rate": 0.0002644508670520231,
+      "loss": 2.3926,
       "step": 709
     },
     {
       "epoch": 1.59,
-      "learning_rate": 0.00026410404624277456,
-      "loss": 0.9825,
+      "learning_rate": 0.0002642774566473988,
+      "loss": 2.1589,
       "step": 710
     },
     {
       "epoch": 1.59,
-      "learning_rate": 0.00026393063583815025,
-      "loss": 0.9905,
+      "learning_rate": 0.00026410404624277456,
+      "loss": 2.0983,
       "step": 711
     },
     {
       "epoch": 1.6,
-      "learning_rate": 0.000263757225433526,
-      "loss": 0.8542,
+      "learning_rate": 0.00026393063583815025,
+      "loss": 1.9613,
       "step": 712
     },
     {
       "epoch": 1.6,
-      "learning_rate": 0.0002635838150289017,
-      "loss": 0.9222,
+      "learning_rate": 0.000263757225433526,
+      "loss": 2.0088,
       "step": 713
     },
     {
       "epoch": 1.6,
-      "learning_rate": 0.00026341040462427744,
-      "loss": 0.8739,
+      "learning_rate": 0.0002635838150289017,
+      "loss": 2.0606,
       "step": 714
     },
     {
       "epoch": 1.6,
-      "learning_rate": 0.00026323699421965314,
-      "loss": 0.8896,
+      "learning_rate": 0.00026341040462427744,
+      "loss": 1.9965,
       "step": 715
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.0002630635838150289,
-      "loss": 0.827,
+      "learning_rate": 0.00026323699421965314,
+      "loss": 1.8923,
       "step": 716
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.0002628901734104046,
-      "loss": 0.8345,
+      "learning_rate": 0.0002630635838150289,
+      "loss": 1.83,
       "step": 717
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.0002627167630057803,
-      "loss": 0.8356,
+      "learning_rate": 0.0002628901734104046,
+      "loss": 1.8531,
       "step": 718
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.000262543352601156,
-      "loss": 0.787,
+      "learning_rate": 0.0002627167630057803,
+      "loss": 1.6699,
       "step": 719
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.00026236994219653177,
-      "loss": 0.7632,
+      "learning_rate": 0.000262543352601156,
+      "loss": 1.6279,
       "step": 720
     },
     {
       "epoch": 1.62,
-      "learning_rate": 0.0002621965317919075,
-      "loss": 0.7849,
+      "learning_rate": 0.00026236994219653177,
+      "loss": 1.6581,
       "step": 721
     },
     {
       "epoch": 1.62,
-      "learning_rate": 0.0002620231213872832,
-      "loss": 0.9055,
+      "learning_rate": 0.0002621965317919075,
+      "loss": 1.6376,
       "step": 722
     },
     {
       "epoch": 1.62,
-      "learning_rate": 0.0002618497109826589,
-      "loss": 0.88,
+      "learning_rate": 0.0002620231213872832,
+      "loss": 1.6408,
       "step": 723
     },
     {
       "epoch": 1.62,
-      "learning_rate": 0.00026167630057803465,
-      "loss": 0.6907,
+      "learning_rate": 0.0002618497109826589,
+      "loss": 1.4039,
       "step": 724
     },
     {
       "epoch": 1.63,
-      "learning_rate": 0.0002615028901734104,
-      "loss": 0.7317,
+      "learning_rate": 0.00026167630057803465,
+      "loss": 1.4447,
       "step": 725
     },
     {
       "epoch": 1.63,
-      "learning_rate": 0.0002613294797687861,
-      "loss": 0.7804,
+      "learning_rate": 0.0002615028901734104,
+      "loss": 1.655,
       "step": 726
     },
     {
       "epoch": 1.63,
-      "learning_rate": 0.00026115606936416184,
-      "loss": 0.6707,
+      "learning_rate": 0.0002613294797687861,
+      "loss": 1.3933,
       "step": 727
     },
     {
       "epoch": 1.63,
-      "learning_rate": 0.0002609826589595376,
-      "loss": 0.6019,
+      "learning_rate": 0.00026115606936416184,
+      "loss": 1.3983,
       "step": 728
     },
     {
       "epoch": 1.63,
-      "learning_rate": 0.0002608092485549133,
-      "loss": 0.6316,
+      "learning_rate": 0.0002609826589595376,
+      "loss": 1.2188,
       "step": 729
     },
     {
       "epoch": 1.64,
-      "learning_rate": 0.000260635838150289,
-      "loss": 0.7106,
+      "learning_rate": 0.0002608092485549133,
+      "loss": 1.3292,
       "step": 730
     },
     {
       "epoch": 1.64,
-      "learning_rate": 0.0002604624277456647,
-      "loss": 0.6806,
+      "learning_rate": 0.000260635838150289,
+      "loss": 1.3526,
       "step": 731
     },
     {
       "epoch": 1.64,
-      "learning_rate": 0.00026028901734104047,
-      "loss": 0.7568,
+      "learning_rate": 0.0002604624277456647,
+      "loss": 1.3375,
       "step": 732
     },
     {
       "epoch": 1.64,
-      "learning_rate": 0.00026011560693641616,
-      "loss": 0.6379,
+      "learning_rate": 0.00026028901734104047,
+      "loss": 1.1941,
       "step": 733
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.00025994219653179186,
-      "loss": 0.7025,
+      "learning_rate": 0.00026011560693641616,
+      "loss": 1.2204,
       "step": 734
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.0002597687861271676,
-      "loss": 0.7221,
+      "learning_rate": 0.00025994219653179186,
+      "loss": 1.2544,
       "step": 735
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.00025959537572254335,
-      "loss": 0.8241,
+      "learning_rate": 0.0002597687861271676,
+      "loss": 1.3232,
       "step": 736
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.00025942196531791905,
-      "loss": 0.7472,
+      "learning_rate": 0.00025959537572254335,
+      "loss": 1.1686,
       "step": 737
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.0002592485549132948,
-      "loss": 0.7684,
+      "learning_rate": 0.00025942196531791905,
+      "loss": 1.1723,
       "step": 738
     },
     {
       "epoch": 1.66,
-      "learning_rate": 0.0002590751445086705,
-      "loss": 0.7034,
+      "learning_rate": 0.0002592485549132948,
+      "loss": 1.1758,
       "step": 739
     },
     {
       "epoch": 1.66,
-      "learning_rate": 0.00025890173410404624,
-      "loss": 0.6973,
+      "learning_rate": 0.0002590751445086705,
+      "loss": 1.1159,
       "step": 740
     },
     {
       "epoch": 1.66,
-      "learning_rate": 0.00025872832369942193,
-      "loss": 0.8219,
+      "learning_rate": 0.00025890173410404624,
+      "loss": 1.3023,
       "step": 741
     },
     {
       "epoch": 1.66,
-      "learning_rate": 0.0002585549132947977,
-      "loss": 0.6853,
+      "learning_rate": 0.00025872832369942193,
+      "loss": 1.1216,
       "step": 742
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.00025838150289017337,
-      "loss": 0.7122,
+      "learning_rate": 0.0002585549132947977,
+      "loss": 1.0869,
       "step": 743
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.0002582080924855491,
-      "loss": 0.9218,
+      "learning_rate": 0.00025838150289017337,
+      "loss": 1.3723,
       "step": 744
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.0002580346820809248,
-      "loss": 0.6941,
+      "learning_rate": 0.0002582080924855491,
+      "loss": 0.9013,
       "step": 745
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.00025786127167630056,
-      "loss": 1.0277,
+      "learning_rate": 0.0002580346820809248,
+      "loss": 1.2461,
       "step": 746
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.00025768786127167625,
-      "loss": 3.3368,
+      "learning_rate": 0.00025786127167630056,
+      "loss": 6.9996,
       "step": 747
     },
     {
       "epoch": 1.68,
-      "learning_rate": 0.000257514450867052,
-      "loss": 2.5557,
+      "learning_rate": 0.00025768786127167625,
+      "loss": 6.0548,
       "step": 748
     },
     {
       "epoch": 1.68,
-      "learning_rate": 0.00025734104046242775,
-      "loss": 1.5549,
+      "learning_rate": 0.000257514450867052,
+      "loss": 4.8795,
       "step": 749
     },
     {
       "epoch": 1.68,
-      "learning_rate": 0.00025716763005780344,
-      "loss": 1.2491,
+      "learning_rate": 0.00025734104046242775,
+      "loss": 3.2843,
       "step": 750
     },
     {
       "epoch": 1.68,
-      "learning_rate": 0.00025699421965317914,
-      "loss": 1.2336,
+      "learning_rate": 0.00025716763005780344,
+      "loss": 2.8835,
       "step": 751
     },
     {
       "epoch": 1.69,
-      "learning_rate": 0.0002568208092485549,
-      "loss": 1.1964,
+      "learning_rate": 0.00025699421965317914,
+      "loss": 2.3855,
       "step": 752
     },
     {
       "epoch": 1.69,
-      "learning_rate": 0.00025664739884393063,
-      "loss": 1.1691,
+      "learning_rate": 0.0002568208092485549,
+      "loss": 2.2957,
       "step": 753
     },
     {
       "epoch": 1.69,
-      "learning_rate": 0.0002564739884393063,
-      "loss": 0.959,
+      "learning_rate": 0.00025664739884393063,
+      "loss": 2.001,
       "step": 754
     },
     {
       "epoch": 1.69,
-      "learning_rate": 0.00025630057803468207,
-      "loss": 0.9412,
+      "learning_rate": 0.0002564739884393063,
+      "loss": 1.824,
       "step": 755
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.0002561271676300578,
-      "loss": 0.9436,
+      "learning_rate": 0.00025630057803468207,
+      "loss": 1.9435,
       "step": 756
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.0002559537572254335,
-      "loss": 0.7702,
+      "learning_rate": 0.0002561271676300578,
+      "loss": 1.5901,
       "step": 757
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.0002557803468208092,
-      "loss": 0.7077,
+      "learning_rate": 0.0002559537572254335,
+      "loss": 1.447,
       "step": 758
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.00025560693641618496,
-      "loss": 0.6719,
+      "learning_rate": 0.0002557803468208092,
+      "loss": 1.4363,
       "step": 759
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.0002554335260115607,
-      "loss": 0.6603,
+      "learning_rate": 0.00025560693641618496,
+      "loss": 1.3356,
       "step": 760
     },
     {
       "epoch": 1.71,
-      "learning_rate": 0.0002552601156069364,
-      "loss": 0.7441,
+      "learning_rate": 0.0002554335260115607,
+      "loss": 1.3838,
       "step": 761
     },
     {
       "epoch": 1.71,
-      "learning_rate": 0.0002550867052023121,
-      "loss": 0.6565,
+      "learning_rate": 0.0002552601156069364,
+      "loss": 1.1423,
       "step": 762
     },
     {
       "epoch": 1.71,
-      "learning_rate": 0.00025491329479768784,
-      "loss": 0.6591,
+      "learning_rate": 0.0002550867052023121,
+      "loss": 1.2079,
       "step": 763
     },
     {
       "epoch": 1.71,
-      "learning_rate": 0.0002547398843930636,
-      "loss": 0.6369,
+      "learning_rate": 0.00025491329479768784,
+      "loss": 1.1538,
       "step": 764
     },
     {
       "epoch": 1.72,
-      "learning_rate": 0.0002545664739884393,
-      "loss": 0.6335,
+      "learning_rate": 0.0002547398843930636,
+      "loss": 1.0943,
       "step": 765
     },
     {
       "epoch": 1.72,
-      "learning_rate": 0.00025439306358381503,
-      "loss": 0.6988,
+      "learning_rate": 0.0002545664739884393,
+      "loss": 1.1982,
       "step": 766
     },
     {
       "epoch": 1.72,
-      "learning_rate": 0.0002542196531791907,
-      "loss": 0.7381,
+      "learning_rate": 0.00025439306358381503,
+      "loss": 1.1676,
       "step": 767
     },
     {
       "epoch": 1.72,
-      "learning_rate": 0.00025404624277456647,
-      "loss": 0.7094,
+      "learning_rate": 0.0002542196531791907,
+      "loss": 1.0995,
       "step": 768
     },
     {
       "epoch": 1.72,
-      "learning_rate": 0.00025387283236994216,
-      "loss": 0.6374,
+      "learning_rate": 0.00025404624277456647,
+      "loss": 0.9632,
       "step": 769
     },
     {
       "epoch": 1.73,
-      "learning_rate": 0.0002536994219653179,
-      "loss": 0.6439,
+      "learning_rate": 0.00025387283236994216,
+      "loss": 1.0703,
       "step": 770
     },
     {
       "epoch": 1.73,
-      "learning_rate": 0.0002535260115606936,
-      "loss": 0.5836,
+      "learning_rate": 0.0002536994219653179,
+      "loss": 0.9207,
       "step": 771
     },
     {
       "epoch": 1.73,
-      "learning_rate": 0.00025335260115606935,
-      "loss": 0.6539,
+      "learning_rate": 0.0002535260115606936,
+      "loss": 0.9262,
       "step": 772
     },
     {
       "epoch": 1.73,
-      "learning_rate": 0.00025317919075144504,
-      "loss": 0.5807,
+      "learning_rate": 0.00025335260115606935,
+      "loss": 0.9645,
       "step": 773
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.0002530057803468208,
-      "loss": 0.6986,
+      "learning_rate": 0.00025317919075144504,
+      "loss": 1.0928,
       "step": 774
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.0002528323699421965,
-      "loss": 0.6707,
+      "learning_rate": 0.0002530057803468208,
+      "loss": 0.9605,
       "step": 775
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.00025265895953757223,
-      "loss": 0.6523,
+      "learning_rate": 0.0002528323699421965,
+      "loss": 1.0233,
       "step": 776
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.000252485549132948,
-      "loss": 0.6065,
+      "learning_rate": 0.00025265895953757223,
+      "loss": 0.9323,
       "step": 777
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.0002523121387283237,
-      "loss": 0.591,
+      "learning_rate": 0.000252485549132948,
+      "loss": 0.8583,
       "step": 778
     },
     {
       "epoch": 1.75,
-      "learning_rate": 0.00025213872832369937,
-      "loss": 0.6904,
+      "learning_rate": 0.0002523121387283237,
+      "loss": 1.039,
       "step": 779
     },
     {
       "epoch": 1.75,
-      "learning_rate": 0.0002519653179190751,
-      "loss": 0.5763,
+      "learning_rate": 0.00025213872832369937,
+      "loss": 0.9294,
       "step": 780
     },
     {
       "epoch": 1.75,
-      "learning_rate": 0.00025179190751445086,
-      "loss": 0.6093,
+      "learning_rate": 0.0002519653179190751,
+      "loss": 0.9475,
       "step": 781
     },
     {
       "epoch": 1.75,
-      "learning_rate": 0.00025161849710982656,
-      "loss": 0.6226,
+      "learning_rate": 0.00025179190751445086,
+      "loss": 0.9068,
       "step": 782
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.0002514450867052023,
-      "loss": 0.6242,
+      "learning_rate": 0.00025161849710982656,
+      "loss": 0.8889,
       "step": 783
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.00025127167630057805,
-      "loss": 0.6007,
+      "learning_rate": 0.0002514450867052023,
+      "loss": 0.8678,
       "step": 784
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.00025109826589595375,
-      "loss": 0.5843,
+      "learning_rate": 0.00025127167630057805,
+      "loss": 0.8449,
       "step": 785
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.00025092485549132944,
-      "loss": 0.5994,
+      "learning_rate": 0.00025109826589595375,
+      "loss": 0.885,
       "step": 786
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.0002507514450867052,
-      "loss": 0.5859,
+      "learning_rate": 0.00025092485549132944,
+      "loss": 0.7919,
       "step": 787
     },
     {
       "epoch": 1.77,
-      "learning_rate": 0.00025057803468208094,
-      "loss": 0.6308,
+      "learning_rate": 0.0002507514450867052,
+      "loss": 0.9106,
       "step": 788
     },
     {
       "epoch": 1.77,
-      "learning_rate": 0.00025040462427745663,
-      "loss": 0.5669,
+      "learning_rate": 0.00025057803468208094,
+      "loss": 0.8597,
       "step": 789
     },
     {
       "epoch": 1.77,
-      "learning_rate": 0.0002502312138728323,
-      "loss": 0.6313,
+      "learning_rate": 0.00025040462427745663,
+      "loss": 0.8783,
       "step": 790
     },
     {
       "epoch": 1.77,
-      "learning_rate": 0.00025005780346820807,
-      "loss": 0.6797,
+      "learning_rate": 0.0002502312138728323,
+      "loss": 0.897,
       "step": 791
     },
     {
       "epoch": 1.78,
-      "learning_rate": 0.00024988439306358376,
-      "loss": 0.6099,
+      "learning_rate": 0.00025005780346820807,
+      "loss": 0.7942,
       "step": 792
     },
     {
       "epoch": 1.78,
-      "learning_rate": 0.0002497109826589595,
-      "loss": 0.6774,
+      "learning_rate": 0.00024988439306358376,
+      "loss": 0.8107,
       "step": 793
     },
     {
       "epoch": 1.78,
-      "learning_rate": 0.00024953757225433526,
-      "loss": 0.6585,
+      "learning_rate": 0.0002497109826589595,
+      "loss": 0.8347,
       "step": 794
     },
     {
       "epoch": 1.78,
-      "learning_rate": 0.00024936416184971095,
-      "loss": 0.8394,
+      "learning_rate": 0.00024953757225433526,
+      "loss": 1.0615,
       "step": 795
     },
     {
       "epoch": 1.78,
-      "learning_rate": 0.00024919075144508665,
-      "loss": 0.5475,
+      "learning_rate": 0.00024936416184971095,
+      "loss": 0.7615,
       "step": 796
     },
     {
       "epoch": 1.79,
-      "learning_rate": 0.0002490173410404624,
-      "loss": 2.8929,
+      "learning_rate": 0.00024919075144508665,
+      "loss": 6.4464,
       "step": 797
     },
     {
       "epoch": 1.79,
-      "learning_rate": 0.00024884393063583814,
-      "loss": 2.074,
+      "learning_rate": 0.0002490173410404624,
+      "loss": 5.0751,
       "step": 798
     },
     {
       "epoch": 1.79,
-      "learning_rate": 0.00024867052023121384,
-      "loss": 1.3343,
+      "learning_rate": 0.00024884393063583814,
+      "loss": 3.7757,
       "step": 799
     },
     {
       "epoch": 1.79,
-      "learning_rate": 0.0002484971098265896,
-      "loss": 1.0365,
+      "learning_rate": 0.00024867052023121384,
+      "loss": 2.5764,
       "step": 800
     },
     {
       "epoch": 1.8,
-      "learning_rate": 0.00024832369942196533,
-      "loss": 1.07,
+      "learning_rate": 0.0002484971098265896,
+      "loss": 2.1743,
       "step": 801
     },
     {
       "epoch": 1.8,
-      "learning_rate": 0.000248150289017341,
-      "loss": 1.0102,
+      "learning_rate": 0.00024832369942196533,
+      "loss": 1.9198,
       "step": 802
     },
     {
       "epoch": 1.8,
-      "learning_rate": 0.0002479768786127167,
-      "loss": 0.8969,
+      "learning_rate": 0.000248150289017341,
+      "loss": 1.7835,
       "step": 803
     },
     {
       "epoch": 1.8,
-      "learning_rate": 0.00024780346820809247,
-      "loss": 0.9204,
+      "learning_rate": 0.0002479768786127167,
+      "loss": 1.6071,
       "step": 804
     },
     {
       "epoch": 1.8,
-      "learning_rate": 0.0002476300578034682,
-      "loss": 0.8675,
+      "learning_rate": 0.00024780346820809247,
+      "loss": 1.511,
       "step": 805
     },
     {
       "epoch": 1.81,
-      "learning_rate": 0.0002474566473988439,
-      "loss": 0.8545,
+      "learning_rate": 0.0002476300578034682,
+      "loss": 1.464,
       "step": 806
     },
     {
       "epoch": 1.81,
-      "learning_rate": 0.0002472832369942196,
-      "loss": 0.635,
+      "learning_rate": 0.0002474566473988439,
+      "loss": 1.1126,
       "step": 807
     },
     {
       "epoch": 1.81,
-      "learning_rate": 0.00024710982658959535,
-      "loss": 0.5683,
+      "learning_rate": 0.0002472832369942196,
+      "loss": 1.0599,
       "step": 808
     },
     {
       "epoch": 1.81,
-      "learning_rate": 0.0002469364161849711,
-      "loss": 0.6124,
+      "learning_rate": 0.00024710982658959535,
+      "loss": 1.0413,
       "step": 809
     },
     {
       "epoch": 1.82,
-      "learning_rate": 0.0002467630057803468,
-      "loss": 0.6162,
+      "learning_rate": 0.0002469364161849711,
+      "loss": 1.0306,
       "step": 810
     },
     {
       "epoch": 1.82,
-      "learning_rate": 0.00024658959537572254,
-      "loss": 0.5883,
+      "learning_rate": 0.0002467630057803468,
+      "loss": 0.943,
       "step": 811
     },
     {
       "epoch": 1.82,
-      "learning_rate": 0.00024641618497109823,
-      "loss": 0.6141,
+      "learning_rate": 0.00024658959537572254,
+      "loss": 0.9814,
       "step": 812
     },
     {
       "epoch": 1.82,
-      "learning_rate": 0.000246242774566474,
-      "loss": 0.538,
+      "learning_rate": 0.00024641618497109823,
+      "loss": 0.8165,
       "step": 813
     },
     {
       "epoch": 1.83,
-      "learning_rate": 0.0002460693641618497,
-      "loss": 0.5417,
+      "learning_rate": 0.000246242774566474,
+      "loss": 0.8491,
       "step": 814
     },
     {
       "epoch": 1.83,
-      "learning_rate": 0.0002458959537572254,
-      "loss": 0.6722,
+      "learning_rate": 0.0002460693641618497,
+      "loss": 0.9818,
       "step": 815
     },
     {
       "epoch": 1.83,
-      "learning_rate": 0.0002457225433526011,
-      "loss": 0.5552,
+      "learning_rate": 0.0002458959537572254,
+      "loss": 0.8564,
       "step": 816
     },
     {
       "epoch": 1.83,
-      "learning_rate": 0.00024554913294797686,
-      "loss": 0.6006,
+      "learning_rate": 0.0002457225433526011,
+      "loss": 0.8471,
       "step": 817
     },
     {
       "epoch": 1.83,
-      "learning_rate": 0.00024537572254335256,
-      "loss": 0.5318,
+      "learning_rate": 0.00024554913294797686,
+      "loss": 0.7855,
       "step": 818
     },
     {
       "epoch": 1.84,
-      "learning_rate": 0.0002452023121387283,
-      "loss": 0.5067,
+      "learning_rate": 0.00024537572254335256,
+      "loss": 0.7987,
       "step": 819
     },
     {
       "epoch": 1.84,
-      "learning_rate": 0.000245028901734104,
-      "loss": 0.4765,
+      "learning_rate": 0.0002452023121387283,
+      "loss": 0.7538,
       "step": 820
     },
     {
       "epoch": 1.84,
-      "learning_rate": 0.00024485549132947975,
-      "loss": 0.5454,
+      "learning_rate": 0.000245028901734104,
+      "loss": 0.8812,
       "step": 821
     },
     {
       "epoch": 1.84,
-      "learning_rate": 0.0002446820809248555,
-      "loss": 0.6913,
+      "learning_rate": 0.00024485549132947975,
+      "loss": 0.8958,
       "step": 822
     },
     {
       "epoch": 1.85,
-      "learning_rate": 0.0002445086705202312,
-      "loss": 0.481,
+      "learning_rate": 0.0002446820809248555,
+      "loss": 0.7152,
       "step": 823
     },
     {
       "epoch": 1.85,
-      "learning_rate": 0.0002443352601156069,
-      "loss": 0.5909,
+      "learning_rate": 0.0002445086705202312,
+      "loss": 0.7971,
       "step": 824
     },
     {
       "epoch": 1.85,
-      "learning_rate": 0.00024416184971098263,
-      "loss": 0.5372,
+      "learning_rate": 0.0002443352601156069,
+      "loss": 0.7203,
       "step": 825
     },
     {
       "epoch": 1.85,
-      "learning_rate": 0.00024398843930635838,
-      "loss": 0.4706,
+      "learning_rate": 0.00024416184971098263,
+      "loss": 0.7015,
       "step": 826
     },
     {
       "epoch": 1.85,
-      "learning_rate": 0.00024381502890173407,
-      "loss": 0.6338,
+      "learning_rate": 0.00024398843930635838,
+      "loss": 0.9212,
       "step": 827
     },
     {
       "epoch": 1.86,
-      "learning_rate": 0.0002436416184971098,
-      "loss": 0.4947,
+      "learning_rate": 0.00024381502890173407,
+      "loss": 0.6807,
       "step": 828
     },
     {
       "epoch": 1.86,
-      "learning_rate": 0.00024346820809248554,
-      "loss": 0.4791,
+      "learning_rate": 0.0002436416184971098,
+      "loss": 0.6554,
       "step": 829
     },
     {
       "epoch": 1.86,
-      "learning_rate": 0.00024329479768786126,
-      "loss": 0.4903,
+      "learning_rate": 0.00024346820809248554,
+      "loss": 0.7264,
       "step": 830
     },
     {
       "epoch": 1.86,
-      "learning_rate": 0.00024312138728323698,
-      "loss": 0.524,
+      "learning_rate": 0.00024329479768786126,
+      "loss": 0.7385,
       "step": 831
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.00024294797687861267,
-      "loss": 0.5472,
+      "learning_rate": 0.00024312138728323698,
+      "loss": 0.757,
       "step": 832
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.00024277456647398842,
-      "loss": 0.5067,
+      "learning_rate": 0.00024294797687861267,
+      "loss": 0.687,
       "step": 833
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.00024260115606936414,
-      "loss": 0.5282,
+      "learning_rate": 0.00024277456647398842,
+      "loss": 0.7073,
       "step": 834
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.00024242774566473986,
-      "loss": 0.5646,
+      "learning_rate": 0.00024260115606936414,
+      "loss": 0.674,
       "step": 835
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.00024225433526011558,
-      "loss": 0.5299,
+      "learning_rate": 0.00024242774566473986,
+      "loss": 0.6726,
       "step": 836
     },
     {
       "epoch": 1.88,
-      "learning_rate": 0.00024208092485549133,
-      "loss": 0.6317,
+      "learning_rate": 0.00024225433526011558,
+      "loss": 0.7894,
       "step": 837
     },
     {
       "epoch": 1.88,
-      "learning_rate": 0.00024190751445086702,
-      "loss": 0.5351,
+      "learning_rate": 0.00024208092485549133,
+      "loss": 0.6616,
       "step": 838
     },
     {
       "epoch": 1.88,
-      "learning_rate": 0.00024173410404624275,
-      "loss": 0.5208,
+      "learning_rate": 0.00024190751445086702,
+      "loss": 0.7662,
       "step": 839
     },
     {
       "epoch": 1.88,
-      "learning_rate": 0.00024156069364161847,
-      "loss": 0.5411,
+      "learning_rate": 0.00024173410404624275,
+      "loss": 0.644,
       "step": 840
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.0002413872832369942,
-      "loss": 0.484,
+      "learning_rate": 0.00024156069364161847,
+      "loss": 0.6078,
       "step": 841
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.00024121387283236993,
-      "loss": 0.5451,
+      "learning_rate": 0.0002413872832369942,
+      "loss": 0.6854,
       "step": 842
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.00024104046242774563,
-      "loss": 0.5628,
+      "learning_rate": 0.00024121387283236993,
+      "loss": 0.7161,
       "step": 843
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.00024086705202312135,
-      "loss": 0.6342,
+      "learning_rate": 0.00024104046242774563,
+      "loss": 0.7296,
       "step": 844
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.0002406936416184971,
-      "loss": 0.8021,
+      "learning_rate": 0.00024086705202312135,
+      "loss": 0.9602,
       "step": 845
     },
     {
       "epoch": 1.9,
-      "learning_rate": 0.00024052023121387282,
-      "loss": 0.8795,
+      "learning_rate": 0.0002406936416184971,
+      "loss": 0.9445,
       "step": 846
     },
     {
       "epoch": 1.9,
-      "learning_rate": 0.00024034682080924854,
-      "loss": 1.5202,
+      "learning_rate": 0.00024052023121387282,
+      "loss": 4.7021,
       "step": 847
     },
     {
       "epoch": 1.9,
-      "learning_rate": 0.00024017341040462423,
-      "loss": 1.1851,
+      "learning_rate": 0.00024034682080924854,
+      "loss": 3.3921,
       "step": 848
     },
     {
       "epoch": 1.9,
-      "learning_rate": 0.00023999999999999998,
-      "loss": 0.8733,
+      "learning_rate": 0.00024017341040462423,
+      "loss": 1.9051,
       "step": 849
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.0002398265895953757,
-      "loss": 0.7796,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 1.4066,
       "step": 850
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.00023965317919075142,
-      "loss": 0.8909,
+      "learning_rate": 0.0002398265895953757,
+      "loss": 1.3678,
       "step": 851
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.00023947976878612714,
-      "loss": 0.7804,
+      "learning_rate": 0.00023965317919075142,
+      "loss": 1.3079,
       "step": 852
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.0002393063583815029,
-      "loss": 0.6437,
+      "learning_rate": 0.00023947976878612714,
+      "loss": 1.0537,
       "step": 853
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.0002391329479768786,
-      "loss": 0.6786,
+      "learning_rate": 0.0002393063583815029,
+      "loss": 1.0467,
       "step": 854
     },
     {
       "epoch": 1.92,
-      "learning_rate": 0.0002389595375722543,
-      "loss": 0.6218,
+      "learning_rate": 0.0002391329479768786,
+      "loss": 0.9966,
       "step": 855
     },
     {
       "epoch": 1.92,
-      "learning_rate": 0.00023878612716763002,
-      "loss": 0.6243,
+      "learning_rate": 0.0002389595375722543,
+      "loss": 0.8728,
       "step": 856
     },
     {
       "epoch": 1.92,
-      "learning_rate": 0.00023861271676300577,
-      "loss": 0.6316,
+      "learning_rate": 0.00023878612716763002,
+      "loss": 0.8287,
       "step": 857
     },
     {
       "epoch": 1.92,
-      "learning_rate": 0.0002384393063583815,
-      "loss": 0.646,
+      "learning_rate": 0.00023861271676300577,
+      "loss": 0.891,
       "step": 858
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.0002382658959537572,
-      "loss": 0.4941,
+      "learning_rate": 0.0002384393063583815,
+      "loss": 0.6969,
       "step": 859
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.0002380924855491329,
-      "loss": 0.6629,
+      "learning_rate": 0.0002382658959537572,
+      "loss": 0.7537,
       "step": 860
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.00023791907514450865,
-      "loss": 0.5397,
+      "learning_rate": 0.0002380924855491329,
+      "loss": 0.7415,
       "step": 861
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.00023774566473988437,
-      "loss": 0.4732,
+      "learning_rate": 0.00023791907514450865,
+      "loss": 0.6404,
       "step": 862
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.0002375722543352601,
-      "loss": 0.4691,
+      "learning_rate": 0.00023774566473988437,
+      "loss": 0.6738,
       "step": 863
     },
     {
       "epoch": 1.94,
-      "learning_rate": 0.00023739884393063582,
-      "loss": 0.5437,
+      "learning_rate": 0.0002375722543352601,
+      "loss": 0.7438,
       "step": 864
     },
     {
       "epoch": 1.94,
-      "learning_rate": 0.00023722543352601156,
-      "loss": 0.4784,
+      "learning_rate": 0.00023739884393063582,
+      "loss": 0.5957,
       "step": 865
     },
     {
       "epoch": 1.94,
-      "learning_rate": 0.00023705202312138726,
-      "loss": 0.4133,
+      "learning_rate": 0.00023722543352601156,
+      "loss": 0.5418,
       "step": 866
     },
     {
       "epoch": 1.94,
-      "learning_rate": 0.00023687861271676298,
-      "loss": 0.6664,
+      "learning_rate": 0.00023705202312138726,
+      "loss": 0.6583,
       "step": 867
     },
     {
       "epoch": 1.95,
-      "learning_rate": 0.0002367052023121387,
-      "loss": 0.5039,
+      "learning_rate": 0.00023687861271676298,
+      "loss": 0.7395,
       "step": 868
     },
     {
       "epoch": 1.95,
-      "learning_rate": 0.00023653179190751445,
-      "loss": 0.6268,
+      "learning_rate": 0.0002367052023121387,
+      "loss": 0.6711,
       "step": 869
     },
     {
       "epoch": 1.95,
-      "learning_rate": 0.00023635838150289017,
-      "loss": 0.4324,
+      "learning_rate": 0.00023653179190751445,
+      "loss": 0.5824,
       "step": 870
     },
     {
       "epoch": 1.95,
-      "learning_rate": 0.00023618497109826586,
-      "loss": 0.5077,
+      "learning_rate": 0.00023635838150289017,
+      "loss": 0.6412,
       "step": 871
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.00023601156069364158,
-      "loss": 0.4605,
+      "learning_rate": 0.00023618497109826586,
+      "loss": 0.5811,
       "step": 872
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.00023583815028901733,
-      "loss": 0.4617,
+      "learning_rate": 0.00023601156069364158,
+      "loss": 0.611,
       "step": 873
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.00023566473988439305,
-      "loss": 0.5816,
+      "learning_rate": 0.00023583815028901733,
+      "loss": 0.6409,
       "step": 874
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.00023549132947976877,
-      "loss": 0.5213,
+      "learning_rate": 0.00023566473988439305,
+      "loss": 0.6914,
       "step": 875
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.0002353179190751445,
-      "loss": 0.4203,
+      "learning_rate": 0.00023549132947976877,
+      "loss": 0.5173,
       "step": 876
     },
     {
       "epoch": 1.97,
-      "learning_rate": 0.00023514450867052024,
-      "loss": 0.4604,
+      "learning_rate": 0.0002353179190751445,
+      "loss": 0.5725,
       "step": 877
     },
     {
       "epoch": 1.97,
-      "learning_rate": 0.00023497109826589593,
-      "loss": 0.4604,
+      "learning_rate": 0.00023514450867052024,
+      "loss": 0.5752,
       "step": 878
     },
     {
       "epoch": 1.97,
-      "learning_rate": 0.00023479768786127165,
-      "loss": 0.5035,
+      "learning_rate": 0.00023497109826589593,
+      "loss": 0.5661,
       "step": 879
     },
     {
       "epoch": 1.97,
-      "learning_rate": 0.00023462427745664737,
-      "loss": 0.3755,
+      "learning_rate": 0.00023479768786127165,
+      "loss": 0.51,
       "step": 880
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.00023445086705202312,
-      "loss": 0.4899,
+      "learning_rate": 0.00023462427745664737,
+      "loss": 0.5564,
       "step": 881
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.00023427745664739884,
-      "loss": 0.4343,
+      "learning_rate": 0.00023445086705202312,
+      "loss": 0.5911,
       "step": 882
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.00023410404624277454,
-      "loss": 0.4574,
+      "learning_rate": 0.00023427745664739884,
+      "loss": 0.5626,
       "step": 883
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.00023393063583815026,
-      "loss": 0.4703,
+      "learning_rate": 0.00023410404624277454,
+      "loss": 0.5812,
       "step": 884
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.000233757225433526,
-      "loss": 0.4654,
+      "learning_rate": 0.00023393063583815026,
+      "loss": 0.6222,
       "step": 885
     },
     {
       "epoch": 1.99,
-      "learning_rate": 0.00023358381502890173,
-      "loss": 0.6055,
+      "learning_rate": 0.000233757225433526,
+      "loss": 0.7041,
       "step": 886
     },
     {
       "epoch": 1.99,
-      "learning_rate": 0.00023341040462427745,
-      "loss": 0.6367,
+      "learning_rate": 0.00023358381502890173,
+      "loss": 0.7997,
       "step": 887
     },
     {
       "epoch": 1.99,
-      "learning_rate": 0.00023323699421965314,
-      "loss": 0.5017,
+      "learning_rate": 0.00023341040462427745,
+      "loss": 0.6179,
       "step": 888
     },
     {
       "epoch": 1.99,
-      "learning_rate": 0.00023306358381502886,
-      "loss": 0.4949,
+      "learning_rate": 0.00023323699421965314,
+      "loss": 0.7034,
       "step": 889
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.0002328901734104046,
-      "loss": 0.5986,
+      "learning_rate": 0.00023306358381502886,
+      "loss": 0.7888,
       "step": 890
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00023271676300578033,
-      "loss": 0.6107,
+      "learning_rate": 0.0002328901734104046,
+      "loss": 0.7726,
       "step": 891
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00023254335260115605,
-      "loss": 0.6763,
+      "learning_rate": 0.00023271676300578033,
+      "loss": 0.7381,
       "step": 892
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.00023236994219653174,
-      "loss": 1.1307,
+      "learning_rate": 0.00023254335260115605,
+      "loss": 3.1339,
       "step": 893
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.0002321965317919075,
-      "loss": 0.6294,
+      "learning_rate": 0.00023236994219653174,
+      "loss": 1.8651,
       "step": 894
     },
     {
       "epoch": 2.01,
-      "learning_rate": 0.0002320231213872832,
-      "loss": 0.5784,
+      "learning_rate": 0.0002321965317919075,
+      "loss": 1.2221,
       "step": 895
     },
     {
       "epoch": 2.01,
-      "learning_rate": 0.00023184971098265893,
-      "loss": 0.5724,
+      "learning_rate": 0.0002320231213872832,
+      "loss": 0.9329,
       "step": 896
     },
     {
       "epoch": 2.01,
-      "learning_rate": 0.00023167630057803465,
-      "loss": 0.5684,
+      "learning_rate": 0.00023184971098265893,
+      "loss": 0.8674,
       "step": 897
     },
     {
       "epoch": 2.01,
-      "learning_rate": 0.0002315028901734104,
-      "loss": 0.5024,
+      "learning_rate": 0.00023167630057803465,
+      "loss": 0.8294,
       "step": 898
     },
     {
       "epoch": 2.02,
-      "learning_rate": 0.00023132947976878612,
-      "loss": 0.4798,
+      "learning_rate": 0.0002315028901734104,
+      "loss": 0.807,
       "step": 899
     },
     {
       "epoch": 2.02,
-      "learning_rate": 0.00023115606936416181,
-      "loss": 0.4793,
+      "learning_rate": 0.00023132947976878612,
+      "loss": 0.6646,
       "step": 900
     },
     {
       "epoch": 2.02,
-      "learning_rate": 0.00023098265895953754,
-      "loss": 0.3836,
+      "learning_rate": 0.00023115606936416181,
+      "loss": 0.5948,
       "step": 901
     },
     {
       "epoch": 2.02,
-      "learning_rate": 0.00023080924855491328,
-      "loss": 0.4929,
+      "learning_rate": 0.00023098265895953754,
+      "loss": 0.7274,
       "step": 902
     },
     {
       "epoch": 2.02,
-      "learning_rate": 0.000230635838150289,
-      "loss": 0.4115,
+      "learning_rate": 0.00023080924855491328,
+      "loss": 0.6304,
       "step": 903
     },
     {
       "epoch": 2.03,
-      "learning_rate": 0.00023046242774566472,
-      "loss": 0.4635,
+      "learning_rate": 0.000230635838150289,
+      "loss": 0.606,
       "step": 904
     },
     {
       "epoch": 2.03,
-      "learning_rate": 0.00023028901734104042,
-      "loss": 0.3872,
+      "learning_rate": 0.00023046242774566472,
+      "loss": 0.5263,
       "step": 905
     },
     {
       "epoch": 2.03,
-      "learning_rate": 0.00023011560693641617,
-      "loss": 0.3065,
+      "learning_rate": 0.00023028901734104042,
+      "loss": 0.4577,
       "step": 906
     },
     {
       "epoch": 2.03,
-      "learning_rate": 0.0002299421965317919,
-      "loss": 0.3686,
+      "learning_rate": 0.00023011560693641617,
+      "loss": 0.4746,
       "step": 907
     },
     {
       "epoch": 2.04,
-      "learning_rate": 0.0002297687861271676,
-      "loss": 0.377,
+      "learning_rate": 0.0002299421965317919,
+      "loss": 0.5186,
       "step": 908
     },
     {
       "epoch": 2.04,
-      "learning_rate": 0.00022959537572254333,
-      "loss": 0.3566,
+      "learning_rate": 0.0002297687861271676,
+      "loss": 0.4732,
       "step": 909
     },
     {
       "epoch": 2.04,
-      "learning_rate": 0.00022942196531791908,
-      "loss": 0.3976,
+      "learning_rate": 0.00022959537572254333,
+      "loss": 0.5486,
       "step": 910
     },
     {
       "epoch": 2.04,
-      "learning_rate": 0.00022924855491329477,
-      "loss": 0.4026,
+      "learning_rate": 0.00022942196531791908,
+      "loss": 0.5361,
       "step": 911
     },
     {
       "epoch": 2.04,
-      "learning_rate": 0.0002290751445086705,
-      "loss": 0.3069,
+      "learning_rate": 0.00022924855491329477,
+      "loss": 0.3955,
       "step": 912
     },
     {
       "epoch": 2.05,
-      "learning_rate": 0.0002289017341040462,
-      "loss": 0.3355,
+      "learning_rate": 0.0002290751445086705,
+      "loss": 0.4232,
       "step": 913
     },
     {
       "epoch": 2.05,
-      "learning_rate": 0.00022872832369942196,
-      "loss": 0.2953,
+      "learning_rate": 0.0002289017341040462,
+      "loss": 0.4127,
       "step": 914
     },
     {
       "epoch": 2.05,
-      "learning_rate": 0.00022855491329479768,
-      "loss": 0.4171,
+      "learning_rate": 0.00022872832369942196,
+      "loss": 0.5143,
       "step": 915
     },
     {
       "epoch": 2.05,
-      "learning_rate": 0.00022838150289017337,
-      "loss": 0.3288,
+      "learning_rate": 0.00022855491329479768,
+      "loss": 0.4155,
       "step": 916
     },
     {
       "epoch": 2.06,
-      "learning_rate": 0.0002282080924855491,
-      "loss": 0.3462,
+      "learning_rate": 0.00022838150289017337,
+      "loss": 0.4304,
       "step": 917
     },
     {
       "epoch": 2.06,
-      "learning_rate": 0.00022803468208092484,
-      "loss": 0.3657,
+      "learning_rate": 0.0002282080924855491,
+      "loss": 0.4103,
       "step": 918
     },
     {
       "epoch": 2.06,
-      "learning_rate": 0.00022786127167630056,
-      "loss": 0.3701,
+      "learning_rate": 0.00022803468208092484,
+      "loss": 0.5457,
       "step": 919
     },
     {
       "epoch": 2.06,
-      "learning_rate": 0.00022768786127167628,
-      "loss": 0.3076,
+      "learning_rate": 0.00022786127167630056,
+      "loss": 0.3894,
       "step": 920
     },
     {
       "epoch": 2.07,
-      "learning_rate": 0.00022751445086705198,
-      "loss": 0.3014,
+      "learning_rate": 0.00022768786127167628,
+      "loss": 0.4171,
       "step": 921
     },
     {
       "epoch": 2.07,
-      "learning_rate": 0.00022734104046242772,
-      "loss": 0.3676,
+      "learning_rate": 0.00022751445086705198,
+      "loss": 0.453,
       "step": 922
     },
     {
       "epoch": 2.07,
-      "learning_rate": 0.00022716763005780344,
-      "loss": 0.322,
+      "learning_rate": 0.00022734104046242772,
+      "loss": 0.4221,
       "step": 923
     },
     {
       "epoch": 2.07,
-      "learning_rate": 0.00022699421965317917,
-      "loss": 0.3089,
+      "learning_rate": 0.00022716763005780344,
+      "loss": 0.3982,
       "step": 924
     },
     {
       "epoch": 2.07,
-      "learning_rate": 0.00022682080924855489,
-      "loss": 0.3649,
+      "learning_rate": 0.00022699421965317917,
+      "loss": 0.4735,
       "step": 925
     },
     {
       "epoch": 2.08,
-      "learning_rate": 0.00022664739884393063,
-      "loss": 0.2995,
+      "learning_rate": 0.00022682080924855489,
+      "loss": 0.4474,
       "step": 926
     },
     {
       "epoch": 2.08,
-      "learning_rate": 0.00022647398843930635,
-      "loss": 0.3248,
+      "learning_rate": 0.00022664739884393063,
+      "loss": 0.4077,
       "step": 927
     },
     {
       "epoch": 2.08,
-      "learning_rate": 0.00022630057803468205,
-      "loss": 0.3239,
+      "learning_rate": 0.00022647398843930635,
+      "loss": 0.3909,
       "step": 928
     },
     {
       "epoch": 2.08,
-      "learning_rate": 0.00022612716763005777,
-      "loss": 0.3388,
+      "learning_rate": 0.00022630057803468205,
+      "loss": 0.4458,
       "step": 929
     },
     {
       "epoch": 2.09,
-      "learning_rate": 0.00022595375722543352,
-      "loss": 0.291,
+      "learning_rate": 0.00022612716763005777,
+      "loss": 0.4183,
       "step": 930
     },
     {
       "epoch": 2.09,
-      "learning_rate": 0.00022578034682080924,
-      "loss": 0.3173,
+      "learning_rate": 0.00022595375722543352,
+      "loss": 0.4089,
       "step": 931
     },
     {
       "epoch": 2.09,
-      "learning_rate": 0.00022560693641618496,
-      "loss": 0.3731,
+      "learning_rate": 0.00022578034682080924,
+      "loss": 0.4871,
       "step": 932
     },
     {
       "epoch": 2.09,
-      "learning_rate": 0.00022543352601156065,
-      "loss": 0.3987,
+      "learning_rate": 0.00022560693641618496,
+      "loss": 0.4871,
       "step": 933
     },
     {
       "epoch": 2.09,
-      "learning_rate": 0.0002252601156069364,
-      "loss": 0.447,
+      "learning_rate": 0.00022543352601156065,
+      "loss": 0.5296,
       "step": 934
     },
     {
       "epoch": 2.1,
-      "learning_rate": 0.00022508670520231212,
-      "loss": 0.2759,
+      "learning_rate": 0.0002252601156069364,
+      "loss": 0.3508,
       "step": 935
     },
     {
       "epoch": 2.1,
-      "learning_rate": 0.00022491329479768784,
-      "loss": 0.2705,
+      "learning_rate": 0.00022508670520231212,
+      "loss": 0.3258,
       "step": 936
     },
     {
       "epoch": 2.1,
-      "learning_rate": 0.00022473988439306356,
-      "loss": 0.3484,
+      "learning_rate": 0.00022491329479768784,
+      "loss": 0.3775,
       "step": 937
     },
     {
       "epoch": 2.1,
-      "learning_rate": 0.0002245664739884393,
-      "loss": 0.3363,
+      "learning_rate": 0.00022473988439306356,
+      "loss": 0.3574,
       "step": 938
     },
     {
       "epoch": 2.11,
-      "learning_rate": 0.000224393063583815,
-      "loss": 0.3411,
+      "learning_rate": 0.0002245664739884393,
+      "loss": 0.4213,
       "step": 939
     },
     {
       "epoch": 2.11,
-      "learning_rate": 0.00022421965317919072,
-      "loss": 0.2715,
+      "learning_rate": 0.000224393063583815,
+      "loss": 0.3603,
       "step": 940
     },
     {
       "epoch": 2.11,
-      "learning_rate": 0.00022404624277456644,
-      "loss": 0.3929,
+      "learning_rate": 0.00022421965317919072,
+      "loss": 0.4983,
       "step": 941
     },
     {
       "epoch": 2.11,
-      "learning_rate": 0.0002238728323699422,
-      "loss": 0.3012,
+      "learning_rate": 0.00022404624277456644,
+      "loss": 0.2657,
       "step": 942
     },
     {
       "epoch": 2.11,
-      "learning_rate": 0.0002236994219653179,
-      "loss": 1.1621,
+      "learning_rate": 0.0002238728323699422,
+      "loss": 1.5798,
       "step": 943
     },
     {
       "epoch": 2.12,
-      "learning_rate": 0.0002235260115606936,
-      "loss": 0.8901,
+      "learning_rate": 0.0002236994219653179,
+      "loss": 1.1552,
       "step": 944
     },
     {
       "epoch": 2.12,
-      "learning_rate": 0.00022335260115606933,
-      "loss": 0.663,
+      "learning_rate": 0.0002235260115606936,
+      "loss": 0.9436,
       "step": 945
     },
     {
       "epoch": 2.12,
-      "learning_rate": 0.00022317919075144507,
-      "loss": 0.5875,
+      "learning_rate": 0.00022335260115606933,
+      "loss": 0.7038,
       "step": 946
     },
     {
       "epoch": 2.12,
-      "learning_rate": 0.0002230057803468208,
-      "loss": 0.5918,
+      "learning_rate": 0.00022317919075144507,
+      "loss": 0.6321,
       "step": 947
     },
     {
       "epoch": 2.13,
-      "learning_rate": 0.00022283236994219652,
-      "loss": 0.6611,
+      "learning_rate": 0.0002230057803468208,
+      "loss": 0.656,
       "step": 948
     },
     {
       "epoch": 2.13,
-      "learning_rate": 0.00022265895953757224,
-      "loss": 0.521,
+      "learning_rate": 0.00022283236994219652,
+      "loss": 0.613,
       "step": 949
     },
     {
       "epoch": 2.13,
-      "learning_rate": 0.00022248554913294798,
-      "loss": 0.4735,
+      "learning_rate": 0.00022265895953757224,
+      "loss": 0.4863,
       "step": 950
     },
     {
       "epoch": 2.13,
-      "learning_rate": 0.00022231213872832368,
-      "loss": 0.4599,
+      "learning_rate": 0.00022248554913294798,
+      "loss": 0.6139,
       "step": 951
     },
     {
       "epoch": 2.13,
-      "learning_rate": 0.0002221387283236994,
-      "loss": 0.4246,
+      "learning_rate": 0.00022231213872832368,
+      "loss": 0.5303,
       "step": 952
     },
     {
       "epoch": 2.14,
-      "learning_rate": 0.00022196531791907512,
-      "loss": 0.4222,
+      "learning_rate": 0.0002221387283236994,
+      "loss": 0.4703,
       "step": 953
     },
     {
       "epoch": 2.14,
-      "learning_rate": 0.00022179190751445087,
-      "loss": 0.4073,
+      "learning_rate": 0.00022196531791907512,
+      "loss": 0.5082,
       "step": 954
     },
     {
       "epoch": 2.14,
-      "learning_rate": 0.0002216184971098266,
-      "loss": 0.2894,
+      "learning_rate": 0.00022179190751445087,
+      "loss": 0.3705,
       "step": 955
     },
     {
       "epoch": 2.14,
-      "learning_rate": 0.00022144508670520228,
-      "loss": 0.3121,
+      "learning_rate": 0.0002216184971098266,
+      "loss": 0.3861,
       "step": 956
     },
     {
       "epoch": 2.15,
-      "learning_rate": 0.000221271676300578,
-      "loss": 0.4046,
+      "learning_rate": 0.00022144508670520228,
+      "loss": 0.4343,
       "step": 957
     },
     {
       "epoch": 2.15,
-      "learning_rate": 0.00022109826589595375,
-      "loss": 0.3066,
+      "learning_rate": 0.000221271676300578,
+      "loss": 0.3629,
       "step": 958
     },
     {
       "epoch": 2.15,
-      "learning_rate": 0.00022092485549132947,
-      "loss": 0.2934,
+      "learning_rate": 0.00022109826589595375,
+      "loss": 0.3247,
       "step": 959
     },
     {
       "epoch": 2.15,
-      "learning_rate": 0.0002207514450867052,
-      "loss": 0.3144,
+      "learning_rate": 0.00022092485549132947,
+      "loss": 0.3734,
       "step": 960
     },
     {
       "epoch": 2.15,
-      "learning_rate": 0.00022057803468208088,
-      "loss": 0.3406,
+      "learning_rate": 0.0002207514450867052,
+      "loss": 0.4231,
       "step": 961
     },
     {
       "epoch": 2.16,
-      "learning_rate": 0.00022040462427745663,
-      "loss": 0.2953,
+      "learning_rate": 0.00022057803468208088,
+      "loss": 0.369,
       "step": 962
     },
     {
       "epoch": 2.16,
-      "learning_rate": 0.00022023121387283235,
-      "loss": 0.311,
+      "learning_rate": 0.00022040462427745663,
+      "loss": 0.3606,
       "step": 963
     },
     {
       "epoch": 2.16,
-      "learning_rate": 0.00022005780346820807,
-      "loss": 0.3591,
+      "learning_rate": 0.00022023121387283235,
+      "loss": 0.4322,
       "step": 964
     },
     {
       "epoch": 2.16,
-      "learning_rate": 0.0002198843930635838,
-      "loss": 0.3022,
+      "learning_rate": 0.00022005780346820807,
+      "loss": 0.3928,
       "step": 965
     },
     {
       "epoch": 2.17,
-      "learning_rate": 0.00021971098265895954,
-      "loss": 0.3229,
+      "learning_rate": 0.0002198843930635838,
+      "loss": 0.4072,
       "step": 966
     },
     {
       "epoch": 2.17,
-      "learning_rate": 0.00021953757225433524,
-      "loss": 0.2787,
+      "learning_rate": 0.00021971098265895954,
+      "loss": 0.3673,
       "step": 967
     },
     {
       "epoch": 2.17,
-      "learning_rate": 0.00021936416184971096,
-      "loss": 0.3134,
+      "learning_rate": 0.00021953757225433524,
+      "loss": 0.3822,
       "step": 968
     },
     {
       "epoch": 2.17,
-      "learning_rate": 0.00021919075144508668,
-      "loss": 0.3083,
+      "learning_rate": 0.00021936416184971096,
+      "loss": 0.3869,
       "step": 969
     },
     {
       "epoch": 2.17,
-      "learning_rate": 0.00021901734104046243,
-      "loss": 0.3141,
+      "learning_rate": 0.00021919075144508668,
+      "loss": 0.3596,
       "step": 970
     },
     {
       "epoch": 2.18,
-      "learning_rate": 0.00021884393063583815,
-      "loss": 0.3645,
+      "learning_rate": 0.00021901734104046243,
+      "loss": 0.4073,
       "step": 971
     },
     {
       "epoch": 2.18,
-      "learning_rate": 0.00021867052023121384,
-      "loss": 0.3452,
+      "learning_rate": 0.00021884393063583815,
+      "loss": 0.3531,
       "step": 972
     },
     {
       "epoch": 2.18,
-      "learning_rate": 0.00021849710982658956,
-      "loss": 0.3692,
+      "learning_rate": 0.00021867052023121384,
+      "loss": 0.4388,
       "step": 973
     },
     {
       "epoch": 2.18,
-      "learning_rate": 0.0002183236994219653,
-      "loss": 0.314,
+      "learning_rate": 0.00021849710982658956,
+      "loss": 0.3399,
       "step": 974
     },
     {
       "epoch": 2.19,
-      "learning_rate": 0.00021815028901734103,
-      "loss": 0.2614,
+      "learning_rate": 0.0002183236994219653,
+      "loss": 0.286,
       "step": 975
     },
     {
       "epoch": 2.19,
-      "learning_rate": 0.00021797687861271675,
-      "loss": 0.3473,
+      "learning_rate": 0.00021815028901734103,
+      "loss": 0.3678,
       "step": 976
     },
     {
       "epoch": 2.19,
-      "learning_rate": 0.00021780346820809247,
-      "loss": 0.36,
+      "learning_rate": 0.00021797687861271675,
+      "loss": 0.402,
       "step": 977
     },
     {
       "epoch": 2.19,
-      "learning_rate": 0.00021763005780346822,
-      "loss": 0.3494,
+      "learning_rate": 0.00021780346820809247,
+      "loss": 0.4259,
       "step": 978
     },
     {
       "epoch": 2.2,
-      "learning_rate": 0.0002174566473988439,
-      "loss": 0.3341,
+      "learning_rate": 0.00021763005780346822,
+      "loss": 0.3776,
       "step": 979
     },
     {
       "epoch": 2.2,
-      "learning_rate": 0.00021728323699421963,
-      "loss": 0.3583,
+      "learning_rate": 0.0002174566473988439,
+      "loss": 0.3641,
       "step": 980
     },
     {
       "epoch": 2.2,
-      "learning_rate": 0.00021710982658959535,
-      "loss": 0.2914,
+      "learning_rate": 0.00021728323699421963,
+      "loss": 0.3582,
       "step": 981
     },
     {
       "epoch": 2.2,
-      "learning_rate": 0.0002169364161849711,
-      "loss": 0.2711,
+      "learning_rate": 0.00021710982658959535,
+      "loss": 0.4073,
       "step": 982
     },
     {
       "epoch": 2.2,
-      "learning_rate": 0.00021676300578034682,
-      "loss": 0.4029,
+      "learning_rate": 0.0002169364161849711,
+      "loss": 0.4097,
       "step": 983
     },
     {
       "epoch": 2.21,
-      "learning_rate": 0.00021658959537572251,
-      "loss": 0.3212,
+      "learning_rate": 0.00021676300578034682,
+      "loss": 0.3869,
       "step": 984
     },
     {
       "epoch": 2.21,
-      "learning_rate": 0.00021641618497109824,
-      "loss": 0.2561,
+      "learning_rate": 0.00021658959537572251,
+      "loss": 0.3229,
       "step": 985
     },
     {
       "epoch": 2.21,
-      "learning_rate": 0.00021624277456647396,
-      "loss": 0.2668,
+      "learning_rate": 0.00021641618497109824,
+      "loss": 0.304,
       "step": 986
     },
     {
       "epoch": 2.21,
-      "learning_rate": 0.0002160693641618497,
-      "loss": 0.3586,
+      "learning_rate": 0.00021624277456647396,
+      "loss": 0.3122,
       "step": 987
     },
     {
       "epoch": 2.22,
-      "learning_rate": 0.00021589595375722542,
-      "loss": 0.3214,
+      "learning_rate": 0.0002160693641618497,
+      "loss": 0.3282,
       "step": 988
     },
     {
       "epoch": 2.22,
-      "learning_rate": 0.00021572254335260112,
-      "loss": 0.2136,
+      "learning_rate": 0.00021589595375722542,
+      "loss": 0.3119,
       "step": 989
     },
     {
       "epoch": 2.22,
-      "learning_rate": 0.00021554913294797684,
-      "loss": 0.3358,
+      "learning_rate": 0.00021572254335260112,
+      "loss": 0.4257,
       "step": 990
     },
     {
       "epoch": 2.22,
-      "learning_rate": 0.00021537572254335259,
-      "loss": 0.3889,
+      "learning_rate": 0.00021554913294797684,
+      "loss": 0.4143,
       "step": 991
     },
     {
       "epoch": 2.22,
-      "learning_rate": 0.0002152023121387283,
-      "loss": 0.3912,
+      "learning_rate": 0.00021537572254335259,
+      "loss": 0.4064,
       "step": 992
     },
     {
       "epoch": 2.23,
-      "learning_rate": 0.00021502890173410403,
-      "loss": 1.052,
+      "learning_rate": 0.0002152023121387283,
+      "loss": 1.685,
       "step": 993
     },
     {
       "epoch": 2.23,
-      "learning_rate": 0.00021485549132947972,
-      "loss": 0.7914,
+      "learning_rate": 0.00021502890173410403,
+      "loss": 1.1262,
       "step": 994
     },
     {
       "epoch": 2.23,
-      "learning_rate": 0.00021468208092485547,
-      "loss": 0.6188,
+      "learning_rate": 0.00021485549132947972,
+      "loss": 0.8902,
       "step": 995
     },
     {
       "epoch": 2.23,
-      "learning_rate": 0.0002145086705202312,
-      "loss": 0.5849,
+      "learning_rate": 0.00021468208092485547,
+      "loss": 0.7312,
       "step": 996
     },
     {
       "epoch": 2.24,
-      "learning_rate": 0.0002143352601156069,
-      "loss": 0.4353,
+      "learning_rate": 0.0002145086705202312,
+      "loss": 0.5592,
       "step": 997
     },
     {
       "epoch": 2.24,
-      "learning_rate": 0.00021416184971098263,
-      "loss": 0.5717,
+      "learning_rate": 0.0002143352601156069,
+      "loss": 0.6647,
       "step": 998
     },
     {
       "epoch": 2.24,
-      "learning_rate": 0.00021398843930635838,
-      "loss": 0.5272,
+      "learning_rate": 0.00021416184971098263,
+      "loss": 0.5828,
       "step": 999
     },
     {
       "epoch": 2.24,
-      "learning_rate": 0.0002138150289017341,
-      "loss": 0.464,
+      "learning_rate": 0.00021398843930635838,
+      "loss": 0.5486,
       "step": 1000
     },
     {
       "epoch": 2.24,
-      "eval_loss": 0.4856905937194824,
-      "eval_runtime": 570.5354,
-      "eval_samples_per_second": 4.631,
-      "eval_steps_per_second": 0.58,
-      "eval_wer": 0.14488298294327648,
+      "eval_loss": 0.5549545288085938,
+      "eval_runtime": 557.4926,
+      "eval_samples_per_second": 4.739,
+      "eval_steps_per_second": 0.594,
+      "eval_wer": 0.17532725109083697,
       "step": 1000
     },
     {
       "epoch": 2.24,
-      "learning_rate": 0.0002136416184971098,
-      "loss": 0.4389,
+      "learning_rate": 0.0002138150289017341,
+      "loss": 0.5099,
       "step": 1001
     },
     {
       "epoch": 2.25,
-      "learning_rate": 0.00021346820809248551,
-      "loss": 0.3892,
+      "learning_rate": 0.0002136416184971098,
+      "loss": 0.4194,
       "step": 1002
     },
     {
       "epoch": 2.25,
-      "learning_rate": 0.00021329479768786126,
-      "loss": 0.3714,
+      "learning_rate": 0.00021346820809248551,
+      "loss": 0.4526,
       "step": 1003
     },
     {
       "epoch": 2.25,
-      "learning_rate": 0.00021312138728323698,
-      "loss": 0.4208,
+      "learning_rate": 0.00021329479768786126,
+      "loss": 0.499,
       "step": 1004
     },
     {
       "epoch": 2.25,
-      "learning_rate": 0.0002129479768786127,
-      "loss": 0.3307,
+      "learning_rate": 0.00021312138728323698,
+      "loss": 0.4094,
       "step": 1005
     },
     {
       "epoch": 2.26,
-      "learning_rate": 0.0002127745664739884,
-      "loss": 0.2909,
+      "learning_rate": 0.0002129479768786127,
+      "loss": 0.3473,
       "step": 1006
     },
     {
       "epoch": 2.26,
-      "learning_rate": 0.00021260115606936414,
-      "loss": 0.388,
+      "learning_rate": 0.0002127745664739884,
+      "loss": 0.4647,
       "step": 1007
     },
     {
       "epoch": 2.26,
-      "learning_rate": 0.00021242774566473987,
-      "loss": 0.3202,
+      "learning_rate": 0.00021260115606936414,
+      "loss": 0.3955,
       "step": 1008
     },
     {
       "epoch": 2.26,
-      "learning_rate": 0.00021225433526011559,
-      "loss": 0.313,
+      "learning_rate": 0.00021242774566473987,
+      "loss": 0.4039,
       "step": 1009
     },
     {
       "epoch": 2.26,
-      "learning_rate": 0.0002120809248554913,
-      "loss": 0.3205,
+      "learning_rate": 0.00021225433526011559,
+      "loss": 0.4174,
       "step": 1010
     },
     {
       "epoch": 2.27,
-      "learning_rate": 0.00021190751445086705,
-      "loss": 0.364,
+      "learning_rate": 0.0002120809248554913,
+      "loss": 0.3858,
       "step": 1011
     },
     {
       "epoch": 2.27,
-      "learning_rate": 0.00021173410404624275,
-      "loss": 0.4024,
+      "learning_rate": 0.00021190751445086705,
+      "loss": 0.4267,
       "step": 1012
     },
     {
       "epoch": 2.27,
-      "learning_rate": 0.00021156069364161847,
-      "loss": 0.3499,
+      "learning_rate": 0.00021173410404624275,
+      "loss": 0.4297,
       "step": 1013
     },
     {
       "epoch": 2.27,
-      "learning_rate": 0.0002113872832369942,
-      "loss": 0.3411,
+      "learning_rate": 0.00021156069364161847,
+      "loss": 0.4191,
       "step": 1014
     },
     {
       "epoch": 2.28,
-      "learning_rate": 0.00021121387283236994,
-      "loss": 0.3433,
+      "learning_rate": 0.0002113872832369942,
+      "loss": 0.402,
       "step": 1015
     },
     {
       "epoch": 2.28,
-      "learning_rate": 0.00021104046242774566,
-      "loss": 0.2937,
+      "learning_rate": 0.00021121387283236994,
+      "loss": 0.3399,
       "step": 1016
     },
     {
       "epoch": 2.28,
-      "learning_rate": 0.00021086705202312135,
-      "loss": 0.3456,
+      "learning_rate": 0.00021104046242774566,
+      "loss": 0.3722,
       "step": 1017
     },
     {
       "epoch": 2.28,
-      "learning_rate": 0.00021069364161849707,
-      "loss": 0.2884,
+      "learning_rate": 0.00021086705202312135,
+      "loss": 0.3687,
       "step": 1018
     },
     {
       "epoch": 2.28,
-      "learning_rate": 0.00021052023121387282,
-      "loss": 0.3717,
+      "learning_rate": 0.00021069364161849707,
+      "loss": 0.4594,
       "step": 1019
     },
     {
       "epoch": 2.29,
-      "learning_rate": 0.00021034682080924854,
-      "loss": 0.3088,
+      "learning_rate": 0.00021052023121387282,
+      "loss": 0.3473,
       "step": 1020
     },
     {
       "epoch": 2.29,
-      "learning_rate": 0.00021017341040462426,
-      "loss": 0.2876,
+      "learning_rate": 0.00021034682080924854,
+      "loss": 0.3437,
       "step": 1021
     },
     {
       "epoch": 2.29,
-      "learning_rate": 0.00020999999999999998,
-      "loss": 0.3473,
+      "learning_rate": 0.00021017341040462426,
+      "loss": 0.3998,
       "step": 1022
     },
     {
       "epoch": 2.29,
-      "learning_rate": 0.00020982658959537573,
-      "loss": 0.3416,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 0.3963,
       "step": 1023
     },
     {
       "epoch": 2.3,
-      "learning_rate": 0.00020965317919075142,
-      "loss": 0.2797,
+      "learning_rate": 0.00020982658959537573,
+      "loss": 0.3373,
       "step": 1024
     },
     {
       "epoch": 2.3,
-      "learning_rate": 0.00020947976878612714,
-      "loss": 0.2986,
+      "learning_rate": 0.00020965317919075142,
+      "loss": 0.3339,
       "step": 1025
     },
     {
       "epoch": 2.3,
-      "learning_rate": 0.00020930635838150286,
-      "loss": 0.3016,
+      "learning_rate": 0.00020947976878612714,
+      "loss": 0.3122,
       "step": 1026
     },
     {
       "epoch": 2.3,
-      "learning_rate": 0.0002091329479768786,
-      "loss": 0.3332,
+      "learning_rate": 0.00020930635838150286,
+      "loss": 0.3959,
       "step": 1027
     },
     {
       "epoch": 2.3,
-      "learning_rate": 0.00020895953757225433,
-      "loss": 0.3596,
+      "learning_rate": 0.0002091329479768786,
+      "loss": 0.38,
       "step": 1028
     },
     {
       "epoch": 2.31,
-      "learning_rate": 0.00020878612716763003,
-      "loss": 0.2761,
+      "learning_rate": 0.00020895953757225433,
+      "loss": 0.2978,
       "step": 1029
     },
     {
       "epoch": 2.31,
-      "learning_rate": 0.00020861271676300575,
-      "loss": 0.2575,
+      "learning_rate": 0.00020878612716763003,
+      "loss": 0.309,
       "step": 1030
     },
     {
       "epoch": 2.31,
-      "learning_rate": 0.0002084393063583815,
-      "loss": 0.2251,
+      "learning_rate": 0.00020861271676300575,
+      "loss": 0.3022,
       "step": 1031
     },
     {
       "epoch": 2.31,
-      "learning_rate": 0.00020826589595375722,
-      "loss": 0.3015,
+      "learning_rate": 0.0002084393063583815,
+      "loss": 0.3544,
       "step": 1032
     },
     {
       "epoch": 2.32,
-      "learning_rate": 0.00020809248554913294,
-      "loss": 0.3175,
+      "learning_rate": 0.00020826589595375722,
+      "loss": 0.3822,
       "step": 1033
     },
     {
       "epoch": 2.32,
-      "learning_rate": 0.00020791907514450863,
-      "loss": 0.2838,
+      "learning_rate": 0.00020809248554913294,
+      "loss": 0.3033,
       "step": 1034
     },
     {
       "epoch": 2.32,
-      "learning_rate": 0.00020774566473988438,
-      "loss": 0.3077,
+      "learning_rate": 0.00020791907514450863,
+      "loss": 0.3179,
       "step": 1035
     },
     {
       "epoch": 2.32,
-      "learning_rate": 0.0002075722543352601,
-      "loss": 0.3122,
+      "learning_rate": 0.00020774566473988438,
+      "loss": 0.3592,
       "step": 1036
     },
     {
       "epoch": 2.33,
-      "learning_rate": 0.00020739884393063582,
-      "loss": 0.275,
+      "learning_rate": 0.0002075722543352601,
+      "loss": 0.293,
       "step": 1037
     },
     {
       "epoch": 2.33,
-      "learning_rate": 0.00020722543352601154,
-      "loss": 0.4268,
+      "learning_rate": 0.00020739884393063582,
+      "loss": 0.4728,
       "step": 1038
     },
     {
       "epoch": 2.33,
-      "learning_rate": 0.0002070520231213873,
-      "loss": 0.3256,
+      "learning_rate": 0.00020722543352601154,
+      "loss": 0.3278,
       "step": 1039
     },
     {
       "epoch": 2.33,
-      "learning_rate": 0.00020687861271676298,
-      "loss": 0.3512,
+      "learning_rate": 0.0002070520231213873,
+      "loss": 0.3808,
       "step": 1040
     },
     {
       "epoch": 2.33,
-      "learning_rate": 0.0002067052023121387,
-      "loss": 0.2547,
+      "learning_rate": 0.00020687861271676298,
+      "loss": 0.3197,
       "step": 1041
     },
     {
       "epoch": 2.34,
-      "learning_rate": 0.00020653179190751442,
-      "loss": 0.3941,
+      "learning_rate": 0.0002067052023121387,
+      "loss": 0.3421,
       "step": 1042
     },
     {
       "epoch": 2.34,
-      "learning_rate": 0.00020635838150289017,
-      "loss": 0.8999,
+      "learning_rate": 0.00020653179190751442,
+      "loss": 0.9862,
       "step": 1043
     },
     {
       "epoch": 2.34,
-      "learning_rate": 0.0002061849710982659,
-      "loss": 0.8205,
+      "learning_rate": 0.00020635838150289017,
+      "loss": 0.9364,
       "step": 1044
     },
     {
       "epoch": 2.34,
-      "learning_rate": 0.00020601156069364158,
-      "loss": 0.4755,
+      "learning_rate": 0.0002061849710982659,
+      "loss": 0.5846,
       "step": 1045
     },
     {
       "epoch": 2.35,
-      "learning_rate": 0.0002058381502890173,
-      "loss": 0.5231,
+      "learning_rate": 0.00020601156069364158,
+      "loss": 0.5967,
       "step": 1046
     },
     {
       "epoch": 2.35,
-      "learning_rate": 0.00020566473988439305,
-      "loss": 0.5203,
+      "learning_rate": 0.0002058381502890173,
+      "loss": 0.4982,
       "step": 1047
     },
     {
       "epoch": 2.35,
-      "learning_rate": 0.00020549132947976877,
-      "loss": 0.4873,
+      "learning_rate": 0.00020566473988439305,
+      "loss": 0.4853,
       "step": 1048
     },
     {
       "epoch": 2.35,
-      "learning_rate": 0.0002053179190751445,
-      "loss": 0.4232,
+      "learning_rate": 0.00020549132947976877,
+      "loss": 0.4355,
       "step": 1049
     },
     {
       "epoch": 2.35,
-      "learning_rate": 0.00020514450867052021,
-      "loss": 0.4999,
+      "learning_rate": 0.0002053179190751445,
+      "loss": 0.5347,
       "step": 1050
     },
     {
       "epoch": 2.36,
-      "learning_rate": 0.00020497109826589596,
-      "loss": 0.4522,
+      "learning_rate": 0.00020514450867052021,
+      "loss": 0.4829,
       "step": 1051
     },
     {
       "epoch": 2.36,
-      "learning_rate": 0.00020479768786127166,
-      "loss": 0.2964,
+      "learning_rate": 0.00020497109826589596,
+      "loss": 0.3672,
       "step": 1052
     },
     {
       "epoch": 2.36,
-      "learning_rate": 0.00020462427745664738,
-      "loss": 0.3664,
+      "learning_rate": 0.00020479768786127166,
+      "loss": 0.4293,
       "step": 1053
     },
     {
       "epoch": 2.36,
-      "learning_rate": 0.0002044508670520231,
-      "loss": 0.3669,
+      "learning_rate": 0.00020462427745664738,
+      "loss": 0.4222,
       "step": 1054
     },
     {
       "epoch": 2.37,
-      "learning_rate": 0.00020427745664739885,
-      "loss": 0.3585,
+      "learning_rate": 0.0002044508670520231,
+      "loss": 0.4015,
       "step": 1055
     },
     {
       "epoch": 2.37,
-      "learning_rate": 0.00020410404624277457,
-      "loss": 0.4454,
+      "learning_rate": 0.00020427745664739885,
+      "loss": 0.4734,
       "step": 1056
     },
     {
       "epoch": 2.37,
-      "learning_rate": 0.00020393063583815026,
-      "loss": 0.348,
+      "learning_rate": 0.00020410404624277457,
+      "loss": 0.3764,
       "step": 1057
     },
     {
       "epoch": 2.37,
-      "learning_rate": 0.00020375722543352598,
-      "loss": 0.4164,
+      "learning_rate": 0.00020393063583815026,
+      "loss": 0.4224,
       "step": 1058
     },
     {
       "epoch": 2.37,
-      "learning_rate": 0.00020358381502890173,
-      "loss": 0.324,
+      "learning_rate": 0.00020375722543352598,
+      "loss": 0.3608,
       "step": 1059
     },
     {
       "epoch": 2.38,
-      "learning_rate": 0.00020341040462427745,
-      "loss": 0.3249,
+      "learning_rate": 0.00020358381502890173,
+      "loss": 0.3688,
       "step": 1060
     },
     {
       "epoch": 2.38,
-      "learning_rate": 0.00020323699421965317,
-      "loss": 0.3287,
+      "learning_rate": 0.00020341040462427745,
+      "loss": 0.3761,
       "step": 1061
     },
     {
       "epoch": 2.38,
-      "learning_rate": 0.00020306358381502886,
-      "loss": 0.323,
+      "learning_rate": 0.00020323699421965317,
+      "loss": 0.3932,
       "step": 1062
     },
     {
       "epoch": 2.38,
-      "learning_rate": 0.0002028901734104046,
-      "loss": 0.3061,
+      "learning_rate": 0.00020306358381502886,
+      "loss": 0.4005,
       "step": 1063
     },
     {
       "epoch": 2.39,
-      "learning_rate": 0.00020271676300578033,
-      "loss": 0.3294,
+      "learning_rate": 0.0002028901734104046,
+      "loss": 0.4029,
       "step": 1064
     },
     {
       "epoch": 2.39,
-      "learning_rate": 0.00020254335260115605,
-      "loss": 0.2192,
+      "learning_rate": 0.00020271676300578033,
+      "loss": 0.2571,
       "step": 1065
     },
     {
       "epoch": 2.39,
-      "learning_rate": 0.00020236994219653177,
-      "loss": 0.275,
+      "learning_rate": 0.00020254335260115605,
+      "loss": 0.2915,
       "step": 1066
     },
     {
       "epoch": 2.39,
-      "learning_rate": 0.00020219653179190752,
-      "loss": 0.282,
+      "learning_rate": 0.00020236994219653177,
+      "loss": 0.3363,
       "step": 1067
     },
     {
       "epoch": 2.39,
-      "learning_rate": 0.00020202312138728321,
-      "loss": 0.2662,
+      "learning_rate": 0.00020219653179190752,
+      "loss": 0.3423,
       "step": 1068
     },
     {
       "epoch": 2.4,
-      "learning_rate": 0.00020184971098265893,
-      "loss": 0.2716,
+      "learning_rate": 0.00020202312138728321,
+      "loss": 0.3071,
       "step": 1069
     },
     {
       "epoch": 2.4,
-      "learning_rate": 0.00020167630057803466,
-      "loss": 0.3011,
+      "learning_rate": 0.00020184971098265893,
+      "loss": 0.3336,
       "step": 1070
     },
     {
       "epoch": 2.4,
-      "learning_rate": 0.0002015028901734104,
-      "loss": 0.2728,
+      "learning_rate": 0.00020167630057803466,
+      "loss": 0.3494,
       "step": 1071
     },
     {
       "epoch": 2.4,
-      "learning_rate": 0.00020132947976878612,
-      "loss": 0.3624,
+      "learning_rate": 0.0002015028901734104,
+      "loss": 0.3916,
       "step": 1072
     },
     {
       "epoch": 2.41,
-      "learning_rate": 0.00020115606936416184,
-      "loss": 0.2688,
+      "learning_rate": 0.00020132947976878612,
+      "loss": 0.3232,
       "step": 1073
     },
     {
       "epoch": 2.41,
-      "learning_rate": 0.00020098265895953754,
-      "loss": 0.2764,
+      "learning_rate": 0.00020115606936416184,
+      "loss": 0.325,
       "step": 1074
     },
     {
       "epoch": 2.41,
-      "learning_rate": 0.00020080924855491329,
-      "loss": 0.2469,
+      "learning_rate": 0.00020098265895953754,
+      "loss": 0.2942,
       "step": 1075
     },
     {
       "epoch": 2.41,
-      "learning_rate": 0.000200635838150289,
-      "loss": 0.2857,
+      "learning_rate": 0.00020080924855491329,
+      "loss": 0.3155,
       "step": 1076
     },
     {
       "epoch": 2.41,
-      "learning_rate": 0.00020046242774566473,
-      "loss": 0.3131,
+      "learning_rate": 0.000200635838150289,
+      "loss": 0.3347,
       "step": 1077
     },
     {
       "epoch": 2.42,
-      "learning_rate": 0.00020028901734104045,
-      "loss": 0.2928,
+      "learning_rate": 0.00020046242774566473,
+      "loss": 0.3422,
       "step": 1078
     },
     {
       "epoch": 2.42,
-      "learning_rate": 0.0002001156069364162,
-      "loss": 0.3083,
+      "learning_rate": 0.00020028901734104045,
+      "loss": 0.3474,
       "step": 1079
     },
     {
       "epoch": 2.42,
-      "learning_rate": 0.0001999421965317919,
-      "loss": 0.2648,
+      "learning_rate": 0.0002001156069364162,
+      "loss": 0.3095,
       "step": 1080
     },
     {
       "epoch": 2.42,
-      "learning_rate": 0.0001997687861271676,
-      "loss": 0.2294,
+      "learning_rate": 0.0001999421965317919,
+      "loss": 0.2851,
       "step": 1081
     },
     {
       "epoch": 2.43,
-      "learning_rate": 0.00019959537572254333,
-      "loss": 0.3392,
+      "learning_rate": 0.0001997687861271676,
+      "loss": 0.3231,
       "step": 1082
     },
     {
       "epoch": 2.43,
-      "learning_rate": 0.00019942196531791905,
-      "loss": 0.3088,
+      "learning_rate": 0.00019959537572254333,
+      "loss": 0.329,
       "step": 1083
     },
     {
       "epoch": 2.43,
-      "learning_rate": 0.0001992485549132948,
-      "loss": 0.2787,
+      "learning_rate": 0.00019942196531791905,
+      "loss": 0.2926,
       "step": 1084
     },
     {
       "epoch": 2.43,
-      "learning_rate": 0.0001990751445086705,
-      "loss": 0.334,
+      "learning_rate": 0.0001992485549132948,
+      "loss": 0.4081,
       "step": 1085
     },
     {
       "epoch": 2.43,
-      "learning_rate": 0.0001989017341040462,
-      "loss": 0.2807,
+      "learning_rate": 0.0001990751445086705,
+      "loss": 0.3161,
       "step": 1086
     },
     {
       "epoch": 2.44,
-      "learning_rate": 0.00019872832369942193,
-      "loss": 0.3032,
+      "learning_rate": 0.0001989017341040462,
+      "loss": 0.3599,
       "step": 1087
     },
     {
       "epoch": 2.44,
-      "learning_rate": 0.00019855491329479768,
-      "loss": 0.29,
+      "learning_rate": 0.00019872832369942193,
+      "loss": 0.371,
       "step": 1088
     },
     {
       "epoch": 2.44,
-      "learning_rate": 0.0001983815028901734,
-      "loss": 0.3273,
+      "learning_rate": 0.00019855491329479768,
+      "loss": 0.4117,
       "step": 1089
     },
     {
       "epoch": 2.44,
-      "learning_rate": 0.0001982080924855491,
-      "loss": 0.291,
+      "learning_rate": 0.0001983815028901734,
+      "loss": 0.2874,
       "step": 1090
     },
     {
       "epoch": 2.45,
-      "learning_rate": 0.00019803468208092482,
-      "loss": 0.273,
+      "learning_rate": 0.0001982080924855491,
+      "loss": 0.2715,
       "step": 1091
     },
     {
       "epoch": 2.45,
-      "learning_rate": 0.00019786127167630056,
-      "loss": 0.3742,
+      "learning_rate": 0.00019803468208092482,
+      "loss": 0.4361,
       "step": 1092
     },
     {
       "epoch": 2.45,
-      "learning_rate": 0.00019768786127167629,
-      "loss": 0.9491,
+      "learning_rate": 0.00019786127167630056,
+      "loss": 0.9283,
       "step": 1093
     },
     {
       "epoch": 2.45,
-      "learning_rate": 0.000197514450867052,
-      "loss": 0.7334,
+      "learning_rate": 0.00019768786127167629,
+      "loss": 0.7911,
       "step": 1094
     },
     {
       "epoch": 2.46,
-      "learning_rate": 0.0001973410404624277,
-      "loss": 0.6034,
+      "learning_rate": 0.000197514450867052,
+      "loss": 0.6643,
       "step": 1095
     },
     {
       "epoch": 2.46,
-      "learning_rate": 0.00019716763005780345,
-      "loss": 0.4765,
+      "learning_rate": 0.0001973410404624277,
+      "loss": 0.56,
       "step": 1096
     },
     {
       "epoch": 2.46,
-      "learning_rate": 0.00019699421965317917,
-      "loss": 0.4759,
+      "learning_rate": 0.00019716763005780345,
+      "loss": 0.4258,
       "step": 1097
     },
     {
       "epoch": 2.46,
-      "learning_rate": 0.0001968208092485549,
-      "loss": 0.4384,
+      "learning_rate": 0.00019699421965317917,
+      "loss": 0.4349,
       "step": 1098
     },
     {
       "epoch": 2.46,
-      "learning_rate": 0.0001966473988439306,
-      "loss": 0.4113,
+      "learning_rate": 0.0001968208092485549,
+      "loss": 0.4175,
       "step": 1099
     },
     {
       "epoch": 2.47,
-      "learning_rate": 0.00019647398843930636,
-      "loss": 0.4254,
+      "learning_rate": 0.0001966473988439306,
+      "loss": 0.453,
       "step": 1100
     },
     {
       "epoch": 2.47,
-      "learning_rate": 0.00019630057803468208,
-      "loss": 0.361,
+      "learning_rate": 0.00019647398843930636,
+      "loss": 0.4161,
       "step": 1101
     },
     {
       "epoch": 2.47,
-      "learning_rate": 0.00019612716763005777,
-      "loss": 0.401,
+      "learning_rate": 0.00019630057803468208,
+      "loss": 0.4631,
       "step": 1102
     },
     {
       "epoch": 2.47,
-      "learning_rate": 0.0001959537572254335,
-      "loss": 0.3375,
+      "learning_rate": 0.00019612716763005777,
+      "loss": 0.3479,
       "step": 1103
     },
     {
       "epoch": 2.48,
-      "learning_rate": 0.00019578034682080924,
-      "loss": 0.3193,
+      "learning_rate": 0.0001959537572254335,
+      "loss": 0.3392,
       "step": 1104
     },
     {
       "epoch": 2.48,
-      "learning_rate": 0.00019560693641618496,
-      "loss": 0.3294,
+      "learning_rate": 0.00019578034682080924,
+      "loss": 0.3378,
       "step": 1105
     },
     {
       "epoch": 2.48,
-      "learning_rate": 0.00019543352601156068,
-      "loss": 0.3724,
+      "learning_rate": 0.00019560693641618496,
+      "loss": 0.3503,
       "step": 1106
     },
     {
       "epoch": 2.48,
-      "learning_rate": 0.00019526011560693637,
-      "loss": 0.2624,
+      "learning_rate": 0.00019543352601156068,
+      "loss": 0.3082,
       "step": 1107
     },
     {
       "epoch": 2.48,
-      "learning_rate": 0.00019508670520231212,
-      "loss": 0.3209,
+      "learning_rate": 0.00019526011560693637,
+      "loss": 0.3127,
       "step": 1108
     },
     {
       "epoch": 2.49,
-      "learning_rate": 0.00019491329479768784,
-      "loss": 0.3297,
+      "learning_rate": 0.00019508670520231212,
+      "loss": 0.3518,
       "step": 1109
     },
     {
       "epoch": 2.49,
-      "learning_rate": 0.00019473988439306356,
-      "loss": 0.2509,
+      "learning_rate": 0.00019491329479768784,
+      "loss": 0.2584,
       "step": 1110
     },
     {
       "epoch": 2.49,
-      "learning_rate": 0.00019456647398843928,
-      "loss": 0.2808,
+      "learning_rate": 0.00019473988439306356,
+      "loss": 0.2994,
       "step": 1111
     },
     {
       "epoch": 2.49,
-      "learning_rate": 0.00019439306358381503,
-      "loss": 0.3131,
+      "learning_rate": 0.00019456647398843928,
+      "loss": 0.3466,
       "step": 1112
     },
     {
       "epoch": 2.5,
-      "learning_rate": 0.00019421965317919073,
-      "loss": 0.301,
+      "learning_rate": 0.00019439306358381503,
+      "loss": 0.3337,
       "step": 1113
     },
     {
       "epoch": 2.5,
-      "learning_rate": 0.00019404624277456645,
-      "loss": 0.2962,
+      "learning_rate": 0.00019421965317919073,
+      "loss": 0.3306,
       "step": 1114
     },
     {
       "epoch": 2.5,
-      "learning_rate": 0.00019387283236994217,
-      "loss": 0.2827,
+      "learning_rate": 0.00019404624277456645,
+      "loss": 0.2953,
       "step": 1115
     },
     {
       "epoch": 2.5,
-      "learning_rate": 0.00019369942196531792,
-      "loss": 0.2973,
+      "learning_rate": 0.00019387283236994217,
+      "loss": 0.3109,
       "step": 1116
     },
     {
       "epoch": 2.5,
-      "learning_rate": 0.00019352601156069364,
-      "loss": 0.2817,
+      "learning_rate": 0.00019369942196531792,
+      "loss": 0.3141,
       "step": 1117
     },
     {
       "epoch": 2.51,
-      "learning_rate": 0.00019335260115606933,
-      "loss": 0.2612,
+      "learning_rate": 0.00019352601156069364,
+      "loss": 0.2826,
       "step": 1118
     },
     {
       "epoch": 2.51,
-      "learning_rate": 0.00019317919075144505,
-      "loss": 0.253,
+      "learning_rate": 0.00019335260115606933,
+      "loss": 0.2811,
       "step": 1119
     },
     {
       "epoch": 2.51,
-      "learning_rate": 0.0001930057803468208,
-      "loss": 0.2817,
+      "learning_rate": 0.00019317919075144505,
+      "loss": 0.308,
       "step": 1120
     },
     {
       "epoch": 2.51,
-      "learning_rate": 0.00019283236994219652,
-      "loss": 0.2823,
+      "learning_rate": 0.0001930057803468208,
+      "loss": 0.3758,
       "step": 1121
     },
     {
       "epoch": 2.52,
-      "learning_rate": 0.00019265895953757224,
-      "loss": 0.3166,
+      "learning_rate": 0.00019283236994219652,
+      "loss": 0.3172,
       "step": 1122
     },
     {
       "epoch": 2.52,
-      "learning_rate": 0.00019248554913294796,
-      "loss": 0.3035,
+      "learning_rate": 0.00019265895953757224,
+      "loss": 0.3789,
       "step": 1123
     },
     {
       "epoch": 2.52,
-      "learning_rate": 0.0001923121387283237,
-      "loss": 0.2608,
+      "learning_rate": 0.00019248554913294796,
+      "loss": 0.326,
       "step": 1124
     },
     {
       "epoch": 2.52,
-      "learning_rate": 0.0001921387283236994,
-      "loss": 0.3112,
+      "learning_rate": 0.0001923121387283237,
+      "loss": 0.3814,
       "step": 1125
     },
     {
       "epoch": 2.52,
-      "learning_rate": 0.00019196531791907512,
-      "loss": 0.3573,
+      "learning_rate": 0.0001921387283236994,
+      "loss": 0.3633,
       "step": 1126
     },
     {
       "epoch": 2.53,
-      "learning_rate": 0.00019179190751445084,
-      "loss": 0.2999,
+      "learning_rate": 0.00019196531791907512,
+      "loss": 0.33,
       "step": 1127
     },
     {
       "epoch": 2.53,
-      "learning_rate": 0.0001916184971098266,
-      "loss": 0.269,
+      "learning_rate": 0.00019179190751445084,
+      "loss": 0.3174,
       "step": 1128
     },
     {
       "epoch": 2.53,
-      "learning_rate": 0.0001914450867052023,
-      "loss": 0.2425,
+      "learning_rate": 0.0001916184971098266,
+      "loss": 0.3249,
       "step": 1129
     },
     {
       "epoch": 2.53,
-      "learning_rate": 0.000191271676300578,
-      "loss": 0.243,
+      "learning_rate": 0.0001914450867052023,
+      "loss": 0.2607,
       "step": 1130
     },
     {
       "epoch": 2.54,
-      "learning_rate": 0.00019109826589595373,
-      "loss": 0.3348,
+      "learning_rate": 0.000191271676300578,
+      "loss": 0.405,
       "step": 1131
     },
     {
       "epoch": 2.54,
-      "learning_rate": 0.00019092485549132947,
-      "loss": 0.2862,
+      "learning_rate": 0.00019109826589595373,
+      "loss": 0.3608,
       "step": 1132
     },
     {
       "epoch": 2.54,
-      "learning_rate": 0.0001907514450867052,
-      "loss": 0.2596,
+      "learning_rate": 0.00019092485549132947,
+      "loss": 0.3364,
       "step": 1133
     },
     {
       "epoch": 2.54,
-      "learning_rate": 0.00019057803468208091,
-      "loss": 0.2915,
+      "learning_rate": 0.0001907514450867052,
+      "loss": 0.3467,
       "step": 1134
     },
     {
       "epoch": 2.54,
-      "learning_rate": 0.0001904046242774566,
-      "loss": 0.269,
+      "learning_rate": 0.00019057803468208091,
+      "loss": 0.2971,
       "step": 1135
     },
     {
       "epoch": 2.55,
-      "learning_rate": 0.00019023121387283236,
-      "loss": 0.2449,
+      "learning_rate": 0.0001904046242774566,
+      "loss": 0.2842,
       "step": 1136
     },
     {
       "epoch": 2.55,
-      "learning_rate": 0.00019005780346820808,
-      "loss": 0.318,
+      "learning_rate": 0.00019023121387283236,
+      "loss": 0.3429,
       "step": 1137
     },
     {
       "epoch": 2.55,
-      "learning_rate": 0.0001898843930635838,
-      "loss": 0.2883,
+      "learning_rate": 0.00019005780346820808,
+      "loss": 0.3082,
       "step": 1138
     },
     {
       "epoch": 2.55,
-      "learning_rate": 0.00018971098265895952,
-      "loss": 0.3835,
+      "learning_rate": 0.0001898843930635838,
+      "loss": 0.4151,
       "step": 1139
     },
     {
       "epoch": 2.56,
-      "learning_rate": 0.00018953757225433527,
-      "loss": 0.3351,
+      "learning_rate": 0.00018971098265895952,
+      "loss": 0.3784,
       "step": 1140
     },
     {
       "epoch": 2.56,
-      "learning_rate": 0.00018936416184971096,
-      "loss": 0.2019,
+      "learning_rate": 0.00018953757225433527,
+      "loss": 0.2453,
       "step": 1141
     },
     {
       "epoch": 2.56,
-      "learning_rate": 0.00018919075144508668,
-      "loss": 0.4093,
+      "learning_rate": 0.00018936416184971096,
+      "loss": 0.4695,
       "step": 1142
     },
     {
       "epoch": 2.56,
-      "learning_rate": 0.0001890173410404624,
-      "loss": 0.5879,
+      "learning_rate": 0.00018919075144508668,
+      "loss": 0.7567,
       "step": 1143
     },
     {
       "epoch": 2.57,
-      "learning_rate": 0.00018884393063583815,
-      "loss": 0.4834,
+      "learning_rate": 0.0001890173410404624,
+      "loss": 0.6189,
       "step": 1144
     },
     {
       "epoch": 2.57,
-      "learning_rate": 0.00018867052023121387,
-      "loss": 0.4007,
+      "learning_rate": 0.00018884393063583815,
+      "loss": 0.504,
       "step": 1145
     },
     {
       "epoch": 2.57,
-      "learning_rate": 0.0001884971098265896,
-      "loss": 0.3683,
+      "learning_rate": 0.00018867052023121387,
+      "loss": 0.4176,
       "step": 1146
     },
     {
       "epoch": 2.57,
-      "learning_rate": 0.00018832369942196528,
-      "loss": 0.3969,
+      "learning_rate": 0.0001884971098265896,
+      "loss": 0.4394,
       "step": 1147
     },
     {
       "epoch": 2.57,
-      "learning_rate": 0.00018815028901734103,
-      "loss": 0.386,
+      "learning_rate": 0.00018832369942196528,
+      "loss": 0.3826,
       "step": 1148
     },
     {
       "epoch": 2.58,
-      "learning_rate": 0.00018797687861271675,
-      "loss": 0.3589,
+      "learning_rate": 0.00018815028901734103,
+      "loss": 0.4139,
       "step": 1149
     },
     {
       "epoch": 2.58,
-      "learning_rate": 0.00018780346820809247,
-      "loss": 0.3318,
+      "learning_rate": 0.00018797687861271675,
+      "loss": 0.4348,
       "step": 1150
     },
     {
       "epoch": 2.58,
-      "learning_rate": 0.0001876300578034682,
-      "loss": 0.3108,
+      "learning_rate": 0.00018780346820809247,
+      "loss": 0.4796,
       "step": 1151
     },
     {
       "epoch": 2.58,
-      "learning_rate": 0.00018745664739884394,
-      "loss": 0.368,
+      "learning_rate": 0.0001876300578034682,
+      "loss": 0.4092,
       "step": 1152
     },
     {
       "epoch": 2.59,
-      "learning_rate": 0.00018728323699421963,
-      "loss": 0.2717,
+      "learning_rate": 0.00018745664739884394,
+      "loss": 0.3293,
       "step": 1153
     },
     {
       "epoch": 2.59,
-      "learning_rate": 0.00018710982658959536,
-      "loss": 0.3246,
+      "learning_rate": 0.00018728323699421963,
+      "loss": 0.3758,
       "step": 1154
     },
     {
       "epoch": 2.59,
-      "learning_rate": 0.00018693641618497108,
-      "loss": 0.3484,
+      "learning_rate": 0.00018710982658959536,
+      "loss": 0.3988,
       "step": 1155
     },
     {
       "epoch": 2.59,
-      "learning_rate": 0.00018676300578034682,
-      "loss": 0.294,
+      "learning_rate": 0.00018693641618497108,
+      "loss": 0.3302,
       "step": 1156
     },
     {
       "epoch": 2.59,
-      "learning_rate": 0.00018658959537572254,
-      "loss": 0.3197,
+      "learning_rate": 0.00018676300578034682,
+      "loss": 0.4761,
       "step": 1157
     },
     {
       "epoch": 2.6,
-      "learning_rate": 0.00018641618497109824,
-      "loss": 0.2659,
+      "learning_rate": 0.00018658959537572254,
+      "loss": 0.4218,
       "step": 1158
     },
     {
       "epoch": 2.6,
-      "learning_rate": 0.00018624277456647396,
-      "loss": 0.2214,
+      "learning_rate": 0.00018641618497109824,
+      "loss": 0.2673,
       "step": 1159
     },
     {
       "epoch": 2.6,
-      "learning_rate": 0.0001860693641618497,
-      "loss": 0.2667,
+      "learning_rate": 0.00018624277456647396,
+      "loss": 0.3068,
       "step": 1160
     },
     {
       "epoch": 2.6,
-      "learning_rate": 0.00018589595375722543,
-      "loss": 0.2771,
+      "learning_rate": 0.0001860693641618497,
+      "loss": 0.3173,
       "step": 1161
     },
     {
       "epoch": 2.61,
-      "learning_rate": 0.00018572254335260115,
-      "loss": 0.2613,
+      "learning_rate": 0.00018589595375722543,
+      "loss": 0.2941,
       "step": 1162
     },
     {
       "epoch": 2.61,
-      "learning_rate": 0.00018554913294797684,
-      "loss": 0.2699,
+      "learning_rate": 0.00018572254335260115,
+      "loss": 0.3206,
       "step": 1163
     },
     {
       "epoch": 2.61,
-      "learning_rate": 0.0001853757225433526,
-      "loss": 0.2633,
+      "learning_rate": 0.00018554913294797684,
+      "loss": 0.4003,
       "step": 1164
     },
     {
       "epoch": 2.61,
-      "learning_rate": 0.0001852023121387283,
-      "loss": 0.2831,
+      "learning_rate": 0.0001853757225433526,
+      "loss": 0.302,
       "step": 1165
     },
     {
       "epoch": 2.61,
-      "learning_rate": 0.00018502890173410403,
-      "loss": 0.286,
+      "learning_rate": 0.0001852023121387283,
+      "loss": 0.3248,
       "step": 1166
     },
     {
       "epoch": 2.62,
-      "learning_rate": 0.00018485549132947975,
-      "loss": 0.2913,
+      "learning_rate": 0.00018502890173410403,
+      "loss": 0.334,
       "step": 1167
     },
     {
       "epoch": 2.62,
-      "learning_rate": 0.0001846820809248555,
-      "loss": 0.2439,
+      "learning_rate": 0.00018485549132947975,
+      "loss": 0.2525,
       "step": 1168
     },
     {
       "epoch": 2.62,
-      "learning_rate": 0.0001845086705202312,
-      "loss": 0.2759,
+      "learning_rate": 0.0001846820809248555,
+      "loss": 0.3136,
       "step": 1169
     },
     {
       "epoch": 2.62,
-      "learning_rate": 0.0001843352601156069,
-      "loss": 0.2754,
+      "learning_rate": 0.0001845086705202312,
+      "loss": 0.2709,
       "step": 1170
     },
     {
       "epoch": 2.63,
-      "learning_rate": 0.00018416184971098263,
-      "loss": 0.2704,
+      "learning_rate": 0.0001843352601156069,
+      "loss": 0.2941,
       "step": 1171
     },
     {
       "epoch": 2.63,
-      "learning_rate": 0.00018398843930635838,
-      "loss": 0.2555,
+      "learning_rate": 0.00018416184971098263,
+      "loss": 0.301,
       "step": 1172
     },
     {
       "epoch": 2.63,
-      "learning_rate": 0.0001838150289017341,
-      "loss": 0.2783,
+      "learning_rate": 0.00018398843930635838,
+      "loss": 0.3081,
       "step": 1173
     },
     {
       "epoch": 2.63,
-      "learning_rate": 0.00018364161849710982,
-      "loss": 0.234,
+      "learning_rate": 0.0001838150289017341,
+      "loss": 0.2646,
       "step": 1174
     },
     {
       "epoch": 2.63,
-      "learning_rate": 0.00018346820809248552,
-      "loss": 0.3144,
+      "learning_rate": 0.00018364161849710982,
+      "loss": 0.3387,
       "step": 1175
     },
     {
       "epoch": 2.64,
-      "learning_rate": 0.00018329479768786124,
-      "loss": 0.2811,
+      "learning_rate": 0.00018346820809248552,
+      "loss": 0.3115,
       "step": 1176
     },
     {
       "epoch": 2.64,
-      "learning_rate": 0.00018312138728323698,
-      "loss": 0.2591,
+      "learning_rate": 0.00018329479768786124,
+      "loss": 0.3095,
       "step": 1177
     },
     {
       "epoch": 2.64,
-      "learning_rate": 0.0001829479768786127,
-      "loss": 0.2926,
+      "learning_rate": 0.00018312138728323698,
+      "loss": 0.2983,
       "step": 1178
     },
     {
       "epoch": 2.64,
-      "learning_rate": 0.00018277456647398843,
-      "loss": 0.3121,
+      "learning_rate": 0.0001829479768786127,
+      "loss": 0.3246,
       "step": 1179
     },
     {
       "epoch": 2.65,
-      "learning_rate": 0.00018260115606936412,
-      "loss": 0.2776,
+      "learning_rate": 0.00018277456647398843,
+      "loss": 0.3024,
       "step": 1180
     },
     {
       "epoch": 2.65,
-      "learning_rate": 0.00018242774566473987,
-      "loss": 0.2272,
+      "learning_rate": 0.00018260115606936412,
+      "loss": 0.2837,
       "step": 1181
     },
     {
       "epoch": 2.65,
-      "learning_rate": 0.0001822543352601156,
-      "loss": 0.1928,
+      "learning_rate": 0.00018242774566473987,
+      "loss": 0.2031,
       "step": 1182
     },
     {
       "epoch": 2.65,
-      "learning_rate": 0.0001820809248554913,
-      "loss": 0.2614,
+      "learning_rate": 0.0001822543352601156,
+      "loss": 0.2731,
       "step": 1183
     },
     {
       "epoch": 2.65,
-      "learning_rate": 0.00018190751445086703,
-      "loss": 0.2515,
+      "learning_rate": 0.0001820809248554913,
+      "loss": 0.4192,
       "step": 1184
     },
     {
       "epoch": 2.66,
-      "learning_rate": 0.00018173410404624278,
-      "loss": 0.2994,
+      "learning_rate": 0.00018190751445086703,
+      "loss": 0.4043,
       "step": 1185
     },
     {
       "epoch": 2.66,
-      "learning_rate": 0.00018156069364161847,
-      "loss": 0.3348,
+      "learning_rate": 0.00018173410404624278,
+      "loss": 0.3666,
       "step": 1186
     },
     {
       "epoch": 2.66,
-      "learning_rate": 0.0001813872832369942,
-      "loss": 0.2476,
+      "learning_rate": 0.00018156069364161847,
+      "loss": 0.3159,
       "step": 1187
     },
     {
       "epoch": 2.66,
-      "learning_rate": 0.0001812138728323699,
-      "loss": 0.2636,
+      "learning_rate": 0.0001813872832369942,
+      "loss": 0.2942,
       "step": 1188
     },
     {
       "epoch": 2.67,
-      "learning_rate": 0.00018104046242774566,
-      "loss": 0.2217,
+      "learning_rate": 0.0001812138728323699,
+      "loss": 0.2372,
       "step": 1189
     },
     {
       "epoch": 2.67,
-      "learning_rate": 0.00018086705202312138,
-      "loss": 0.3463,
+      "learning_rate": 0.00018104046242774566,
+      "loss": 0.3847,
       "step": 1190
     },
     {
       "epoch": 2.67,
-      "learning_rate": 0.00018069364161849707,
-      "loss": 0.24,
+      "learning_rate": 0.00018086705202312138,
+      "loss": 0.3073,
       "step": 1191
     },
     {
       "epoch": 2.67,
-      "learning_rate": 0.0001805202312138728,
-      "loss": 0.3035,
+      "learning_rate": 0.00018069364161849707,
+      "loss": 0.3426,
       "step": 1192
     },
     {
       "epoch": 2.67,
-      "learning_rate": 0.00018034682080924854,
-      "loss": 0.7377,
+      "learning_rate": 0.0001805202312138728,
+      "loss": 0.8319,
       "step": 1193
     },
     {
       "epoch": 2.68,
-      "learning_rate": 0.00018017341040462426,
-      "loss": 0.5811,
+      "learning_rate": 0.00018034682080924854,
+      "loss": 0.6785,
       "step": 1194
     },
     {
       "epoch": 2.68,
-      "learning_rate": 0.00017999999999999998,
-      "loss": 0.4344,
+      "learning_rate": 0.00018017341040462426,
+      "loss": 0.5124,
       "step": 1195
     },
     {
       "epoch": 2.68,
-      "learning_rate": 0.0001798265895953757,
-      "loss": 0.4348,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.4609,
       "step": 1196
     },
     {
       "epoch": 2.68,
-      "learning_rate": 0.00017965317919075145,
-      "loss": 0.3972,
+      "learning_rate": 0.0001798265895953757,
+      "loss": 0.4206,
       "step": 1197
     },
     {
       "epoch": 2.69,
-      "learning_rate": 0.00017947976878612715,
-      "loss": 0.4097,
+      "learning_rate": 0.00017965317919075145,
+      "loss": 0.4402,
       "step": 1198
     },
     {
       "epoch": 2.69,
-      "learning_rate": 0.00017930635838150287,
-      "loss": 0.386,
+      "learning_rate": 0.00017947976878612715,
+      "loss": 0.3913,
       "step": 1199
     },
     {
       "epoch": 2.69,
-      "learning_rate": 0.0001791329479768786,
-      "loss": 0.3258,
+      "learning_rate": 0.00017930635838150287,
+      "loss": 0.3591,
       "step": 1200
     },
     {
       "epoch": 2.69,
-      "learning_rate": 0.00017895953757225434,
-      "loss": 0.3626,
+      "learning_rate": 0.0001791329479768786,
+      "loss": 0.3962,
       "step": 1201
     },
     {
       "epoch": 2.7,
-      "learning_rate": 0.00017878612716763006,
-      "loss": 0.3213,
+      "learning_rate": 0.00017895953757225434,
+      "loss": 0.3368,
       "step": 1202
     },
     {
       "epoch": 2.7,
-      "learning_rate": 0.00017861271676300575,
-      "loss": 0.2808,
+      "learning_rate": 0.00017878612716763006,
+      "loss": 0.339,
       "step": 1203
     },
     {
       "epoch": 2.7,
-      "learning_rate": 0.00017843930635838147,
-      "loss": 0.2454,
+      "learning_rate": 0.00017861271676300575,
+      "loss": 0.2718,
       "step": 1204
     },
     {
       "epoch": 2.7,
-      "learning_rate": 0.00017826589595375722,
-      "loss": 0.2311,
+      "learning_rate": 0.00017843930635838147,
+      "loss": 0.2501,
       "step": 1205
     },
     {
       "epoch": 2.7,
-      "learning_rate": 0.00017809248554913294,
-      "loss": 0.2546,
+      "learning_rate": 0.00017826589595375722,
+      "loss": 0.3284,
       "step": 1206
     },
     {
       "epoch": 2.71,
-      "learning_rate": 0.00017791907514450866,
-      "loss": 0.2878,
+      "learning_rate": 0.00017809248554913294,
+      "loss": 0.2951,
       "step": 1207
     },
     {
       "epoch": 2.71,
-      "learning_rate": 0.00017774566473988435,
-      "loss": 0.2705,
+      "learning_rate": 0.00017791907514450866,
+      "loss": 0.2785,
       "step": 1208
     },
     {
       "epoch": 2.71,
-      "learning_rate": 0.0001775722543352601,
-      "loss": 0.3434,
+      "learning_rate": 0.00017774566473988435,
+      "loss": 0.3452,
       "step": 1209
     },
     {
       "epoch": 2.71,
-      "learning_rate": 0.00017739884393063582,
-      "loss": 0.2801,
+      "learning_rate": 0.0001775722543352601,
+      "loss": 0.3002,
       "step": 1210
     },
     {
       "epoch": 2.72,
-      "learning_rate": 0.00017722543352601154,
-      "loss": 0.3088,
+      "learning_rate": 0.00017739884393063582,
+      "loss": 0.3082,
       "step": 1211
     },
     {
       "epoch": 2.72,
-      "learning_rate": 0.00017705202312138726,
-      "loss": 0.2821,
+      "learning_rate": 0.00017722543352601154,
+      "loss": 0.2915,
       "step": 1212
     },
     {
       "epoch": 2.72,
-      "learning_rate": 0.000176878612716763,
-      "loss": 0.3169,
+      "learning_rate": 0.00017705202312138726,
+      "loss": 0.3227,
       "step": 1213
     },
     {
       "epoch": 2.72,
-      "learning_rate": 0.0001767052023121387,
-      "loss": 0.2789,
+      "learning_rate": 0.000176878612716763,
+      "loss": 0.2923,
       "step": 1214
     },
     {
       "epoch": 2.72,
-      "learning_rate": 0.00017653179190751442,
-      "loss": 0.1741,
+      "learning_rate": 0.0001767052023121387,
+      "loss": 0.2174,
       "step": 1215
     },
     {
       "epoch": 2.73,
-      "learning_rate": 0.00017635838150289015,
-      "loss": 0.2652,
+      "learning_rate": 0.00017653179190751442,
+      "loss": 0.3016,
       "step": 1216
     },
     {
       "epoch": 2.73,
-      "learning_rate": 0.0001761849710982659,
-      "loss": 0.2357,
+      "learning_rate": 0.00017635838150289015,
+      "loss": 0.2278,
       "step": 1217
     },
     {
       "epoch": 2.73,
-      "learning_rate": 0.00017601156069364161,
-      "loss": 0.2809,
+      "learning_rate": 0.0001761849710982659,
+      "loss": 0.2899,
       "step": 1218
     },
     {
       "epoch": 2.73,
-      "learning_rate": 0.0001758381502890173,
-      "loss": 0.2503,
+      "learning_rate": 0.00017601156069364161,
+      "loss": 0.2595,
       "step": 1219
     },
     {
       "epoch": 2.74,
-      "learning_rate": 0.00017566473988439303,
-      "loss": 0.2239,
+      "learning_rate": 0.0001758381502890173,
+      "loss": 0.2312,
       "step": 1220
     },
     {
       "epoch": 2.74,
-      "learning_rate": 0.00017549132947976878,
-      "loss": 0.2508,
+      "learning_rate": 0.00017566473988439303,
+      "loss": 0.2796,
       "step": 1221
     },
     {
       "epoch": 2.74,
-      "learning_rate": 0.0001753179190751445,
-      "loss": 0.2548,
+      "learning_rate": 0.00017549132947976878,
+      "loss": 0.2577,
       "step": 1222
     },
     {
       "epoch": 2.74,
-      "learning_rate": 0.00017514450867052022,
-      "loss": 0.3294,
+      "learning_rate": 0.0001753179190751445,
+      "loss": 0.3298,
       "step": 1223
     },
     {
       "epoch": 2.74,
-      "learning_rate": 0.00017497109826589594,
-      "loss": 0.272,
+      "learning_rate": 0.00017514450867052022,
+      "loss": 0.2904,
       "step": 1224
     },
     {
       "epoch": 2.75,
-      "learning_rate": 0.00017479768786127169,
-      "loss": 0.229,
+      "learning_rate": 0.00017497109826589594,
+      "loss": 0.2401,
       "step": 1225
     },
     {
       "epoch": 2.75,
-      "learning_rate": 0.00017462427745664738,
-      "loss": 0.2473,
+      "learning_rate": 0.00017479768786127169,
+      "loss": 0.273,
       "step": 1226
     },
     {
       "epoch": 2.75,
-      "learning_rate": 0.0001744508670520231,
-      "loss": 0.2432,
+      "learning_rate": 0.00017462427745664738,
+      "loss": 0.2479,
       "step": 1227
     },
     {
       "epoch": 2.75,
-      "learning_rate": 0.00017427745664739882,
-      "loss": 0.3006,
+      "learning_rate": 0.0001744508670520231,
+      "loss": 0.325,
       "step": 1228
     },
     {
       "epoch": 2.76,
-      "learning_rate": 0.00017410404624277457,
-      "loss": 0.2579,
+      "learning_rate": 0.00017427745664739882,
+      "loss": 0.2803,
       "step": 1229
     },
     {
       "epoch": 2.76,
-      "learning_rate": 0.0001739306358381503,
-      "loss": 0.2355,
+      "learning_rate": 0.00017410404624277457,
+      "loss": 0.2399,
       "step": 1230
     },
     {
       "epoch": 2.76,
-      "learning_rate": 0.00017375722543352598,
-      "loss": 0.2419,
+      "learning_rate": 0.0001739306358381503,
+      "loss": 0.2507,
       "step": 1231
     },
     {
       "epoch": 2.76,
-      "learning_rate": 0.0001735838150289017,
-      "loss": 0.2779,
+      "learning_rate": 0.00017375722543352598,
+      "loss": 0.2888,
       "step": 1232
     },
     {
       "epoch": 2.76,
-      "learning_rate": 0.00017341040462427745,
-      "loss": 0.2763,
+      "learning_rate": 0.0001735838150289017,
+      "loss": 0.2838,
       "step": 1233
     },
     {
       "epoch": 2.77,
-      "learning_rate": 0.00017323699421965317,
-      "loss": 0.2985,
+      "learning_rate": 0.00017341040462427745,
+      "loss": 0.2881,
       "step": 1234
     },
     {
       "epoch": 2.77,
-      "learning_rate": 0.0001730635838150289,
-      "loss": 0.2795,
+      "learning_rate": 0.00017323699421965317,
+      "loss": 0.2587,
       "step": 1235
     },
     {
       "epoch": 2.77,
-      "learning_rate": 0.00017289017341040459,
-      "loss": 0.2494,
+      "learning_rate": 0.0001730635838150289,
+      "loss": 0.2566,
       "step": 1236
     },
     {
       "epoch": 2.77,
-      "learning_rate": 0.00017271676300578033,
-      "loss": 0.3492,
+      "learning_rate": 0.00017289017341040459,
+      "loss": 0.3898,
       "step": 1237
     },
     {
       "epoch": 2.78,
-      "learning_rate": 0.00017254335260115605,
-      "loss": 0.2718,
+      "learning_rate": 0.00017271676300578033,
+      "loss": 0.2496,
       "step": 1238
     },
     {
       "epoch": 2.78,
-      "learning_rate": 0.00017236994219653178,
-      "loss": 0.211,
+      "learning_rate": 0.00017254335260115605,
+      "loss": 0.2205,
       "step": 1239
     },
     {
       "epoch": 2.78,
-      "learning_rate": 0.0001721965317919075,
-      "loss": 0.2588,
+      "learning_rate": 0.00017236994219653178,
+      "loss": 0.2732,
       "step": 1240
     },
     {
       "epoch": 2.78,
-      "learning_rate": 0.00017202312138728324,
-      "loss": 0.2744,
+      "learning_rate": 0.0001721965317919075,
+      "loss": 0.2794,
       "step": 1241
     },
     {
       "epoch": 2.78,
-      "learning_rate": 0.00017184971098265894,
-      "loss": 0.3484,
+      "learning_rate": 0.00017202312138728324,
+      "loss": 0.3402,
       "step": 1242
     },
     {
       "epoch": 2.79,
-      "learning_rate": 0.00017167630057803466,
-      "loss": 0.4737,
+      "learning_rate": 0.00017184971098265894,
+      "loss": 0.5904,
       "step": 1243
     },
     {
       "epoch": 2.79,
-      "learning_rate": 0.00017150289017341038,
-      "loss": 0.4863,
+      "learning_rate": 0.00017167630057803466,
+      "loss": 0.5435,
       "step": 1244
     },
     {
       "epoch": 2.79,
-      "learning_rate": 0.00017132947976878613,
-      "loss": 0.3562,
+      "learning_rate": 0.00017150289017341038,
+      "loss": 0.4598,
       "step": 1245
     },
     {
       "epoch": 2.79,
-      "learning_rate": 0.00017115606936416185,
-      "loss": 0.3472,
+      "learning_rate": 0.00017132947976878613,
+      "loss": 0.4237,
       "step": 1246
     },
     {
       "epoch": 2.8,
-      "learning_rate": 0.00017098265895953757,
-      "loss": 0.3903,
+      "learning_rate": 0.00017115606936416185,
+      "loss": 0.4052,
       "step": 1247
     },
     {
       "epoch": 2.8,
-      "learning_rate": 0.00017080924855491326,
-      "loss": 0.3271,
+      "learning_rate": 0.00017098265895953757,
+      "loss": 0.3596,
       "step": 1248
     },
     {
       "epoch": 2.8,
-      "learning_rate": 0.000170635838150289,
-      "loss": 0.3256,
+      "learning_rate": 0.00017080924855491326,
+      "loss": 0.3375,
       "step": 1249
     },
     {
       "epoch": 2.8,
-      "learning_rate": 0.00017046242774566473,
-      "loss": 0.3254,
+      "learning_rate": 0.000170635838150289,
+      "loss": 0.3509,
       "step": 1250
     },
     {
       "epoch": 2.8,
-      "learning_rate": 0.00017028901734104045,
-      "loss": 0.3303,
+      "learning_rate": 0.00017046242774566473,
+      "loss": 0.3337,
       "step": 1251
     },
     {
       "epoch": 2.81,
-      "learning_rate": 0.00017011560693641617,
-      "loss": 0.3209,
+      "learning_rate": 0.00017028901734104045,
+      "loss": 0.4023,
       "step": 1252
     },
     {
       "epoch": 2.81,
-      "learning_rate": 0.00016994219653179192,
-      "loss": 0.27,
+      "learning_rate": 0.00017011560693641617,
+      "loss": 0.2961,
       "step": 1253
     },
     {
       "epoch": 2.81,
-      "learning_rate": 0.0001697687861271676,
-      "loss": 0.2704,
+      "learning_rate": 0.00016994219653179192,
+      "loss": 0.2642,
       "step": 1254
     },
     {
       "epoch": 2.81,
-      "learning_rate": 0.00016959537572254333,
-      "loss": 0.3389,
+      "learning_rate": 0.0001697687861271676,
+      "loss": 0.3523,
       "step": 1255
     },
     {
       "epoch": 2.82,
-      "learning_rate": 0.00016942196531791905,
-      "loss": 0.254,
+      "learning_rate": 0.00016959537572254333,
+      "loss": 0.2547,
       "step": 1256
     },
     {
       "epoch": 2.82,
-      "learning_rate": 0.0001692485549132948,
-      "loss": 0.2372,
-      "step": 1257
+      "learning_rate": 0.00016942196531791905,
+      "loss": 0.2283,
+      "step": 1257
     },
     {
       "epoch": 2.82,
-      "learning_rate": 0.00016907514450867052,
-      "loss": 0.2977,
+      "learning_rate": 0.0001692485549132948,
+      "loss": 0.2924,
       "step": 1258
     },
     {
       "epoch": 2.82,
-      "learning_rate": 0.00016890173410404622,
-      "loss": 0.3098,
+      "learning_rate": 0.00016907514450867052,
+      "loss": 0.3059,
       "step": 1259
     },
     {
       "epoch": 2.83,
-      "learning_rate": 0.00016872832369942194,
-      "loss": 0.2283,
+      "learning_rate": 0.00016890173410404622,
+      "loss": 0.2438,
       "step": 1260
     },
     {
       "epoch": 2.83,
-      "learning_rate": 0.00016855491329479768,
-      "loss": 0.2234,
+      "learning_rate": 0.00016872832369942194,
+      "loss": 0.2312,
       "step": 1261
     },
     {
       "epoch": 2.83,
-      "learning_rate": 0.0001683815028901734,
-      "loss": 0.2372,
+      "learning_rate": 0.00016855491329479768,
+      "loss": 0.2383,
       "step": 1262
     },
     {
       "epoch": 2.83,
-      "learning_rate": 0.00016820809248554913,
-      "loss": 0.2552,
+      "learning_rate": 0.0001683815028901734,
+      "loss": 0.2653,
       "step": 1263
     },
     {
       "epoch": 2.83,
-      "learning_rate": 0.00016803468208092482,
-      "loss": 0.2208,
+      "learning_rate": 0.00016820809248554913,
+      "loss": 0.2651,
       "step": 1264
     },
     {
       "epoch": 2.84,
-      "learning_rate": 0.00016786127167630057,
-      "loss": 0.2961,
+      "learning_rate": 0.00016803468208092482,
+      "loss": 0.3351,
       "step": 1265
     },
     {
       "epoch": 2.84,
-      "learning_rate": 0.0001676878612716763,
-      "loss": 0.2374,
+      "learning_rate": 0.00016786127167630057,
+      "loss": 0.2531,
       "step": 1266
     },
     {
       "epoch": 2.84,
-      "learning_rate": 0.000167514450867052,
-      "loss": 0.2974,
+      "learning_rate": 0.0001676878612716763,
+      "loss": 0.3217,
       "step": 1267
     },
     {
       "epoch": 2.84,
-      "learning_rate": 0.00016734104046242773,
-      "loss": 0.2306,
+      "learning_rate": 0.000167514450867052,
+      "loss": 0.2335,
       "step": 1268
     },
     {
       "epoch": 2.85,
-      "learning_rate": 0.00016716763005780348,
-      "loss": 0.2881,
+      "learning_rate": 0.00016734104046242773,
+      "loss": 0.2898,
       "step": 1269
     },
     {
       "epoch": 2.85,
-      "learning_rate": 0.0001669942196531792,
-      "loss": 0.1941,
+      "learning_rate": 0.00016716763005780348,
+      "loss": 0.2112,
       "step": 1270
     },
     {
       "epoch": 2.85,
-      "learning_rate": 0.0001668208092485549,
-      "loss": 0.2276,
+      "learning_rate": 0.0001669942196531792,
+      "loss": 0.2423,
       "step": 1271
     },
     {
       "epoch": 2.85,
-      "learning_rate": 0.0001666473988439306,
-      "loss": 0.2499,
+      "learning_rate": 0.0001668208092485549,
+      "loss": 0.2758,
       "step": 1272
     },
     {
       "epoch": 2.85,
-      "learning_rate": 0.00016647398843930633,
-      "loss": 0.2944,
+      "learning_rate": 0.0001666473988439306,
+      "loss": 0.2943,
       "step": 1273
     },
     {
       "epoch": 2.86,
-      "learning_rate": 0.00016630057803468208,
-      "loss": 0.2228,
+      "learning_rate": 0.00016647398843930633,
+      "loss": 0.2411,
       "step": 1274
     },
     {
       "epoch": 2.86,
-      "learning_rate": 0.0001661271676300578,
-      "loss": 0.2056,
+      "learning_rate": 0.00016630057803468208,
+      "loss": 0.2418,
       "step": 1275
     },
     {
       "epoch": 2.86,
-      "learning_rate": 0.0001659537572254335,
-      "loss": 0.2833,
+      "learning_rate": 0.0001661271676300578,
+      "loss": 0.2899,
       "step": 1276
     },
     {
       "epoch": 2.86,
-      "learning_rate": 0.00016578034682080922,
-      "loss": 0.2251,
+      "learning_rate": 0.0001659537572254335,
+      "loss": 0.2431,
       "step": 1277
     },
     {
       "epoch": 2.87,
-      "learning_rate": 0.00016560693641618496,
-      "loss": 0.2508,
+      "learning_rate": 0.00016578034682080922,
+      "loss": 0.2374,
       "step": 1278
     },
     {
       "epoch": 2.87,
-      "learning_rate": 0.00016543352601156068,
-      "loss": 0.2835,
+      "learning_rate": 0.00016560693641618496,
+      "loss": 0.3077,
       "step": 1279
     },
     {
       "epoch": 2.87,
-      "learning_rate": 0.0001652601156069364,
-      "loss": 0.2011,
+      "learning_rate": 0.00016543352601156068,
+      "loss": 0.1856,
       "step": 1280
     },
     {
       "epoch": 2.87,
-      "learning_rate": 0.0001650867052023121,
-      "loss": 0.2674,
+      "learning_rate": 0.0001652601156069364,
+      "loss": 0.2465,
       "step": 1281
     },
     {
       "epoch": 2.87,
-      "learning_rate": 0.00016491329479768785,
-      "loss": 0.2861,
+      "learning_rate": 0.0001650867052023121,
+      "loss": 0.301,
       "step": 1282
     },
     {
       "epoch": 2.88,
-      "learning_rate": 0.00016473988439306357,
-      "loss": 0.2339,
+      "learning_rate": 0.00016491329479768785,
+      "loss": 0.2545,
       "step": 1283
     },
     {
       "epoch": 2.88,
-      "learning_rate": 0.0001645664739884393,
-      "loss": 0.2834,
+      "learning_rate": 0.00016473988439306357,
+      "loss": 0.3224,
       "step": 1284
     },
     {
       "epoch": 2.88,
-      "learning_rate": 0.000164393063583815,
-      "loss": 0.2822,
+      "learning_rate": 0.0001645664739884393,
+      "loss": 0.3098,
       "step": 1285
     },
     {
       "epoch": 2.88,
-      "learning_rate": 0.00016421965317919076,
-      "loss": 0.2153,
+      "learning_rate": 0.000164393063583815,
+      "loss": 0.255,
       "step": 1286
     },
     {
       "epoch": 2.89,
-      "learning_rate": 0.00016404624277456645,
-      "loss": 0.2266,
+      "learning_rate": 0.00016421965317919076,
+      "loss": 0.2056,
       "step": 1287
     },
     {
       "epoch": 2.89,
-      "learning_rate": 0.00016387283236994217,
-      "loss": 0.3552,
+      "learning_rate": 0.00016404624277456645,
+      "loss": 0.3752,
       "step": 1288
     },
     {
       "epoch": 2.89,
-      "learning_rate": 0.0001636994219653179,
-      "loss": 0.2672,
+      "learning_rate": 0.00016387283236994217,
+      "loss": 0.2446,
       "step": 1289
     },
     {
       "epoch": 2.89,
-      "learning_rate": 0.00016352601156069364,
-      "loss": 0.3913,
+      "learning_rate": 0.0001636994219653179,
+      "loss": 0.4176,
       "step": 1290
     },
     {
       "epoch": 2.89,
-      "learning_rate": 0.00016335260115606936,
-      "loss": 0.2523,
+      "learning_rate": 0.00016352601156069364,
+      "loss": 0.2583,
       "step": 1291
     },
     {
       "epoch": 2.9,
-      "learning_rate": 0.00016317919075144505,
-      "loss": 0.4232,
+      "learning_rate": 0.00016335260115606936,
+      "loss": 0.5816,
       "step": 1292
     },
     {
       "epoch": 2.9,
-      "learning_rate": 0.00016300578034682077,
-      "loss": 0.491,
+      "learning_rate": 0.00016317919075144505,
+      "loss": 0.591,
       "step": 1293
     },
     {
       "epoch": 2.9,
-      "learning_rate": 0.00016283236994219652,
-      "loss": 0.4315,
+      "learning_rate": 0.00016300578034682077,
+      "loss": 0.4989,
       "step": 1294
     },
     {
       "epoch": 2.9,
-      "learning_rate": 0.00016265895953757224,
-      "loss": 0.3455,
+      "learning_rate": 0.00016283236994219652,
+      "loss": 0.4007,
       "step": 1295
     },
     {
       "epoch": 2.91,
-      "learning_rate": 0.00016248554913294796,
-      "loss": 0.3502,
+      "learning_rate": 0.00016265895953757224,
+      "loss": 0.3692,
       "step": 1296
     },
     {
       "epoch": 2.91,
-      "learning_rate": 0.00016231213872832368,
-      "loss": 0.3384,
+      "learning_rate": 0.00016248554913294796,
+      "loss": 0.3135,
       "step": 1297
     },
     {
       "epoch": 2.91,
-      "learning_rate": 0.00016213872832369943,
-      "loss": 0.3879,
+      "learning_rate": 0.00016231213872832368,
+      "loss": 0.3618,
       "step": 1298
     },
     {
       "epoch": 2.91,
-      "learning_rate": 0.00016196531791907512,
-      "loss": 0.3678,
+      "learning_rate": 0.00016213872832369943,
+      "loss": 0.3442,
       "step": 1299
     },
     {
       "epoch": 2.91,
-      "learning_rate": 0.00016179190751445085,
-      "loss": 0.3088,
+      "learning_rate": 0.00016196531791907512,
+      "loss": 0.3581,
       "step": 1300
     },
     {
       "epoch": 2.92,
-      "learning_rate": 0.00016161849710982657,
-      "loss": 0.2579,
+      "learning_rate": 0.00016179190751445085,
+      "loss": 0.3073,
       "step": 1301
     },
     {
       "epoch": 2.92,
-      "learning_rate": 0.00016144508670520231,
-      "loss": 0.2767,
+      "learning_rate": 0.00016161849710982657,
+      "loss": 0.3005,
       "step": 1302
     },
     {
       "epoch": 2.92,
-      "learning_rate": 0.00016127167630057803,
-      "loss": 0.2344,
+      "learning_rate": 0.00016144508670520231,
+      "loss": 0.2618,
       "step": 1303
     },
     {
       "epoch": 2.92,
-      "learning_rate": 0.00016109826589595373,
-      "loss": 0.2729,
+      "learning_rate": 0.00016127167630057803,
+      "loss": 0.3217,
       "step": 1304
     },
     {
       "epoch": 2.93,
-      "learning_rate": 0.00016092485549132945,
-      "loss": 0.2734,
+      "learning_rate": 0.00016109826589595373,
+      "loss": 0.2993,
       "step": 1305
     },
     {
       "epoch": 2.93,
-      "learning_rate": 0.0001607514450867052,
-      "loss": 0.2517,
+      "learning_rate": 0.00016092485549132945,
+      "loss": 0.2586,
       "step": 1306
     },
     {
       "epoch": 2.93,
-      "learning_rate": 0.00016057803468208092,
-      "loss": 0.228,
+      "learning_rate": 0.0001607514450867052,
+      "loss": 0.2481,
       "step": 1307
     },
     {
       "epoch": 2.93,
-      "learning_rate": 0.00016040462427745664,
-      "loss": 0.2728,
+      "learning_rate": 0.00016057803468208092,
+      "loss": 0.2872,
       "step": 1308
     },
     {
       "epoch": 2.93,
-      "learning_rate": 0.00016023121387283233,
-      "loss": 0.2453,
+      "learning_rate": 0.00016040462427745664,
+      "loss": 0.2772,
       "step": 1309
     },
     {
       "epoch": 2.94,
-      "learning_rate": 0.00016005780346820808,
-      "loss": 0.2424,
+      "learning_rate": 0.00016023121387283233,
+      "loss": 0.2466,
       "step": 1310
     },
     {
       "epoch": 2.94,
-      "learning_rate": 0.0001598843930635838,
-      "loss": 0.3022,
+      "learning_rate": 0.00016005780346820808,
+      "loss": 0.3004,
       "step": 1311
     },
     {
       "epoch": 2.94,
-      "learning_rate": 0.00015971098265895952,
-      "loss": 0.15,
+      "learning_rate": 0.0001598843930635838,
+      "loss": 0.1622,
       "step": 1312
     },
     {
       "epoch": 2.94,
-      "learning_rate": 0.00015953757225433524,
-      "loss": 0.2304,
+      "learning_rate": 0.00015971098265895952,
+      "loss": 0.2447,
       "step": 1313
     },
     {
       "epoch": 2.95,
-      "learning_rate": 0.000159364161849711,
-      "loss": 0.2228,
+      "learning_rate": 0.00015953757225433524,
+      "loss": 0.2302,
       "step": 1314
     },
     {
       "epoch": 2.95,
-      "learning_rate": 0.00015919075144508668,
-      "loss": 0.2411,
+      "learning_rate": 0.000159364161849711,
+      "loss": 0.2657,
       "step": 1315
     },
     {
       "epoch": 2.95,
-      "learning_rate": 0.0001590173410404624,
-      "loss": 0.2497,
+      "learning_rate": 0.00015919075144508668,
+      "loss": 0.267,
       "step": 1316
     },
     {
       "epoch": 2.95,
-      "learning_rate": 0.00015884393063583812,
-      "loss": 0.2196,
+      "learning_rate": 0.0001590173410404624,
+      "loss": 0.243,
       "step": 1317
     },
     {
       "epoch": 2.96,
-      "learning_rate": 0.00015867052023121387,
-      "loss": 0.2327,
+      "learning_rate": 0.00015884393063583812,
+      "loss": 0.2366,
       "step": 1318
     },
     {
       "epoch": 2.96,
-      "learning_rate": 0.0001584971098265896,
-      "loss": 0.2496,
+      "learning_rate": 0.00015867052023121387,
+      "loss": 0.2866,
       "step": 1319
     },
     {
       "epoch": 2.96,
-      "learning_rate": 0.0001583236994219653,
-      "loss": 0.2477,
+      "learning_rate": 0.0001584971098265896,
+      "loss": 0.2798,
       "step": 1320
     },
     {
       "epoch": 2.96,
-      "learning_rate": 0.000158150289017341,
-      "loss": 0.3474,
+      "learning_rate": 0.0001583236994219653,
+      "loss": 0.3047,
       "step": 1321
     },
     {
       "epoch": 2.96,
-      "learning_rate": 0.00015797687861271675,
-      "loss": 0.2032,
+      "learning_rate": 0.000158150289017341,
+      "loss": 0.2196,
       "step": 1322
     },
     {
       "epoch": 2.97,
-      "learning_rate": 0.00015780346820809248,
-      "loss": 0.2555,
+      "learning_rate": 0.00015797687861271675,
+      "loss": 0.2324,
       "step": 1323
     },
     {
       "epoch": 2.97,
-      "learning_rate": 0.0001576300578034682,
-      "loss": 0.2592,
+      "learning_rate": 0.00015780346820809248,
+      "loss": 0.2934,
       "step": 1324
     },
     {
       "epoch": 2.97,
-      "learning_rate": 0.00015745664739884392,
-      "loss": 0.2183,
+      "learning_rate": 0.0001576300578034682,
+      "loss": 0.2647,
       "step": 1325
     },
     {
       "epoch": 2.97,
-      "learning_rate": 0.00015728323699421966,
-      "loss": 0.2331,
+      "learning_rate": 0.00015745664739884392,
+      "loss": 0.245,
       "step": 1326
     },
     {
       "epoch": 2.98,
-      "learning_rate": 0.00015710982658959536,
-      "loss": 0.1846,
+      "learning_rate": 0.00015728323699421966,
+      "loss": 0.2012,
       "step": 1327
     },
     {
       "epoch": 2.98,
-      "learning_rate": 0.00015693641618497108,
-      "loss": 0.2293,
+      "learning_rate": 0.00015710982658959536,
+      "loss": 0.2425,
       "step": 1328
     },
     {
       "epoch": 2.98,
-      "learning_rate": 0.0001567630057803468,
-      "loss": 0.194,
+      "learning_rate": 0.00015693641618497108,
+      "loss": 0.229,
       "step": 1329
     },
     {
       "epoch": 2.98,
-      "learning_rate": 0.00015658959537572255,
-      "loss": 0.2433,
+      "learning_rate": 0.0001567630057803468,
+      "loss": 0.2368,
       "step": 1330
     },
     {
       "epoch": 2.98,
-      "learning_rate": 0.00015641618497109827,
-      "loss": 0.2455,
+      "learning_rate": 0.00015658959537572255,
+      "loss": 0.2612,
       "step": 1331
     },
     {
       "epoch": 2.99,
-      "learning_rate": 0.00015624277456647396,
-      "loss": 0.3042,
+      "learning_rate": 0.00015641618497109827,
+      "loss": 0.2954,
       "step": 1332
     },
     {
       "epoch": 2.99,
-      "learning_rate": 0.00015606936416184968,
-      "loss": 0.2411,
+      "learning_rate": 0.00015624277456647396,
+      "loss": 0.2647,
       "step": 1333
     },
     {
       "epoch": 2.99,
-      "learning_rate": 0.00015589595375722543,
-      "loss": 0.2174,
+      "learning_rate": 0.00015606936416184968,
+      "loss": 0.2506,
       "step": 1334
     },
     {
       "epoch": 2.99,
-      "learning_rate": 0.00015572254335260115,
-      "loss": 0.3234,
+      "learning_rate": 0.00015589595375722543,
+      "loss": 0.2866,
       "step": 1335
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00015554913294797687,
-      "loss": 0.2283,
+      "learning_rate": 0.00015572254335260115,
+      "loss": 0.2059,
       "step": 1336
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00015537572254335256,
-      "loss": 0.2508,
+      "learning_rate": 0.00015554913294797687,
+      "loss": 0.2379,
       "step": 1337
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.0001552023121387283,
-      "loss": 0.3316,
+      "learning_rate": 0.00015537572254335256,
+      "loss": 0.3188,
       "step": 1338
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00015502890173410403,
-      "loss": 0.3798,
+      "learning_rate": 0.0001552023121387283,
+      "loss": 0.3332,
       "step": 1339
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00015485549132947975,
-      "loss": 0.3203,
+      "learning_rate": 0.00015502890173410403,
+      "loss": 0.3164,
       "step": 1340
     },
     {
       "epoch": 3.01,
-      "learning_rate": 0.00015468208092485547,
-      "loss": 0.325,
+      "learning_rate": 0.00015485549132947975,
+      "loss": 0.3005,
       "step": 1341
     },
     {
       "epoch": 3.01,
-      "learning_rate": 0.00015450867052023122,
-      "loss": 0.2678,
+      "learning_rate": 0.00015468208092485547,
+      "loss": 0.1913,
       "step": 1342
     },
     {
       "epoch": 3.01,
-      "learning_rate": 0.00015433526011560692,
-      "loss": 0.2495,
+      "learning_rate": 0.00015450867052023122,
+      "loss": 0.2485,
       "step": 1343
     },
     {
       "epoch": 3.01,
-      "learning_rate": 0.00015416184971098264,
-      "loss": 0.2213,
+      "learning_rate": 0.00015433526011560692,
+      "loss": 0.2041,
       "step": 1344
     },
     {
       "epoch": 3.02,
-      "learning_rate": 0.00015398843930635836,
-      "loss": 0.2451,
+      "learning_rate": 0.00015416184971098264,
+      "loss": 0.2291,
       "step": 1345
     },
     {
       "epoch": 3.02,
-      "learning_rate": 0.0001538150289017341,
-      "loss": 0.2073,
+      "learning_rate": 0.00015398843930635836,
+      "loss": 0.1931,
       "step": 1346
     },
     {
       "epoch": 3.02,
-      "learning_rate": 0.00015364161849710983,
-      "loss": 0.205,
+      "learning_rate": 0.0001538150289017341,
+      "loss": 0.1707,
       "step": 1347
     },
     {
       "epoch": 3.02,
-      "learning_rate": 0.00015346820809248555,
-      "loss": 0.2282,
+      "learning_rate": 0.00015364161849710983,
+      "loss": 0.2488,
       "step": 1348
     },
     {
       "epoch": 3.02,
-      "learning_rate": 0.00015329479768786124,
-      "loss": 0.1877,
+      "learning_rate": 0.00015346820809248555,
+      "loss": 0.167,
       "step": 1349
     },
     {
       "epoch": 3.03,
-      "learning_rate": 0.000153121387283237,
-      "loss": 0.1867,
+      "learning_rate": 0.00015329479768786124,
+      "loss": 0.1798,
       "step": 1350
     },
     {
       "epoch": 3.03,
-      "learning_rate": 0.0001529479768786127,
-      "loss": 0.1768,
+      "learning_rate": 0.000153121387283237,
+      "loss": 0.1798,
       "step": 1351
     },
     {
       "epoch": 3.03,
-      "learning_rate": 0.00015277456647398843,
-      "loss": 0.1304,
+      "learning_rate": 0.0001529479768786127,
+      "loss": 0.1232,
       "step": 1352
     },
     {
       "epoch": 3.03,
-      "learning_rate": 0.00015260115606936415,
-      "loss": 0.1706,
+      "learning_rate": 0.00015277456647398843,
+      "loss": 0.171,
       "step": 1353
     },
     {
       "epoch": 3.04,
-      "learning_rate": 0.0001524277456647399,
-      "loss": 0.2271,
+      "learning_rate": 0.00015260115606936415,
+      "loss": 0.2154,
       "step": 1354
     },
     {
       "epoch": 3.04,
-      "learning_rate": 0.0001522543352601156,
-      "loss": 0.1508,
+      "learning_rate": 0.0001524277456647399,
+      "loss": 0.1341,
       "step": 1355
     },
     {
       "epoch": 3.04,
-      "learning_rate": 0.0001520809248554913,
-      "loss": 0.1696,
+      "learning_rate": 0.0001522543352601156,
+      "loss": 0.1676,
       "step": 1356
     },
     {
       "epoch": 3.04,
-      "learning_rate": 0.00015190751445086703,
-      "loss": 0.1688,
+      "learning_rate": 0.0001520809248554913,
+      "loss": 0.1611,
       "step": 1357
     },
     {
       "epoch": 3.04,
-      "learning_rate": 0.00015173410404624278,
-      "loss": 0.2154,
+      "learning_rate": 0.00015190751445086703,
+      "loss": 0.1838,
       "step": 1358
     },
     {
       "epoch": 3.05,
-      "learning_rate": 0.0001515606936416185,
-      "loss": 0.1489,
+      "learning_rate": 0.00015173410404624278,
+      "loss": 0.164,
       "step": 1359
     },
     {
       "epoch": 3.05,
-      "learning_rate": 0.0001513872832369942,
-      "loss": 0.1893,
+      "learning_rate": 0.0001515606936416185,
+      "loss": 0.1697,
       "step": 1360
     },
     {
       "epoch": 3.05,
-      "learning_rate": 0.00015121387283236992,
-      "loss": 0.188,
+      "learning_rate": 0.0001513872832369942,
+      "loss": 0.1793,
       "step": 1361
     },
     {
       "epoch": 3.05,
-      "learning_rate": 0.00015104046242774566,
-      "loss": 0.172,
+      "learning_rate": 0.00015121387283236992,
+      "loss": 0.1421,
       "step": 1362
     },
     {
       "epoch": 3.06,
-      "learning_rate": 0.00015086705202312138,
-      "loss": 0.1528,
+      "learning_rate": 0.00015104046242774566,
+      "loss": 0.1488,
       "step": 1363
     },
     {
       "epoch": 3.06,
-      "learning_rate": 0.0001506936416184971,
-      "loss": 0.1574,
+      "learning_rate": 0.00015086705202312138,
+      "loss": 0.1655,
       "step": 1364
     },
     {
       "epoch": 3.06,
-      "learning_rate": 0.0001505202312138728,
-      "loss": 0.1721,
+      "learning_rate": 0.0001506936416184971,
+      "loss": 0.2056,
       "step": 1365
     },
     {
       "epoch": 3.06,
-      "learning_rate": 0.00015034682080924855,
-      "loss": 0.1366,
+      "learning_rate": 0.0001505202312138728,
+      "loss": 0.121,
       "step": 1366
     },
     {
       "epoch": 3.07,
-      "learning_rate": 0.00015017341040462427,
-      "loss": 0.1438,
+      "learning_rate": 0.00015034682080924855,
+      "loss": 0.1514,
       "step": 1367
     },
     {
       "epoch": 3.07,
-      "learning_rate": 0.00015,
-      "loss": 0.1591,
+      "learning_rate": 0.00015017341040462427,
+      "loss": 0.1511,
       "step": 1368
     },
     {
       "epoch": 3.07,
-      "learning_rate": 0.0001498265895953757,
-      "loss": 0.1695,
+      "learning_rate": 0.00015,
+      "loss": 0.1363,
       "step": 1369
     },
     {
       "epoch": 3.07,
-      "learning_rate": 0.00014965317919075143,
-      "loss": 0.1676,
+      "learning_rate": 0.0001498265895953757,
+      "loss": 0.1644,
       "step": 1370
     },
     {
       "epoch": 3.07,
-      "learning_rate": 0.00014947976878612715,
-      "loss": 0.1333,
+      "learning_rate": 0.00014965317919075143,
+      "loss": 0.1106,
       "step": 1371
     },
     {
       "epoch": 3.08,
-      "learning_rate": 0.00014930635838150287,
-      "loss": 0.1507,
+      "learning_rate": 0.00014947976878612715,
+      "loss": 0.1259,
       "step": 1372
     },
     {
       "epoch": 3.08,
-      "learning_rate": 0.0001491329479768786,
-      "loss": 0.1445,
+      "learning_rate": 0.00014930635838150287,
+      "loss": 0.1348,
       "step": 1373
     },
     {
       "epoch": 3.08,
-      "learning_rate": 0.0001489595375722543,
-      "loss": 0.1443,
+      "learning_rate": 0.0001491329479768786,
+      "loss": 0.1144,
       "step": 1374
     },
     {
       "epoch": 3.08,
-      "learning_rate": 0.00014878612716763003,
-      "loss": 0.1344,
+      "learning_rate": 0.0001489595375722543,
+      "loss": 0.1318,
       "step": 1375
     },
     {
       "epoch": 3.09,
-      "learning_rate": 0.00014861271676300578,
-      "loss": 0.1514,
+      "learning_rate": 0.00014878612716763003,
+      "loss": 0.1212,
       "step": 1376
     },
     {
       "epoch": 3.09,
-      "learning_rate": 0.00014843930635838147,
-      "loss": 0.1539,
+      "learning_rate": 0.00014861271676300578,
+      "loss": 0.1427,
       "step": 1377
     },
     {
       "epoch": 3.09,
-      "learning_rate": 0.00014826589595375722,
-      "loss": 0.1185,
+      "learning_rate": 0.00014843930635838147,
+      "loss": 0.1204,
       "step": 1378
     },
     {
       "epoch": 3.09,
-      "learning_rate": 0.00014809248554913291,
-      "loss": 0.4735,
+      "learning_rate": 0.00014826589595375722,
+      "loss": 0.6003,
       "step": 1379
     },
     {
       "epoch": 3.09,
-      "learning_rate": 0.00014791907514450866,
-      "loss": 0.169,
+      "learning_rate": 0.00014809248554913291,
+      "loss": 0.1959,
       "step": 1380
     },
     {
       "epoch": 3.1,
-      "learning_rate": 0.00014774566473988438,
-      "loss": 0.1536,
+      "learning_rate": 0.00014791907514450866,
+      "loss": 0.1231,
       "step": 1381
     },
     {
       "epoch": 3.1,
-      "learning_rate": 0.0001475722543352601,
-      "loss": 0.1547,
+      "learning_rate": 0.00014774566473988438,
+      "loss": 0.1274,
       "step": 1382
     },
     {
       "epoch": 3.1,
-      "learning_rate": 0.00014739884393063582,
-      "loss": 0.1347,
+      "learning_rate": 0.0001475722543352601,
+      "loss": 0.1249,
       "step": 1383
     },
     {
       "epoch": 3.1,
-      "learning_rate": 0.00014722543352601154,
-      "loss": 0.1757,
+      "learning_rate": 0.00014739884393063582,
+      "loss": 0.1267,
       "step": 1384
     },
     {
       "epoch": 3.11,
-      "learning_rate": 0.00014705202312138727,
-      "loss": 0.1752,
+      "learning_rate": 0.00014722543352601154,
+      "loss": 0.1784,
       "step": 1385
     },
     {
       "epoch": 3.11,
-      "learning_rate": 0.00014687861271676299,
-      "loss": 0.186,
+      "learning_rate": 0.00014705202312138727,
+      "loss": 0.17,
       "step": 1386
     },
     {
       "epoch": 3.11,
-      "learning_rate": 0.0001467052023121387,
-      "loss": 0.1712,
+      "learning_rate": 0.00014687861271676299,
+      "loss": 0.1838,
       "step": 1387
     },
     {
       "epoch": 3.11,
-      "learning_rate": 0.00014653179190751443,
-      "loss": 0.2316,
+      "learning_rate": 0.0001467052023121387,
+      "loss": 0.2029,
       "step": 1388
     },
     {
       "epoch": 3.11,
-      "learning_rate": 0.00014635838150289015,
-      "loss": 0.4119,
+      "learning_rate": 0.00014653179190751443,
+      "loss": 0.3663,
       "step": 1389
     },
     {
       "epoch": 3.12,
-      "learning_rate": 0.0001461849710982659,
-      "loss": 0.3495,
+      "learning_rate": 0.00014635838150289015,
+      "loss": 0.3393,
       "step": 1390
     },
     {
       "epoch": 3.12,
-      "learning_rate": 0.0001460115606936416,
-      "loss": 0.2835,
+      "learning_rate": 0.0001461849710982659,
+      "loss": 0.2693,
       "step": 1391
     },
     {
       "epoch": 3.12,
-      "learning_rate": 0.00014583815028901734,
-      "loss": 0.2823,
+      "learning_rate": 0.0001460115606936416,
+      "loss": 0.2366,
       "step": 1392
     },
     {
       "epoch": 3.12,
-      "learning_rate": 0.00014566473988439306,
-      "loss": 0.2574,
+      "learning_rate": 0.00014583815028901734,
+      "loss": 0.2271,
       "step": 1393
     },
     {
       "epoch": 3.13,
-      "learning_rate": 0.00014549132947976878,
-      "loss": 0.2407,
+      "learning_rate": 0.00014566473988439306,
+      "loss": 0.2333,
       "step": 1394
     },
     {
       "epoch": 3.13,
-      "learning_rate": 0.0001453179190751445,
-      "loss": 0.2446,
+      "learning_rate": 0.00014549132947976878,
+      "loss": 0.2391,
       "step": 1395
     },
     {
       "epoch": 3.13,
-      "learning_rate": 0.00014514450867052022,
-      "loss": 0.2939,
+      "learning_rate": 0.0001453179190751445,
+      "loss": 0.563,
       "step": 1396
     },
     {
       "epoch": 3.13,
-      "learning_rate": 0.00014497109826589594,
-      "loss": 0.1817,
+      "learning_rate": 0.00014514450867052022,
+      "loss": 0.1767,
       "step": 1397
     },
     {
       "epoch": 3.13,
-      "learning_rate": 0.00014479768786127166,
-      "loss": 0.2134,
+      "learning_rate": 0.00014497109826589594,
+      "loss": 0.2116,
       "step": 1398
     },
     {
       "epoch": 3.14,
-      "learning_rate": 0.00014462427745664738,
-      "loss": 0.177,
+      "learning_rate": 0.00014479768786127166,
+      "loss": 0.1731,
       "step": 1399
     },
     {
       "epoch": 3.14,
-      "learning_rate": 0.0001444508670520231,
-      "loss": 0.2093,
+      "learning_rate": 0.00014462427745664738,
+      "loss": 0.2047,
       "step": 1400
     },
     {
       "epoch": 3.14,
-      "learning_rate": 0.00014427745664739882,
-      "loss": 0.204,
+      "learning_rate": 0.0001444508670520231,
+      "loss": 0.1805,
       "step": 1401
     },
     {
       "epoch": 3.14,
-      "learning_rate": 0.00014410404624277454,
-      "loss": 0.1778,
+      "learning_rate": 0.00014427745664739882,
+      "loss": 0.1929,
       "step": 1402
     },
     {
       "epoch": 3.15,
-      "learning_rate": 0.00014393063583815026,
-      "loss": 0.1262,
+      "learning_rate": 0.00014410404624277454,
+      "loss": 0.1438,
       "step": 1403
     },
     {
       "epoch": 3.15,
-      "learning_rate": 0.000143757225433526,
-      "loss": 0.1809,
+      "learning_rate": 0.00014393063583815026,
+      "loss": 0.1483,
       "step": 1404
     },
     {
       "epoch": 3.15,
-      "learning_rate": 0.0001435838150289017,
-      "loss": 0.2144,
+      "learning_rate": 0.000143757225433526,
+      "loss": 0.1981,
       "step": 1405
     },
     {
       "epoch": 3.15,
-      "learning_rate": 0.00014341040462427745,
-      "loss": 0.208,
+      "learning_rate": 0.0001435838150289017,
+      "loss": 0.1836,
       "step": 1406
     },
     {
       "epoch": 3.15,
-      "learning_rate": 0.00014323699421965317,
-      "loss": 0.1815,
+      "learning_rate": 0.00014341040462427745,
+      "loss": 0.1852,
       "step": 1407
     },
     {
       "epoch": 3.16,
-      "learning_rate": 0.0001430635838150289,
-      "loss": 0.1736,
+      "learning_rate": 0.00014323699421965317,
+      "loss": 0.1698,
       "step": 1408
     },
     {
       "epoch": 3.16,
-      "learning_rate": 0.00014289017341040462,
-      "loss": 0.1365,
+      "learning_rate": 0.0001430635838150289,
+      "loss": 0.1216,
       "step": 1409
     },
     {
       "epoch": 3.16,
-      "learning_rate": 0.00014271676300578034,
-      "loss": 0.192,
+      "learning_rate": 0.00014289017341040462,
+      "loss": 0.1676,
       "step": 1410
     },
     {
       "epoch": 3.16,
-      "learning_rate": 0.00014254335260115606,
-      "loss": 0.1793,
+      "learning_rate": 0.00014271676300578034,
+      "loss": 0.1809,
       "step": 1411
     },
     {
       "epoch": 3.17,
-      "learning_rate": 0.00014236994219653178,
-      "loss": 0.1523,
+      "learning_rate": 0.00014254335260115606,
+      "loss": 0.1413,
       "step": 1412
     },
     {
       "epoch": 3.17,
-      "learning_rate": 0.0001421965317919075,
-      "loss": 0.1462,
+      "learning_rate": 0.00014236994219653178,
+      "loss": 0.137,
       "step": 1413
     },
     {
       "epoch": 3.17,
-      "learning_rate": 0.00014202312138728322,
-      "loss": 0.1728,
+      "learning_rate": 0.0001421965317919075,
+      "loss": 0.1761,
       "step": 1414
     },
     {
       "epoch": 3.17,
-      "learning_rate": 0.00014184971098265894,
-      "loss": 0.1174,
+      "learning_rate": 0.00014202312138728322,
+      "loss": 0.126,
       "step": 1415
     },
     {
       "epoch": 3.17,
-      "learning_rate": 0.00014167630057803466,
-      "loss": 0.1724,
+      "learning_rate": 0.00014184971098265894,
+      "loss": 0.1518,
       "step": 1416
     },
     {
       "epoch": 3.18,
-      "learning_rate": 0.00014150289017341038,
-      "loss": 0.1518,
+      "learning_rate": 0.00014167630057803466,
+      "loss": 0.1303,
       "step": 1417
     },
     {
       "epoch": 3.18,
-      "learning_rate": 0.00014132947976878613,
-      "loss": 0.1748,
+      "learning_rate": 0.00014150289017341038,
+      "loss": 0.1535,
       "step": 1418
     },
     {
       "epoch": 3.18,
-      "learning_rate": 0.00014115606936416182,
-      "loss": 0.1524,
+      "learning_rate": 0.00014132947976878613,
+      "loss": 0.1286,
       "step": 1419
     },
     {
       "epoch": 3.18,
-      "learning_rate": 0.00014098265895953757,
-      "loss": 0.1267,
+      "learning_rate": 0.00014115606936416182,
+      "loss": 0.1395,
       "step": 1420
     },
     {
       "epoch": 3.19,
-      "learning_rate": 0.0001408092485549133,
-      "loss": 0.1518,
+      "learning_rate": 0.00014098265895953757,
+      "loss": 0.1791,
       "step": 1421
     },
     {
       "epoch": 3.19,
-      "learning_rate": 0.000140635838150289,
-      "loss": 0.1754,
+      "learning_rate": 0.0001408092485549133,
+      "loss": 0.1708,
       "step": 1422
     },
     {
       "epoch": 3.19,
-      "learning_rate": 0.00014046242774566473,
-      "loss": 0.1313,
+      "learning_rate": 0.000140635838150289,
+      "loss": 0.5027,
       "step": 1423
     },
     {
       "epoch": 3.19,
-      "learning_rate": 0.00014028901734104045,
-      "loss": 0.1822,
+      "learning_rate": 0.00014046242774566473,
+      "loss": 0.1802,
       "step": 1424
     },
     {
       "epoch": 3.2,
-      "learning_rate": 0.00014011560693641617,
-      "loss": 0.134,
+      "learning_rate": 0.00014028901734104045,
+      "loss": 0.1555,
       "step": 1425
     },
     {
       "epoch": 3.2,
-      "learning_rate": 0.0001399421965317919,
-      "loss": 0.1302,
+      "learning_rate": 0.00014011560693641617,
+      "loss": 0.1257,
       "step": 1426
     },
     {
       "epoch": 3.2,
-      "learning_rate": 0.00013976878612716762,
-      "loss": 0.1575,
+      "learning_rate": 0.0001399421965317919,
+      "loss": 0.1221,
       "step": 1427
     },
     {
       "epoch": 3.2,
-      "learning_rate": 0.00013959537572254334,
-      "loss": 0.1463,
+      "learning_rate": 0.00013976878612716762,
+      "loss": 0.1183,
       "step": 1428
     },
     {
       "epoch": 3.2,
-      "learning_rate": 0.00013942196531791906,
-      "loss": 0.1395,
+      "learning_rate": 0.00013959537572254334,
+      "loss": 0.1121,
       "step": 1429
     },
     {
       "epoch": 3.21,
-      "learning_rate": 0.00013924855491329478,
-      "loss": 0.1407,
+      "learning_rate": 0.00013942196531791906,
+      "loss": 0.1325,
       "step": 1430
     },
     {
       "epoch": 3.21,
-      "learning_rate": 0.0001390751445086705,
-      "loss": 0.1618,
+      "learning_rate": 0.00013924855491329478,
+      "loss": 0.1182,
       "step": 1431
     },
     {
       "epoch": 3.21,
-      "learning_rate": 0.00013890173410404625,
-      "loss": 0.1844,
+      "learning_rate": 0.0001390751445086705,
+      "loss": 0.1615,
       "step": 1432
     },
     {
       "epoch": 3.21,
-      "learning_rate": 0.00013872832369942194,
-      "loss": 0.1514,
+      "learning_rate": 0.00013890173410404625,
+      "loss": 0.1368,
       "step": 1433
     },
     {
       "epoch": 3.22,
-      "learning_rate": 0.0001385549132947977,
-      "loss": 0.1257,
+      "learning_rate": 0.00013872832369942194,
+      "loss": 0.1295,
       "step": 1434
     },
     {
       "epoch": 3.22,
-      "learning_rate": 0.0001383815028901734,
-      "loss": 0.1139,
+      "learning_rate": 0.0001385549132947977,
+      "loss": 0.0937,
       "step": 1435
     },
     {
       "epoch": 3.22,
-      "learning_rate": 0.00013820809248554913,
-      "loss": 0.1438,
+      "learning_rate": 0.0001383815028901734,
+      "loss": 0.113,
       "step": 1436
     },
     {
       "epoch": 3.22,
-      "learning_rate": 0.00013803468208092485,
-      "loss": 0.1302,
+      "learning_rate": 0.00013820809248554913,
+      "loss": 0.1177,
       "step": 1437
     },
     {
       "epoch": 3.22,
-      "learning_rate": 0.00013786127167630057,
-      "loss": 0.1929,
+      "learning_rate": 0.00013803468208092485,
+      "loss": 0.1841,
       "step": 1438
     },
     {
       "epoch": 3.23,
-      "learning_rate": 0.0001376878612716763,
-      "loss": 0.3476,
+      "learning_rate": 0.00013786127167630057,
+      "loss": 0.3102,
       "step": 1439
     },
     {
       "epoch": 3.23,
-      "learning_rate": 0.000137514450867052,
-      "loss": 0.3686,
+      "learning_rate": 0.0001376878612716763,
+      "loss": 0.3234,
       "step": 1440
     },
     {
       "epoch": 3.23,
-      "learning_rate": 0.00013734104046242773,
-      "loss": 0.2973,
+      "learning_rate": 0.000137514450867052,
+      "loss": 0.3114,
       "step": 1441
     },
     {
       "epoch": 3.23,
-      "learning_rate": 0.00013716763005780345,
-      "loss": 0.2733,
+      "learning_rate": 0.00013734104046242773,
+      "loss": 0.2625,
       "step": 1442
     },
     {
       "epoch": 3.24,
-      "learning_rate": 0.00013699421965317917,
-      "loss": 0.2181,
+      "learning_rate": 0.00013716763005780345,
+      "loss": 0.1896,
       "step": 1443
     },
     {
       "epoch": 3.24,
-      "learning_rate": 0.00013682080924855492,
-      "loss": 0.2085,
+      "learning_rate": 0.00013699421965317917,
+      "loss": 0.1894,
       "step": 1444
     },
     {
       "epoch": 3.24,
-      "learning_rate": 0.00013664739884393061,
-      "loss": 0.2227,
+      "learning_rate": 0.00013682080924855492,
+      "loss": 0.1994,
       "step": 1445
     },
     {
       "epoch": 3.24,
-      "learning_rate": 0.00013647398843930636,
-      "loss": 0.2176,
+      "learning_rate": 0.00013664739884393061,
+      "loss": 0.1791,
       "step": 1446
     },
     {
       "epoch": 3.24,
-      "learning_rate": 0.00013630057803468206,
-      "loss": 0.2107,
+      "learning_rate": 0.00013647398843930636,
+      "loss": 0.213,
       "step": 1447
     },
     {
       "epoch": 3.25,
-      "learning_rate": 0.0001361271676300578,
-      "loss": 0.1943,
+      "learning_rate": 0.00013630057803468206,
+      "loss": 0.1923,
       "step": 1448
     },
     {
       "epoch": 3.25,
-      "learning_rate": 0.00013595375722543352,
-      "loss": 0.1552,
+      "learning_rate": 0.0001361271676300578,
+      "loss": 0.1412,
       "step": 1449
     },
     {
       "epoch": 3.25,
-      "learning_rate": 0.00013578034682080925,
-      "loss": 0.2,
+      "learning_rate": 0.00013595375722543352,
+      "loss": 0.1965,
       "step": 1450
     },
     {
       "epoch": 3.25,
-      "learning_rate": 0.00013560693641618497,
-      "loss": 0.2198,
+      "learning_rate": 0.00013578034682080925,
+      "loss": 0.1976,
       "step": 1451
     },
     {
       "epoch": 3.26,
-      "learning_rate": 0.0001354335260115607,
-      "loss": 0.1381,
+      "learning_rate": 0.00013560693641618497,
+      "loss": 0.1655,
       "step": 1452
     },
     {
       "epoch": 3.26,
-      "learning_rate": 0.0001352601156069364,
-      "loss": 0.2021,
+      "learning_rate": 0.0001354335260115607,
+      "loss": 0.2195,
       "step": 1453
     },
     {
       "epoch": 3.26,
-      "learning_rate": 0.00013508670520231213,
-      "loss": 0.19,
+      "learning_rate": 0.0001352601156069364,
+      "loss": 0.1942,
       "step": 1454
     },
     {
       "epoch": 3.26,
-      "learning_rate": 0.00013491329479768785,
-      "loss": 0.1655,
+      "learning_rate": 0.00013508670520231213,
+      "loss": 0.1557,
       "step": 1455
     },
     {
       "epoch": 3.26,
-      "learning_rate": 0.00013473988439306357,
-      "loss": 0.1811,
+      "learning_rate": 0.00013491329479768785,
+      "loss": 0.155,
       "step": 1456
     },
     {
       "epoch": 3.27,
-      "learning_rate": 0.0001345664739884393,
-      "loss": 0.1644,
+      "learning_rate": 0.00013473988439306357,
+      "loss": 0.1443,
       "step": 1457
     },
     {
       "epoch": 3.27,
-      "learning_rate": 0.00013439306358381504,
-      "loss": 0.1778,
+      "learning_rate": 0.0001345664739884393,
+      "loss": 0.1514,
       "step": 1458
     },
     {
       "epoch": 3.27,
-      "learning_rate": 0.00013421965317919073,
-      "loss": 0.1469,
+      "learning_rate": 0.00013439306358381504,
+      "loss": 0.1367,
       "step": 1459
     },
     {
       "epoch": 3.27,
-      "learning_rate": 0.00013404624277456648,
-      "loss": 0.1454,
+      "learning_rate": 0.00013421965317919073,
+      "loss": 0.1459,
       "step": 1460
     },
     {
       "epoch": 3.28,
-      "learning_rate": 0.00013387283236994217,
-      "loss": 0.194,
+      "learning_rate": 0.00013404624277456648,
+      "loss": 0.1873,
       "step": 1461
     },
     {
       "epoch": 3.28,
-      "learning_rate": 0.00013369942196531792,
-      "loss": 0.1461,
+      "learning_rate": 0.00013387283236994217,
+      "loss": 0.1342,
       "step": 1462
     },
     {
       "epoch": 3.28,
-      "learning_rate": 0.00013352601156069364,
-      "loss": 0.1489,
+      "learning_rate": 0.00013369942196531792,
+      "loss": 0.1421,
       "step": 1463
     },
     {
       "epoch": 3.28,
-      "learning_rate": 0.00013335260115606936,
-      "loss": 0.1083,
+      "learning_rate": 0.00013352601156069364,
+      "loss": 0.1153,
       "step": 1464
     },
     {
       "epoch": 3.28,
-      "learning_rate": 0.00013317919075144508,
-      "loss": 0.1315,
+      "learning_rate": 0.00013335260115606936,
+      "loss": 0.1168,
       "step": 1465
     },
     {
       "epoch": 3.29,
-      "learning_rate": 0.00013300578034682078,
-      "loss": 0.176,
+      "learning_rate": 0.00013317919075144508,
+      "loss": 0.1794,
       "step": 1466
     },
     {
       "epoch": 3.29,
       "learning_rate": 0.00013300578034682078,
-      "loss": 0.3109,
+      "loss": 0.2179,
       "step": 1467
     },
     {
       "epoch": 3.29,
       "learning_rate": 0.00013283236994219652,
-      "loss": 0.1548,
+      "loss": 0.1447,
       "step": 1468
     },
     {
       "epoch": 3.29,
       "learning_rate": 0.00013265895953757224,
-      "loss": 0.2042,
+      "loss": 0.1838,
       "step": 1469
     },
     {
       "epoch": 3.3,
       "learning_rate": 0.00013248554913294797,
-      "loss": 0.1323,
+      "loss": 0.1361,
       "step": 1470
     },
     {
       "epoch": 3.3,
       "learning_rate": 0.00013231213872832369,
-      "loss": 0.1398,
+      "loss": 0.1267,
       "step": 1471
     },
     {
       "epoch": 3.3,
       "learning_rate": 0.0001321387283236994,
-      "loss": 0.1565,
+      "loss": 0.1564,
       "step": 1472
     },
     {
       "epoch": 3.3,
       "learning_rate": 0.00013196531791907513,
-      "loss": 0.1623,
+      "loss": 0.1736,
       "step": 1473
     },
     {
       "epoch": 3.3,
       "learning_rate": 0.00013179190751445085,
-      "loss": 0.1512,
+      "loss": 0.1423,
       "step": 1474
     },
     {
       "epoch": 3.31,
       "learning_rate": 0.00013161849710982657,
-      "loss": 0.1239,
+      "loss": 0.1301,
       "step": 1475
     },
     {
       "epoch": 3.31,
       "learning_rate": 0.0001314450867052023,
-      "loss": 0.1257,
+      "loss": 0.1068,
       "step": 1476
     },
     {
       "epoch": 3.31,
       "learning_rate": 0.000131271676300578,
-      "loss": 0.1207,
+      "loss": 0.1306,
       "step": 1477
     },
     {
       "epoch": 3.31,
       "learning_rate": 0.00013109826589595376,
-      "loss": 0.1297,
+      "loss": 0.1315,
       "step": 1478
     },
     {
       "epoch": 3.32,
       "learning_rate": 0.00013092485549132945,
-      "loss": 0.1387,
+      "loss": 0.1319,
       "step": 1479
     },
     {
       "epoch": 3.32,
       "learning_rate": 0.0001307514450867052,
-      "loss": 0.1363,
+      "loss": 0.1319,
       "step": 1480
     },
     {
       "epoch": 3.32,
       "learning_rate": 0.00013057803468208092,
-      "loss": 0.2004,
+      "loss": 0.4801,
       "step": 1481
     },
     {
       "epoch": 3.32,
       "learning_rate": 0.00013040462427745664,
-      "loss": 0.0965,
+      "loss": 0.1113,
       "step": 1482
     },
     {
       "epoch": 3.33,
       "learning_rate": 0.00013023121387283236,
-      "loss": 0.1862,
+      "loss": 0.1629,
       "step": 1483
     },
     {
       "epoch": 3.33,
       "learning_rate": 0.00013005780346820808,
-      "loss": 0.1533,
+      "loss": 0.1464,
       "step": 1484
     },
     {
       "epoch": 3.33,
       "learning_rate": 0.0001298843930635838,
-      "loss": 0.1787,
+      "loss": 0.1699,
       "step": 1485
     },
     {
       "epoch": 3.33,
       "learning_rate": 0.00012971098265895952,
-      "loss": 0.2023,
+      "loss": 0.1579,
       "step": 1486
     },
     {
       "epoch": 3.33,
       "learning_rate": 0.00012953757225433524,
-      "loss": 0.157,
+      "loss": 0.1498,
       "step": 1487
     },
     {
       "epoch": 3.34,
       "learning_rate": 0.00012936416184971096,
-      "loss": 0.154,
+      "loss": 0.1559,
       "step": 1488
     },
     {
       "epoch": 3.34,
       "learning_rate": 0.00012919075144508669,
-      "loss": 0.3683,
+      "loss": 0.4199,
       "step": 1489
     },
     {
       "epoch": 3.34,
       "learning_rate": 0.0001290173410404624,
-      "loss": 0.2544,
+      "loss": 0.2768,
       "step": 1490
     },
     {
       "epoch": 3.34,
       "learning_rate": 0.00012884393063583813,
-      "loss": 0.2392,
+      "loss": 0.2245,
       "step": 1491
     },
     {
       "epoch": 3.35,
       "learning_rate": 0.00012867052023121387,
-      "loss": 0.227,
+      "loss": 0.2052,
       "step": 1492
     },
     {
       "epoch": 3.35,
       "learning_rate": 0.00012849710982658957,
-      "loss": 0.1851,
+      "loss": 0.1749,
       "step": 1493
     },
     {
       "epoch": 3.35,
       "learning_rate": 0.00012832369942196532,
-      "loss": 0.1939,
+      "loss": 0.1611,
       "step": 1494
     },
     {
       "epoch": 3.35,
       "learning_rate": 0.00012815028901734104,
-      "loss": 0.2472,
+      "loss": 0.2097,
       "step": 1495
     },
     {
       "epoch": 3.35,
       "learning_rate": 0.00012797687861271676,
-      "loss": 0.1952,
+      "loss": 0.1994,
       "step": 1496
     },
     {
       "epoch": 3.36,
       "learning_rate": 0.00012780346820809248,
-      "loss": 0.1977,
+      "loss": 0.1656,
       "step": 1497
     },
     {
       "epoch": 3.36,
       "learning_rate": 0.0001276300578034682,
-      "loss": 0.2138,
+      "loss": 0.1881,
       "step": 1498
     },
     {
       "epoch": 3.36,
       "learning_rate": 0.00012745664739884392,
-      "loss": 0.2302,
+      "loss": 0.214,
       "step": 1499
     },
     {
       "epoch": 3.36,
       "learning_rate": 0.00012728323699421964,
-      "loss": 0.1376,
+      "loss": 0.1277,
       "step": 1500
     },
     {
       "epoch": 3.36,
-      "eval_loss": 0.33944663405418396,
-      "eval_runtime": 570.7817,
-      "eval_samples_per_second": 4.629,
-      "eval_steps_per_second": 0.58,
-      "eval_wer": 0.10464101547005157,
+      "eval_loss": 0.34972885251045227,
+      "eval_runtime": 555.1742,
+      "eval_samples_per_second": 4.759,
+      "eval_steps_per_second": 0.596,
+      "eval_wer": 0.1040063466878223,
       "step": 1500
     },
     {
       "epoch": 3.37,
       "learning_rate": 0.00012710982658959536,
-      "loss": 0.1535,
+      "loss": 0.1352,
       "step": 1501
     },
     {
       "epoch": 3.37,
       "learning_rate": 0.00012693641618497108,
-      "loss": 0.159,
+      "loss": 0.1341,
       "step": 1502
     },
     {
       "epoch": 3.37,
       "learning_rate": 0.0001267630057803468,
-      "loss": 0.1576,
+      "loss": 0.1614,
       "step": 1503
     },
     {
       "epoch": 3.37,
       "learning_rate": 0.00012658959537572252,
-      "loss": 0.1656,
+      "loss": 0.1356,
       "step": 1504
     },
     {
       "epoch": 3.37,
       "learning_rate": 0.00012641618497109824,
-      "loss": 0.1746,
+      "loss": 0.1486,
       "step": 1505
     },
     {
       "epoch": 3.38,
       "learning_rate": 0.000126242774566474,
-      "loss": 0.1503,
+      "loss": 0.1557,
       "step": 1506
     },
     {
       "epoch": 3.38,
       "learning_rate": 0.00012606936416184968,
-      "loss": 0.1322,
+      "loss": 0.1015,
       "step": 1507
     },
     {
       "epoch": 3.38,
       "learning_rate": 0.00012589595375722543,
-      "loss": 0.1634,
+      "loss": 0.1544,
       "step": 1508
     },
     {
       "epoch": 3.38,
       "learning_rate": 0.00012572254335260115,
-      "loss": 0.1521,
+      "loss": 0.1357,
       "step": 1509
     },
     {
       "epoch": 3.39,
       "learning_rate": 0.00012554913294797687,
-      "loss": 0.1227,
+      "loss": 0.1212,
       "step": 1510
     },
     {
       "epoch": 3.39,
       "learning_rate": 0.0001253757225433526,
-      "loss": 0.1433,
+      "loss": 0.1688,
       "step": 1511
     },
     {
       "epoch": 3.39,
       "learning_rate": 0.00012520231213872831,
-      "loss": 0.1242,
+      "loss": 0.1234,
       "step": 1512
     },
     {
       "epoch": 3.39,
       "learning_rate": 0.00012502890173410404,
-      "loss": 0.1352,
+      "loss": 0.1307,
       "step": 1513
     },
     {
       "epoch": 3.39,
       "learning_rate": 0.00012485549132947976,
-      "loss": 0.1243,
+      "loss": 0.1111,
       "step": 1514
     },
     {
       "epoch": 3.4,
       "learning_rate": 0.00012468208092485548,
-      "loss": 0.1467,
+      "loss": 0.133,
       "step": 1515
     },
     {
       "epoch": 3.4,
       "learning_rate": 0.0001245086705202312,
-      "loss": 0.1349,
+      "loss": 0.1339,
       "step": 1516
     },
     {
       "epoch": 3.4,
       "learning_rate": 0.00012433526011560692,
-      "loss": 0.1183,
+      "loss": 0.1395,
       "step": 1517
     },
     {
       "epoch": 3.4,
       "learning_rate": 0.00012416184971098267,
-      "loss": 0.1427,
+      "loss": 0.1324,
       "step": 1518
     },
     {
       "epoch": 3.41,
       "learning_rate": 0.00012398843930635836,
-      "loss": 0.1034,
+      "loss": 0.0869,
       "step": 1519
     },
     {
       "epoch": 3.41,
       "learning_rate": 0.0001238150289017341,
-      "loss": 0.1718,
+      "loss": 0.1488,
       "step": 1520
     },
     {
       "epoch": 3.41,
       "learning_rate": 0.0001236416184971098,
-      "loss": 0.1485,
+      "loss": 0.129,
       "step": 1521
     },
     {
       "epoch": 3.41,
       "learning_rate": 0.00012346820809248555,
-      "loss": 0.1333,
+      "loss": 0.1255,
       "step": 1522
     },
     {
       "epoch": 3.41,
       "learning_rate": 0.00012329479768786127,
-      "loss": 0.149,
+      "loss": 0.1568,
       "step": 1523
     },
     {
       "epoch": 3.42,
       "learning_rate": 0.000123121387283237,
-      "loss": 0.1364,
+      "loss": 0.1269,
       "step": 1524
     },
     {
       "epoch": 3.42,
       "learning_rate": 0.0001229479768786127,
-      "loss": 0.1184,
+      "loss": 0.0831,
       "step": 1525
     },
     {
       "epoch": 3.42,
       "learning_rate": 0.00012277456647398843,
-      "loss": 0.1482,
+      "loss": 0.1454,
       "step": 1526
     },
     {
       "epoch": 3.42,
       "learning_rate": 0.00012260115606936415,
-      "loss": 0.1418,
+      "loss": 0.1484,
       "step": 1527
     },
     {
       "epoch": 3.43,
       "learning_rate": 0.00012242774566473987,
-      "loss": 0.1314,
+      "loss": 0.1237,
       "step": 1528
     },
     {
       "epoch": 3.43,
       "learning_rate": 0.0001222543352601156,
-      "loss": 0.1829,
+      "loss": 0.1523,
       "step": 1529
     },
     {
       "epoch": 3.43,
       "learning_rate": 0.00012208092485549131,
-      "loss": 0.1409,
+      "loss": 0.1167,
       "step": 1530
     },
     {
       "epoch": 3.43,
       "learning_rate": 0.00012190751445086703,
-      "loss": 0.1908,
+      "loss": 0.1941,
       "step": 1531
     },
     {
       "epoch": 3.43,
       "learning_rate": 0.00012173410404624277,
-      "loss": 0.1689,
+      "loss": 0.1381,
       "step": 1532
     },
     {
       "epoch": 3.44,
       "learning_rate": 0.00012156069364161849,
-      "loss": 0.1731,
+      "loss": 0.1792,
       "step": 1533
     },
     {
       "epoch": 3.44,
       "learning_rate": 0.00012138728323699421,
-      "loss": 0.1157,
+      "loss": 0.0959,
       "step": 1534
     },
     {
       "epoch": 3.44,
       "learning_rate": 0.00012121387283236993,
-      "loss": 0.1351,
+      "loss": 0.1264,
       "step": 1535
     },
     {
       "epoch": 3.44,
       "learning_rate": 0.00012104046242774567,
-      "loss": 0.1151,
+      "loss": 0.0949,
       "step": 1536
     },
     {
       "epoch": 3.45,
       "learning_rate": 0.00012086705202312137,
-      "loss": 0.1389,
+      "loss": 0.1213,
       "step": 1537
     },
     {
       "epoch": 3.45,
       "learning_rate": 0.0001206936416184971,
-      "loss": 0.1776,
+      "loss": 0.2097,
       "step": 1538
     },
     {
       "epoch": 3.45,
       "learning_rate": 0.00012052023121387281,
-      "loss": 0.3006,
+      "loss": 0.292,
       "step": 1539
     },
     {
       "epoch": 3.45,
       "learning_rate": 0.00012034682080924855,
-      "loss": 0.3316,
+      "loss": 0.2704,
       "step": 1540
     },
     {
       "epoch": 3.46,
       "learning_rate": 0.00012017341040462427,
-      "loss": 0.2327,
+      "loss": 0.2344,
       "step": 1541
     },
     {
       "epoch": 3.46,
       "learning_rate": 0.00011999999999999999,
-      "loss": 0.2735,
+      "loss": 0.2288,
       "step": 1542
     },
     {
       "epoch": 3.46,
       "learning_rate": 0.00011982658959537571,
-      "loss": 0.2421,
+      "loss": 0.2087,
       "step": 1543
     },
     {
       "epoch": 3.46,
       "learning_rate": 0.00011965317919075144,
-      "loss": 0.2247,
+      "loss": 0.2058,
       "step": 1544
     },
     {
       "epoch": 3.46,
       "learning_rate": 0.00011947976878612715,
-      "loss": 0.2077,
+      "loss": 0.173,
       "step": 1545
     },
     {
       "epoch": 3.47,
       "learning_rate": 0.00011930635838150289,
-      "loss": 0.1836,
+      "loss": 0.1584,
       "step": 1546
     },
     {
       "epoch": 3.47,
       "learning_rate": 0.0001191329479768786,
-      "loss": 0.2034,
+      "loss": 0.2082,
       "step": 1547
     },
     {
       "epoch": 3.47,
       "learning_rate": 0.00011895953757225433,
-      "loss": 0.1808,
+      "loss": 0.1624,
       "step": 1548
     },
     {
       "epoch": 3.47,
       "learning_rate": 0.00011878612716763005,
-      "loss": 0.1652,
+      "loss": 0.1681,
       "step": 1549
     },
     {
       "epoch": 3.48,
       "learning_rate": 0.00011861271676300578,
-      "loss": 0.1875,
+      "loss": 0.1637,
       "step": 1550
     },
     {
       "epoch": 3.48,
       "learning_rate": 0.00011843930635838149,
-      "loss": 0.146,
+      "loss": 0.1455,
       "step": 1551
     },
     {
       "epoch": 3.48,
       "learning_rate": 0.00011826589595375722,
-      "loss": 0.1162,
+      "loss": 0.1189,
       "step": 1552
     },
     {
       "epoch": 3.48,
       "learning_rate": 0.00011809248554913293,
-      "loss": 0.17,
+      "loss": 0.1575,
       "step": 1553
     },
     {
       "epoch": 3.48,
       "learning_rate": 0.00011791907514450866,
-      "loss": 0.1839,
+      "loss": 0.1567,
       "step": 1554
     },
     {
       "epoch": 3.49,
       "learning_rate": 0.00011774566473988439,
-      "loss": 0.1617,
+      "loss": 0.1484,
       "step": 1555
     },
     {
       "epoch": 3.49,
       "learning_rate": 0.00011757225433526012,
-      "loss": 0.1352,
+      "loss": 0.1359,
       "step": 1556
     },
     {
       "epoch": 3.49,
       "learning_rate": 0.00011739884393063583,
-      "loss": 0.1274,
+      "loss": 0.1224,
       "step": 1557
     },
     {
       "epoch": 3.49,
       "learning_rate": 0.00011722543352601156,
-      "loss": 0.1727,
+      "loss": 0.1439,
       "step": 1558
     },
     {
       "epoch": 3.5,
       "learning_rate": 0.00011705202312138727,
-      "loss": 0.1377,
+      "loss": 0.1238,
       "step": 1559
     },
     {
       "epoch": 3.5,
       "learning_rate": 0.000116878612716763,
-      "loss": 0.1351,
+      "loss": 0.1421,
       "step": 1560
     },
     {
       "epoch": 3.5,
       "learning_rate": 0.00011670520231213872,
-      "loss": 0.1326,
+      "loss": 0.1541,
       "step": 1561
     },
     {
       "epoch": 3.5,
       "learning_rate": 0.00011653179190751443,
-      "loss": 0.1246,
+      "loss": 0.1132,
       "step": 1562
     },
     {
       "epoch": 3.5,
       "learning_rate": 0.00011635838150289016,
-      "loss": 0.1424,
+      "loss": 0.136,
       "step": 1563
     },
     {
       "epoch": 3.51,
       "learning_rate": 0.00011618497109826587,
-      "loss": 0.1224,
+      "loss": 0.134,
       "step": 1564
     },
     {
       "epoch": 3.51,
       "learning_rate": 0.0001160115606936416,
-      "loss": 0.1188,
+      "loss": 0.1325,
       "step": 1565
     },
     {
       "epoch": 3.51,
       "learning_rate": 0.00011583815028901733,
-      "loss": 0.1331,
+      "loss": 0.1259,
       "step": 1566
     },
     {
       "epoch": 3.51,
       "learning_rate": 0.00011566473988439306,
-      "loss": 0.1422,
+      "loss": 0.16,
       "step": 1567
     },
     {
       "epoch": 3.52,
       "learning_rate": 0.00011549132947976877,
-      "loss": 0.1426,
+      "loss": 0.1253,
       "step": 1568
     },
     {
       "epoch": 3.52,
       "learning_rate": 0.0001153179190751445,
-      "loss": 0.1677,
+      "loss": 0.1421,
       "step": 1569
     },
     {
       "epoch": 3.52,
       "learning_rate": 0.00011514450867052021,
-      "loss": 0.1393,
+      "loss": 0.1387,
       "step": 1570
     },
     {
       "epoch": 3.52,
       "learning_rate": 0.00011497109826589594,
-      "loss": 0.1264,
+      "loss": 0.1289,
       "step": 1571
     },
     {
       "epoch": 3.52,
       "learning_rate": 0.00011479768786127166,
-      "loss": 0.154,
+      "loss": 0.1405,
       "step": 1572
     },
     {
       "epoch": 3.53,
       "learning_rate": 0.00011462427745664738,
-      "loss": 0.2037,
+      "loss": 0.171,
       "step": 1573
     },
     {
       "epoch": 3.53,
       "learning_rate": 0.0001144508670520231,
-      "loss": 0.1182,
+      "loss": 0.1136,
       "step": 1574
     },
     {
       "epoch": 3.53,
       "learning_rate": 0.00011427745664739884,
-      "loss": 0.1106,
+      "loss": 0.1325,
       "step": 1575
     },
     {
       "epoch": 3.53,
       "learning_rate": 0.00011410404624277455,
-      "loss": 0.1191,
+      "loss": 0.0906,
       "step": 1576
     },
     {
       "epoch": 3.54,
       "learning_rate": 0.00011393063583815028,
-      "loss": 0.1119,
+      "loss": 0.0978,
       "step": 1577
     },
     {
       "epoch": 3.54,
       "learning_rate": 0.00011375722543352599,
-      "loss": 0.1422,
+      "loss": 0.1221,
       "step": 1578
     },
     {
       "epoch": 3.54,
       "learning_rate": 0.00011358381502890172,
-      "loss": 0.1839,
+      "loss": 0.2028,
       "step": 1579
     },
     {
       "epoch": 3.54,
       "learning_rate": 0.00011341040462427744,
-      "loss": 0.1446,
+      "loss": 0.131,
       "step": 1580
     },
     {
       "epoch": 3.54,
       "learning_rate": 0.00011323699421965318,
-      "loss": 0.1155,
+      "loss": 0.1039,
       "step": 1581
     },
     {
       "epoch": 3.55,
       "learning_rate": 0.00011306358381502888,
-      "loss": 0.1447,
+      "loss": 0.1327,
       "step": 1582
     },
     {
       "epoch": 3.55,
       "learning_rate": 0.00011289017341040462,
-      "loss": 0.0928,
+      "loss": 0.0906,
       "step": 1583
     },
     {
       "epoch": 3.55,
       "learning_rate": 0.00011271676300578033,
-      "loss": 0.1752,
+      "loss": 0.1549,
       "step": 1584
     },
     {
       "epoch": 3.55,
       "learning_rate": 0.00011254335260115606,
-      "loss": 0.1932,
+      "loss": 0.1482,
       "step": 1585
     },
     {
       "epoch": 3.56,
       "learning_rate": 0.00011236994219653178,
-      "loss": 0.0998,
+      "loss": 0.0944,
       "step": 1586
     },
     {
       "epoch": 3.56,
       "learning_rate": 0.0001121965317919075,
-      "loss": 0.0839,
+      "loss": 0.0893,
       "step": 1587
     },
     {
       "epoch": 3.56,
       "learning_rate": 0.00011202312138728322,
-      "loss": 0.1495,
+      "loss": 0.1269,
       "step": 1588
     },
     {
       "epoch": 3.56,
       "learning_rate": 0.00011184971098265896,
-      "loss": 0.318,
+      "loss": 0.3062,
       "step": 1589
     },
     {
       "epoch": 3.57,
       "learning_rate": 0.00011167630057803466,
-      "loss": 0.249,
+      "loss": 0.2444,
       "step": 1590
     },
     {
       "epoch": 3.57,
       "learning_rate": 0.0001115028901734104,
-      "loss": 0.2753,
+      "loss": 0.2537,
       "step": 1591
     },
     {
       "epoch": 3.57,
       "learning_rate": 0.00011132947976878612,
-      "loss": 0.1983,
+      "loss": 0.1882,
       "step": 1592
     },
     {
       "epoch": 3.57,
       "learning_rate": 0.00011115606936416184,
-      "loss": 0.2503,
+      "loss": 0.2053,
       "step": 1593
     },
     {
       "epoch": 3.57,
       "learning_rate": 0.00011098265895953756,
-      "loss": 0.2084,
+      "loss": 0.192,
       "step": 1594
     },
     {
       "epoch": 3.58,
       "learning_rate": 0.0001108092485549133,
-      "loss": 0.2058,
+      "loss": 0.1852,
       "step": 1595
     },
     {
       "epoch": 3.58,
       "learning_rate": 0.000110635838150289,
-      "loss": 0.1842,
+      "loss": 0.1857,
       "step": 1596
     },
     {
       "epoch": 3.58,
       "learning_rate": 0.00011046242774566474,
-      "loss": 0.1766,
+      "loss": 0.1598,
       "step": 1597
     },
     {
       "epoch": 3.58,
       "learning_rate": 0.00011028901734104044,
-      "loss": 0.1668,
+      "loss": 0.1605,
       "step": 1598
     },
     {
       "epoch": 3.59,
       "learning_rate": 0.00011011560693641618,
-      "loss": 0.175,
+      "loss": 0.1873,
       "step": 1599
     },
     {
       "epoch": 3.59,
       "learning_rate": 0.0001099421965317919,
-      "loss": 0.1449,
+      "loss": 0.1568,
       "step": 1600
     },
     {
       "epoch": 3.59,
       "learning_rate": 0.00010976878612716762,
-      "loss": 0.1336,
+      "loss": 0.1084,
       "step": 1601
     },
     {
       "epoch": 3.59,
       "learning_rate": 0.00010959537572254334,
-      "loss": 0.1672,
+      "loss": 0.1516,
       "step": 1602
     },
     {
       "epoch": 3.59,
       "learning_rate": 0.00010942196531791907,
-      "loss": 0.1915,
+      "loss": 0.1708,
       "step": 1603
     },
     {
       "epoch": 3.6,
       "learning_rate": 0.00010924855491329478,
-      "loss": 0.1526,
+      "loss": 0.1408,
       "step": 1604
     },
     {
       "epoch": 3.6,
       "learning_rate": 0.00010907514450867051,
-      "loss": 0.1468,
+      "loss": 0.1296,
       "step": 1605
     },
     {
       "epoch": 3.6,
       "learning_rate": 0.00010890173410404623,
-      "loss": 0.1434,
+      "loss": 0.1297,
       "step": 1606
     },
     {
       "epoch": 3.6,
       "learning_rate": 0.00010872832369942196,
-      "loss": 0.1468,
+      "loss": 0.1442,
       "step": 1607
     },
     {
       "epoch": 3.61,
       "learning_rate": 0.00010855491329479768,
-      "loss": 0.126,
+      "loss": 0.1224,
       "step": 1608
     },
     {
       "epoch": 3.61,
       "learning_rate": 0.00010838150289017341,
-      "loss": 0.1342,
+      "loss": 0.1211,
       "step": 1609
     },
     {
       "epoch": 3.61,
       "learning_rate": 0.00010820809248554912,
-      "loss": 0.1391,
+      "loss": 0.1337,
       "step": 1610
     },
     {
       "epoch": 3.61,
       "learning_rate": 0.00010803468208092485,
-      "loss": 0.1658,
+      "loss": 0.1649,
       "step": 1611
     },
     {
       "epoch": 3.61,
       "learning_rate": 0.00010786127167630056,
-      "loss": 0.1385,
+      "loss": 0.128,
       "step": 1612
     },
     {
       "epoch": 3.62,
       "learning_rate": 0.00010768786127167629,
-      "loss": 0.156,
+      "loss": 0.1535,
       "step": 1613
     },
     {
       "epoch": 3.62,
       "learning_rate": 0.00010751445086705201,
-      "loss": 0.1532,
+      "loss": 0.1381,
       "step": 1614
     },
     {
       "epoch": 3.62,
       "learning_rate": 0.00010734104046242773,
-      "loss": 0.1125,
+      "loss": 0.1441,
       "step": 1615
     },
     {
@@ -9733,367 +9733,367 @@
     {
       "epoch": 3.63,
       "learning_rate": 0.00010699421965317919,
-      "loss": 0.1338,
+      "loss": 0.1301,
       "step": 1617
     },
     {
       "epoch": 3.63,
       "learning_rate": 0.0001068208092485549,
-      "loss": 0.1572,
+      "loss": 0.1528,
       "step": 1618
     },
     {
       "epoch": 3.63,
       "learning_rate": 0.00010664739884393063,
-      "loss": 0.1134,
+      "loss": 0.1144,
       "step": 1619
     },
     {
       "epoch": 3.63,
       "learning_rate": 0.00010647398843930635,
-      "loss": 0.1702,
+      "loss": 0.1606,
       "step": 1620
     },
     {
       "epoch": 3.63,
       "learning_rate": 0.00010630057803468207,
-      "loss": 0.1197,
+      "loss": 0.4729,
       "step": 1621
     },
     {
       "epoch": 3.64,
       "learning_rate": 0.00010612716763005779,
-      "loss": 0.1265,
+      "loss": 0.127,
       "step": 1622
     },
     {
       "epoch": 3.64,
       "learning_rate": 0.00010595375722543353,
-      "loss": 0.1377,
+      "loss": 0.1248,
       "step": 1623
     },
     {
       "epoch": 3.64,
       "learning_rate": 0.00010578034682080923,
-      "loss": 0.1339,
+      "loss": 0.1386,
       "step": 1624
     },
     {
       "epoch": 3.64,
       "learning_rate": 0.00010560693641618497,
-      "loss": 0.1696,
+      "loss": 0.1429,
       "step": 1625
     },
     {
       "epoch": 3.65,
       "learning_rate": 0.00010543352601156068,
-      "loss": 0.1004,
+      "loss": 0.099,
       "step": 1626
     },
     {
       "epoch": 3.65,
       "learning_rate": 0.00010526011560693641,
-      "loss": 0.1684,
+      "loss": 0.148,
       "step": 1627
     },
     {
       "epoch": 3.65,
       "learning_rate": 0.00010508670520231213,
-      "loss": 0.1417,
+      "loss": 0.1268,
       "step": 1628
     },
     {
       "epoch": 3.65,
       "learning_rate": 0.00010491329479768786,
-      "loss": 0.1377,
+      "loss": 0.1344,
       "step": 1629
     },
     {
       "epoch": 3.65,
       "learning_rate": 0.00010473988439306357,
-      "loss": 0.1299,
+      "loss": 0.1011,
       "step": 1630
     },
     {
       "epoch": 3.66,
       "learning_rate": 0.0001045664739884393,
-      "loss": 0.1311,
+      "loss": 0.0999,
       "step": 1631
     },
     {
       "epoch": 3.66,
       "learning_rate": 0.00010439306358381501,
-      "loss": 0.1528,
+      "loss": 0.1456,
       "step": 1632
     },
     {
       "epoch": 3.66,
       "learning_rate": 0.00010421965317919075,
-      "loss": 0.1332,
+      "loss": 0.1284,
       "step": 1633
     },
     {
       "epoch": 3.66,
       "learning_rate": 0.00010404624277456647,
-      "loss": 0.1222,
+      "loss": 0.126,
       "step": 1634
     },
     {
       "epoch": 3.67,
       "learning_rate": 0.00010387283236994219,
-      "loss": 0.0947,
+      "loss": 0.1044,
       "step": 1635
     },
     {
       "epoch": 3.67,
       "learning_rate": 0.00010369942196531791,
-      "loss": 0.1681,
+      "loss": 0.1651,
       "step": 1636
     },
     {
       "epoch": 3.67,
       "learning_rate": 0.00010352601156069364,
-      "loss": 0.1594,
+      "loss": 0.1435,
       "step": 1637
     },
     {
       "epoch": 3.67,
       "learning_rate": 0.00010335260115606935,
-      "loss": 0.2108,
+      "loss": 0.2565,
       "step": 1638
     },
     {
       "epoch": 3.67,
       "learning_rate": 0.00010317919075144509,
-      "loss": 0.3022,
+      "loss": 0.2916,
       "step": 1639
     },
     {
       "epoch": 3.68,
       "learning_rate": 0.00010300578034682079,
-      "loss": 0.2142,
+      "loss": 0.1893,
       "step": 1640
     },
     {
       "epoch": 3.68,
       "learning_rate": 0.00010283236994219653,
-      "loss": 0.2192,
+      "loss": 0.202,
       "step": 1641
     },
     {
       "epoch": 3.68,
       "learning_rate": 0.00010265895953757225,
-      "loss": 0.2243,
+      "loss": 0.2011,
       "step": 1642
     },
     {
       "epoch": 3.68,
       "learning_rate": 0.00010248554913294798,
-      "loss": 0.1595,
+      "loss": 0.1397,
       "step": 1643
     },
     {
       "epoch": 3.69,
       "learning_rate": 0.00010231213872832369,
-      "loss": 0.229,
+      "loss": 0.1991,
       "step": 1644
     },
     {
       "epoch": 3.69,
       "learning_rate": 0.00010213872832369942,
-      "loss": 0.1601,
+      "loss": 0.143,
       "step": 1645
     },
     {
       "epoch": 3.69,
       "learning_rate": 0.00010196531791907513,
-      "loss": 0.176,
+      "loss": 0.181,
       "step": 1646
     },
     {
       "epoch": 3.69,
       "learning_rate": 0.00010179190751445086,
-      "loss": 0.1696,
+      "loss": 0.1497,
       "step": 1647
     },
     {
       "epoch": 3.7,
       "learning_rate": 0.00010161849710982658,
-      "loss": 0.2211,
+      "loss": 0.2158,
       "step": 1648
     },
     {
       "epoch": 3.7,
       "learning_rate": 0.0001014450867052023,
-      "loss": 0.2021,
+      "loss": 0.1534,
       "step": 1649
     },
     {
       "epoch": 3.7,
       "learning_rate": 0.00010127167630057803,
-      "loss": 0.1745,
+      "loss": 0.16,
       "step": 1650
     },
     {
       "epoch": 3.7,
       "learning_rate": 0.00010109826589595376,
-      "loss": 0.1634,
+      "loss": 0.1698,
       "step": 1651
     },
     {
       "epoch": 3.7,
       "learning_rate": 0.00010092485549132947,
-      "loss": 0.1452,
+      "loss": 0.1082,
       "step": 1652
     },
     {
       "epoch": 3.71,
       "learning_rate": 0.0001007514450867052,
-      "loss": 0.1467,
+      "loss": 0.1202,
       "step": 1653
     },
     {
       "epoch": 3.71,
       "learning_rate": 0.00010057803468208092,
-      "loss": 0.1406,
+      "loss": 0.1296,
       "step": 1654
     },
     {
       "epoch": 3.71,
       "learning_rate": 0.00010040462427745664,
-      "loss": 0.1217,
+      "loss": 0.1141,
       "step": 1655
     },
     {
       "epoch": 3.71,
       "learning_rate": 0.00010023121387283236,
-      "loss": 0.1009,
+      "loss": 0.1072,
       "step": 1656
     },
     {
       "epoch": 3.72,
       "learning_rate": 0.0001000578034682081,
-      "loss": 0.1508,
+      "loss": 0.1284,
       "step": 1657
     },
     {
       "epoch": 3.72,
       "learning_rate": 9.98843930635838e-05,
-      "loss": 0.138,
+      "loss": 0.1583,
       "step": 1658
     },
     {
       "epoch": 3.72,
       "learning_rate": 9.971098265895953e-05,
-      "loss": 0.1139,
+      "loss": 0.1118,
       "step": 1659
     },
     {
       "epoch": 3.72,
       "learning_rate": 9.953757225433525e-05,
-      "loss": 0.1276,
+      "loss": 0.1163,
       "step": 1660
     },
     {
       "epoch": 3.72,
       "learning_rate": 9.936416184971097e-05,
-      "loss": 0.0977,
+      "loss": 0.0947,
       "step": 1661
     },
     {
       "epoch": 3.73,
       "learning_rate": 9.91907514450867e-05,
-      "loss": 0.1013,
+      "loss": 0.097,
       "step": 1662
     },
     {
       "epoch": 3.73,
       "learning_rate": 9.901734104046241e-05,
-      "loss": 0.096,
+      "loss": 0.1164,
       "step": 1663
     },
     {
       "epoch": 3.73,
       "learning_rate": 9.884393063583814e-05,
-      "loss": 0.135,
+      "loss": 0.1529,
       "step": 1664
     },
     {
       "epoch": 3.73,
       "learning_rate": 9.867052023121385e-05,
-      "loss": 0.1351,
+      "loss": 0.1325,
       "step": 1665
     },
     {
       "epoch": 3.74,
       "learning_rate": 9.849710982658958e-05,
-      "loss": 0.1233,
+      "loss": 0.1309,
       "step": 1666
     },
     {
       "epoch": 3.74,
       "learning_rate": 9.83236994219653e-05,
-      "loss": 0.1614,
+      "loss": 0.1572,
       "step": 1667
     },
     {
       "epoch": 3.74,
       "learning_rate": 9.815028901734104e-05,
-      "loss": 0.1256,
+      "loss": 0.1051,
       "step": 1668
     },
     {
       "epoch": 3.74,
       "learning_rate": 9.797687861271675e-05,
-      "loss": 0.17,
+      "loss": 0.1572,
       "step": 1669
     },
     {
       "epoch": 3.74,
       "learning_rate": 9.780346820809248e-05,
-      "loss": 0.128,
+      "loss": 0.1192,
       "step": 1670
     },
     {
       "epoch": 3.75,
       "learning_rate": 9.763005780346819e-05,
-      "loss": 0.0963,
+      "loss": 0.1136,
       "step": 1671
     },
     {
       "epoch": 3.75,
       "learning_rate": 9.745664739884392e-05,
-      "loss": 0.1002,
+      "loss": 0.1003,
       "step": 1672
     },
     {
       "epoch": 3.75,
       "learning_rate": 9.728323699421964e-05,
-      "loss": 0.1219,
+      "loss": 0.1195,
       "step": 1673
     },
     {
       "epoch": 3.75,
       "learning_rate": 9.710982658959536e-05,
-      "loss": 0.1356,
+      "loss": 0.1246,
       "step": 1674
     },
     {
       "epoch": 3.76,
       "learning_rate": 9.693641618497108e-05,
-      "loss": 0.1512,
+      "loss": 0.1734,
       "step": 1675
     },
     {
       "epoch": 3.76,
       "learning_rate": 9.676300578034682e-05,
-      "loss": 0.1085,
+      "loss": 0.1257,
       "step": 1676
     },
     {
       "epoch": 3.76,
       "learning_rate": 9.658959537572253e-05,
-      "loss": 0.1254,
+      "loss": 0.122,
       "step": 1677
     },
     {
@@ -10105,1954 +10105,1954 @@
     {
       "epoch": 3.76,
       "learning_rate": 9.624277456647398e-05,
-      "loss": 0.1117,
+      "loss": 0.113,
       "step": 1679
     },
     {
       "epoch": 3.77,
       "learning_rate": 9.60693641618497e-05,
-      "loss": 0.117,
+      "loss": 0.1115,
       "step": 1680
     },
     {
       "epoch": 3.77,
       "learning_rate": 9.589595375722542e-05,
-      "loss": 0.1922,
+      "loss": 0.1696,
       "step": 1681
     },
     {
       "epoch": 3.77,
       "learning_rate": 9.572254335260116e-05,
-      "loss": 0.1505,
+      "loss": 0.1251,
       "step": 1682
     },
     {
       "epoch": 3.77,
       "learning_rate": 9.554913294797686e-05,
-      "loss": 0.1026,
+      "loss": 0.0788,
       "step": 1683
     },
     {
       "epoch": 3.78,
       "learning_rate": 9.53757225433526e-05,
-      "loss": 0.0952,
+      "loss": 0.0783,
       "step": 1684
     },
     {
       "epoch": 3.78,
       "learning_rate": 9.52023121387283e-05,
-      "loss": 0.1228,
+      "loss": 0.0956,
       "step": 1685
     },
     {
       "epoch": 3.78,
       "learning_rate": 9.502890173410404e-05,
-      "loss": 0.1412,
+      "loss": 0.1233,
       "step": 1686
     },
     {
       "epoch": 3.78,
       "learning_rate": 9.485549132947976e-05,
-      "loss": 0.1445,
+      "loss": 0.1435,
       "step": 1687
     },
     {
       "epoch": 3.78,
       "learning_rate": 9.468208092485548e-05,
-      "loss": 0.2098,
+      "loss": 0.2561,
       "step": 1688
     },
     {
       "epoch": 3.79,
       "learning_rate": 9.45086705202312e-05,
-      "loss": 0.3423,
+      "loss": 0.308,
       "step": 1689
     },
     {
       "epoch": 3.79,
       "learning_rate": 9.433526011560693e-05,
-      "loss": 0.2938,
+      "loss": 0.2986,
       "step": 1690
     },
     {
       "epoch": 3.79,
       "learning_rate": 9.416184971098264e-05,
-      "loss": 0.2029,
+      "loss": 0.2049,
       "step": 1691
     },
     {
       "epoch": 3.79,
       "learning_rate": 9.398843930635838e-05,
-      "loss": 0.211,
+      "loss": 0.1799,
       "step": 1692
     },
     {
       "epoch": 3.8,
       "learning_rate": 9.38150289017341e-05,
-      "loss": 0.1836,
+      "loss": 0.1759,
       "step": 1693
     },
     {
       "epoch": 3.8,
       "learning_rate": 9.364161849710982e-05,
-      "loss": 0.1827,
+      "loss": 0.1845,
       "step": 1694
     },
     {
       "epoch": 3.8,
       "learning_rate": 9.346820809248554e-05,
-      "loss": 0.2368,
+      "loss": 0.1882,
       "step": 1695
     },
     {
       "epoch": 3.8,
       "learning_rate": 9.329479768786127e-05,
-      "loss": 0.1681,
+      "loss": 0.1517,
       "step": 1696
     },
     {
       "epoch": 3.8,
       "learning_rate": 9.312138728323698e-05,
-      "loss": 0.1423,
+      "loss": 0.1269,
       "step": 1697
     },
     {
       "epoch": 3.81,
       "learning_rate": 9.294797687861271e-05,
-      "loss": 0.1687,
+      "loss": 0.1485,
       "step": 1698
     },
     {
       "epoch": 3.81,
       "learning_rate": 9.277456647398842e-05,
-      "loss": 0.1647,
+      "loss": 0.1608,
       "step": 1699
     },
     {
       "epoch": 3.81,
       "learning_rate": 9.260115606936415e-05,
-      "loss": 0.1315,
+      "loss": 0.1159,
       "step": 1700
     },
     {
       "epoch": 3.81,
       "learning_rate": 9.242774566473988e-05,
-      "loss": 0.1734,
+      "loss": 0.1598,
       "step": 1701
     },
     {
       "epoch": 3.82,
       "learning_rate": 9.22543352601156e-05,
-      "loss": 0.1152,
+      "loss": 0.1181,
       "step": 1702
     },
     {
       "epoch": 3.82,
       "learning_rate": 9.208092485549132e-05,
-      "loss": 0.1555,
+      "loss": 0.1478,
       "step": 1703
     },
     {
       "epoch": 3.82,
       "learning_rate": 9.190751445086705e-05,
-      "loss": 0.102,
+      "loss": 0.0908,
       "step": 1704
     },
     {
       "epoch": 3.82,
       "learning_rate": 9.173410404624276e-05,
-      "loss": 0.1814,
+      "loss": 0.1451,
       "step": 1705
     },
     {
       "epoch": 3.83,
       "learning_rate": 9.156069364161849e-05,
-      "loss": 0.1371,
+      "loss": 0.1336,
       "step": 1706
     },
     {
       "epoch": 3.83,
       "learning_rate": 9.138728323699421e-05,
-      "loss": 0.136,
+      "loss": 0.1112,
       "step": 1707
     },
     {
       "epoch": 3.83,
       "learning_rate": 9.121387283236993e-05,
-      "loss": 0.1517,
+      "loss": 0.1364,
       "step": 1708
     },
     {
       "epoch": 3.83,
       "learning_rate": 9.104046242774565e-05,
-      "loss": 0.1427,
+      "loss": 0.1621,
       "step": 1709
     },
     {
       "epoch": 3.83,
       "learning_rate": 9.086705202312139e-05,
-      "loss": 0.1287,
+      "loss": 0.1221,
       "step": 1710
     },
     {
       "epoch": 3.84,
       "learning_rate": 9.06936416184971e-05,
-      "loss": 0.1701,
+      "loss": 0.1516,
       "step": 1711
     },
     {
       "epoch": 3.84,
       "learning_rate": 9.052023121387283e-05,
-      "loss": 0.1062,
+      "loss": 0.1168,
       "step": 1712
     },
     {
       "epoch": 3.84,
       "learning_rate": 9.034682080924854e-05,
-      "loss": 0.1352,
+      "loss": 0.1298,
       "step": 1713
     },
     {
       "epoch": 3.84,
       "learning_rate": 9.017341040462427e-05,
-      "loss": 0.1358,
+      "loss": 0.1345,
       "step": 1714
     },
     {
       "epoch": 3.85,
       "learning_rate": 8.999999999999999e-05,
-      "loss": 0.1026,
+      "loss": 0.1037,
       "step": 1715
     },
     {
       "epoch": 3.85,
       "learning_rate": 8.982658959537573e-05,
-      "loss": 0.1243,
+      "loss": 0.1118,
       "step": 1716
     },
     {
       "epoch": 3.85,
       "learning_rate": 8.965317919075143e-05,
-      "loss": 0.0982,
+      "loss": 0.1032,
       "step": 1717
     },
     {
       "epoch": 3.85,
       "learning_rate": 8.947976878612717e-05,
-      "loss": 0.1623,
+      "loss": 0.1485,
       "step": 1718
     },
     {
       "epoch": 3.85,
       "learning_rate": 8.930635838150287e-05,
-      "loss": 0.1325,
+      "loss": 0.1301,
       "step": 1719
     },
     {
       "epoch": 3.86,
       "learning_rate": 8.913294797687861e-05,
-      "loss": 0.1237,
+      "loss": 0.126,
       "step": 1720
     },
     {
       "epoch": 3.86,
       "learning_rate": 8.895953757225433e-05,
-      "loss": 0.1091,
+      "loss": 0.1263,
       "step": 1721
     },
     {
       "epoch": 3.86,
       "learning_rate": 8.878612716763005e-05,
-      "loss": 0.1751,
+      "loss": 0.1456,
       "step": 1722
     },
     {
       "epoch": 3.86,
       "learning_rate": 8.861271676300577e-05,
-      "loss": 0.1028,
+      "loss": 0.1013,
       "step": 1723
     },
     {
       "epoch": 3.87,
       "learning_rate": 8.84393063583815e-05,
-      "loss": 0.0774,
+      "loss": 0.0905,
       "step": 1724
     },
     {
       "epoch": 3.87,
       "learning_rate": 8.826589595375721e-05,
-      "loss": 0.1442,
+      "loss": 0.1383,
       "step": 1725
     },
     {
       "epoch": 3.87,
       "learning_rate": 8.809248554913295e-05,
-      "loss": 0.0841,
+      "loss": 0.0949,
       "step": 1726
     },
     {
       "epoch": 3.87,
       "learning_rate": 8.791907514450865e-05,
-      "loss": 0.1029,
+      "loss": 0.1031,
       "step": 1727
     },
     {
       "epoch": 3.87,
       "learning_rate": 8.774566473988439e-05,
-      "loss": 0.1056,
+      "loss": 0.0967,
       "step": 1728
     },
     {
       "epoch": 3.88,
       "learning_rate": 8.757225433526011e-05,
-      "loss": 0.1083,
+      "loss": 0.0964,
       "step": 1729
     },
     {
       "epoch": 3.88,
       "learning_rate": 8.739884393063584e-05,
-      "loss": 0.2944,
+      "loss": 0.1521,
       "step": 1730
     },
     {
       "epoch": 3.88,
       "learning_rate": 8.722543352601155e-05,
-      "loss": 0.1269,
+      "loss": 0.1167,
       "step": 1731
     },
     {
       "epoch": 3.88,
       "learning_rate": 8.705202312138728e-05,
-      "loss": 0.1041,
+      "loss": 0.1063,
       "step": 1732
     },
     {
       "epoch": 3.89,
       "learning_rate": 8.687861271676299e-05,
-      "loss": 0.1044,
+      "loss": 0.1102,
       "step": 1733
     },
     {
       "epoch": 3.89,
       "learning_rate": 8.670520231213873e-05,
-      "loss": 0.1307,
+      "loss": 0.1269,
       "step": 1734
     },
     {
       "epoch": 3.89,
       "learning_rate": 8.653179190751445e-05,
-      "loss": 0.1562,
+      "loss": 0.151,
       "step": 1735
     },
     {
       "epoch": 3.89,
       "learning_rate": 8.635838150289017e-05,
-      "loss": 0.1526,
+      "loss": 0.1504,
       "step": 1736
     },
     {
       "epoch": 3.89,
       "learning_rate": 8.618497109826589e-05,
-      "loss": 0.1579,
+      "loss": 0.1653,
       "step": 1737
     },
     {
       "epoch": 3.9,
       "learning_rate": 8.601156069364162e-05,
-      "loss": 0.1612,
+      "loss": 0.1394,
       "step": 1738
     },
     {
       "epoch": 3.9,
       "learning_rate": 8.583815028901733e-05,
-      "loss": 0.3042,
+      "loss": 0.2872,
       "step": 1739
     },
     {
       "epoch": 3.9,
       "learning_rate": 8.566473988439306e-05,
-      "loss": 0.2292,
+      "loss": 0.2162,
       "step": 1740
     },
     {
       "epoch": 3.9,
       "learning_rate": 8.549132947976878e-05,
-      "loss": 0.208,
+      "loss": 0.1601,
       "step": 1741
     },
     {
       "epoch": 3.91,
       "learning_rate": 8.53179190751445e-05,
-      "loss": 0.2278,
+      "loss": 0.1933,
       "step": 1742
     },
     {
       "epoch": 3.91,
       "learning_rate": 8.514450867052023e-05,
-      "loss": 0.1905,
+      "loss": 0.1663,
       "step": 1743
     },
     {
       "epoch": 3.91,
       "learning_rate": 8.497109826589596e-05,
-      "loss": 0.1668,
+      "loss": 0.1629,
       "step": 1744
     },
     {
       "epoch": 3.91,
       "learning_rate": 8.479768786127167e-05,
-      "loss": 0.1466,
+      "loss": 0.1432,
       "step": 1745
     },
     {
       "epoch": 3.91,
       "learning_rate": 8.46242774566474e-05,
-      "loss": 0.15,
+      "loss": 0.1363,
       "step": 1746
     },
     {
       "epoch": 3.92,
       "learning_rate": 8.445086705202311e-05,
-      "loss": 0.1258,
+      "loss": 0.1099,
       "step": 1747
     },
     {
       "epoch": 3.92,
       "learning_rate": 8.427745664739884e-05,
-      "loss": 0.1983,
+      "loss": 0.1789,
       "step": 1748
     },
     {
       "epoch": 3.92,
       "learning_rate": 8.410404624277456e-05,
-      "loss": 0.1321,
+      "loss": 0.1364,
       "step": 1749
     },
     {
       "epoch": 3.92,
       "learning_rate": 8.393063583815028e-05,
-      "loss": 0.1364,
+      "loss": 0.1443,
       "step": 1750
     },
     {
       "epoch": 3.93,
       "learning_rate": 8.3757225433526e-05,
-      "loss": 0.161,
+      "loss": 0.1386,
       "step": 1751
     },
     {
       "epoch": 3.93,
       "learning_rate": 8.358381502890174e-05,
-      "loss": 0.1493,
+      "loss": 0.1371,
       "step": 1752
     },
     {
       "epoch": 3.93,
       "learning_rate": 8.341040462427745e-05,
-      "loss": 0.1414,
+      "loss": 0.1255,
       "step": 1753
     },
     {
       "epoch": 3.93,
       "learning_rate": 8.323699421965317e-05,
-      "loss": 0.1325,
+      "loss": 0.1153,
       "step": 1754
     },
     {
       "epoch": 3.93,
       "learning_rate": 8.30635838150289e-05,
-      "loss": 0.1344,
+      "loss": 0.1359,
       "step": 1755
     },
     {
       "epoch": 3.94,
       "learning_rate": 8.289017341040461e-05,
-      "loss": 0.1742,
+      "loss": 0.1598,
       "step": 1756
     },
     {
       "epoch": 3.94,
       "learning_rate": 8.271676300578034e-05,
-      "loss": 0.1388,
+      "loss": 0.1175,
       "step": 1757
     },
     {
       "epoch": 3.94,
       "learning_rate": 8.254335260115605e-05,
-      "loss": 0.101,
+      "loss": 0.0973,
       "step": 1758
     },
     {
       "epoch": 3.94,
       "learning_rate": 8.236994219653178e-05,
-      "loss": 0.1336,
+      "loss": 0.1235,
       "step": 1759
     },
     {
       "epoch": 3.95,
       "learning_rate": 8.21965317919075e-05,
-      "loss": 0.1235,
+      "loss": 0.128,
       "step": 1760
     },
     {
       "epoch": 3.95,
       "learning_rate": 8.202312138728322e-05,
-      "loss": 0.1327,
+      "loss": 0.1383,
       "step": 1761
     },
     {
       "epoch": 3.95,
       "learning_rate": 8.184971098265895e-05,
-      "loss": 0.1269,
+      "loss": 0.1162,
       "step": 1762
     },
     {
       "epoch": 3.95,
       "learning_rate": 8.167630057803468e-05,
-      "loss": 0.0936,
+      "loss": 0.1021,
       "step": 1763
     },
     {
       "epoch": 3.96,
       "learning_rate": 8.150289017341039e-05,
-      "loss": 0.1567,
+      "loss": 0.1703,
       "step": 1764
     },
     {
       "epoch": 3.96,
       "learning_rate": 8.132947976878612e-05,
-      "loss": 0.1126,
+      "loss": 0.1103,
       "step": 1765
     },
     {
       "epoch": 3.96,
       "learning_rate": 8.115606936416184e-05,
-      "loss": 0.1776,
+      "loss": 0.1656,
       "step": 1766
     },
     {
       "epoch": 3.96,
       "learning_rate": 8.098265895953756e-05,
-      "loss": 0.1237,
+      "loss": 0.1121,
       "step": 1767
     },
     {
       "epoch": 3.96,
       "learning_rate": 8.080924855491328e-05,
-      "loss": 0.1129,
+      "loss": 0.0983,
       "step": 1768
     },
     {
       "epoch": 3.97,
       "learning_rate": 8.063583815028902e-05,
-      "loss": 0.1201,
+      "loss": 0.1085,
       "step": 1769
     },
     {
       "epoch": 3.97,
       "learning_rate": 8.046242774566472e-05,
-      "loss": 0.1102,
+      "loss": 0.1103,
       "step": 1770
     },
     {
       "epoch": 3.97,
       "learning_rate": 8.028901734104046e-05,
-      "loss": 0.1233,
+      "loss": 0.1093,
       "step": 1771
     },
     {
       "epoch": 3.97,
       "learning_rate": 8.011560693641617e-05,
-      "loss": 0.1111,
+      "loss": 0.097,
       "step": 1772
     },
     {
       "epoch": 3.98,
       "learning_rate": 7.99421965317919e-05,
-      "loss": 0.1067,
+      "loss": 0.102,
       "step": 1773
     },
     {
       "epoch": 3.98,
       "learning_rate": 7.976878612716762e-05,
-      "loss": 0.1084,
+      "loss": 0.1105,
       "step": 1774
     },
     {
       "epoch": 3.98,
       "learning_rate": 7.959537572254334e-05,
-      "loss": 0.1326,
+      "loss": 0.1144,
       "step": 1775
     },
     {
       "epoch": 3.98,
       "learning_rate": 7.942196531791906e-05,
-      "loss": 0.1046,
+      "loss": 0.1052,
       "step": 1776
     },
     {
       "epoch": 3.98,
       "learning_rate": 7.92485549132948e-05,
-      "loss": 0.1212,
+      "loss": 0.1278,
       "step": 1777
     },
     {
       "epoch": 3.99,
       "learning_rate": 7.90751445086705e-05,
-      "loss": 0.1214,
+      "loss": 0.1123,
       "step": 1778
     },
     {
       "epoch": 3.99,
       "learning_rate": 7.890173410404624e-05,
-      "loss": 0.1194,
+      "loss": 0.1038,
       "step": 1779
     },
     {
       "epoch": 3.99,
       "learning_rate": 7.872832369942196e-05,
-      "loss": 0.1105,
+      "loss": 0.1107,
       "step": 1780
     },
     {
       "epoch": 3.99,
       "learning_rate": 7.855491329479768e-05,
-      "loss": 0.1012,
+      "loss": 0.0981,
       "step": 1781
     },
     {
       "epoch": 4.0,
       "learning_rate": 7.83815028901734e-05,
-      "loss": 0.0891,
+      "loss": 0.0696,
       "step": 1782
     },
     {
       "epoch": 4.0,
       "learning_rate": 7.820809248554913e-05,
-      "loss": 0.1084,
+      "loss": 0.0678,
       "step": 1783
     },
     {
       "epoch": 4.0,
       "learning_rate": 7.803468208092484e-05,
-      "loss": 0.137,
+      "loss": 0.1774,
       "step": 1784
     },
     {
       "epoch": 4.0,
       "learning_rate": 7.786127167630058e-05,
-      "loss": 0.1927,
+      "loss": 0.1757,
       "step": 1785
     },
     {
       "epoch": 4.0,
       "learning_rate": 7.768786127167628e-05,
-      "loss": 0.1612,
+      "loss": 0.1199,
       "step": 1786
     },
     {
       "epoch": 4.01,
       "learning_rate": 7.751445086705202e-05,
-      "loss": 0.1435,
+      "loss": 0.1245,
       "step": 1787
     },
     {
       "epoch": 4.01,
       "learning_rate": 7.734104046242774e-05,
-      "loss": 0.1179,
+      "loss": 0.1052,
       "step": 1788
     },
     {
       "epoch": 4.01,
       "learning_rate": 7.716763005780346e-05,
-      "loss": 0.1074,
+      "loss": 0.0766,
       "step": 1789
     },
     {
       "epoch": 4.01,
       "learning_rate": 7.699421965317918e-05,
-      "loss": 0.1491,
+      "loss": 0.1122,
       "step": 1790
     },
     {
       "epoch": 4.02,
       "learning_rate": 7.682080924855491e-05,
-      "loss": 0.1194,
+      "loss": 0.0899,
       "step": 1791
     },
     {
       "epoch": 4.02,
       "learning_rate": 7.664739884393062e-05,
-      "loss": 0.108,
+      "loss": 0.0932,
       "step": 1792
     },
     {
       "epoch": 4.02,
       "learning_rate": 7.647398843930635e-05,
-      "loss": 0.1506,
+      "loss": 0.1327,
       "step": 1793
     },
     {
       "epoch": 4.02,
       "learning_rate": 7.630057803468207e-05,
-      "loss": 0.1111,
+      "loss": 0.1113,
       "step": 1794
     },
     {
       "epoch": 4.02,
       "learning_rate": 7.61271676300578e-05,
-      "loss": 0.079,
+      "loss": 0.087,
       "step": 1795
     },
     {
       "epoch": 4.03,
       "learning_rate": 7.595375722543352e-05,
-      "loss": 0.1138,
+      "loss": 0.0836,
       "step": 1796
     },
     {
       "epoch": 4.03,
       "learning_rate": 7.578034682080925e-05,
-      "loss": 0.0851,
+      "loss": 0.0778,
       "step": 1797
     },
     {
       "epoch": 4.03,
       "learning_rate": 7.560693641618496e-05,
-      "loss": 0.0961,
+      "loss": 0.0863,
       "step": 1798
     },
     {
       "epoch": 4.03,
       "learning_rate": 7.543352601156069e-05,
-      "loss": 0.0884,
+      "loss": 0.083,
       "step": 1799
     },
     {
       "epoch": 4.04,
       "learning_rate": 7.52601156069364e-05,
-      "loss": 0.1127,
+      "loss": 0.1162,
       "step": 1800
     },
     {
       "epoch": 4.04,
       "learning_rate": 7.508670520231213e-05,
-      "loss": 0.0897,
+      "loss": 0.0966,
       "step": 1801
     },
     {
       "epoch": 4.04,
       "learning_rate": 7.491329479768785e-05,
-      "loss": 0.0773,
+      "loss": 0.0732,
       "step": 1802
     },
     {
       "epoch": 4.04,
       "learning_rate": 7.473988439306357e-05,
-      "loss": 0.0915,
+      "loss": 0.0816,
       "step": 1803
     },
     {
       "epoch": 4.04,
       "learning_rate": 7.45664739884393e-05,
-      "loss": 0.068,
+      "loss": 0.0554,
       "step": 1804
     },
     {
       "epoch": 4.05,
       "learning_rate": 7.439306358381502e-05,
-      "loss": 0.1086,
+      "loss": 0.087,
       "step": 1805
     },
     {
       "epoch": 4.05,
       "learning_rate": 7.421965317919074e-05,
-      "loss": 0.092,
+      "loss": 0.0752,
       "step": 1806
     },
     {
       "epoch": 4.05,
       "learning_rate": 7.404624277456646e-05,
-      "loss": 0.0761,
+      "loss": 0.0659,
       "step": 1807
     },
     {
       "epoch": 4.05,
       "learning_rate": 7.387283236994219e-05,
-      "loss": 0.0756,
+      "loss": 0.0671,
       "step": 1808
     },
     {
       "epoch": 4.06,
       "learning_rate": 7.369942196531791e-05,
-      "loss": 0.0843,
+      "loss": 0.0664,
       "step": 1809
     },
     {
       "epoch": 4.06,
       "learning_rate": 7.352601156069363e-05,
-      "loss": 0.0606,
+      "loss": 0.0496,
       "step": 1810
     },
     {
       "epoch": 4.06,
       "learning_rate": 7.335260115606935e-05,
-      "loss": 0.0804,
+      "loss": 0.0615,
       "step": 1811
     },
     {
       "epoch": 4.06,
       "learning_rate": 7.317919075144507e-05,
-      "loss": 0.0911,
+      "loss": 0.0902,
       "step": 1812
     },
     {
       "epoch": 4.07,
       "learning_rate": 7.30057803468208e-05,
-      "loss": 0.0853,
+      "loss": 0.0814,
       "step": 1813
     },
     {
       "epoch": 4.07,
       "learning_rate": 7.283236994219653e-05,
-      "loss": 0.0711,
+      "loss": 0.0685,
       "step": 1814
     },
     {
       "epoch": 4.07,
       "learning_rate": 7.265895953757225e-05,
-      "loss": 0.0891,
+      "loss": 0.0804,
       "step": 1815
     },
     {
       "epoch": 4.07,
       "learning_rate": 7.248554913294797e-05,
-      "loss": 0.0727,
+      "loss": 0.0699,
       "step": 1816
     },
     {
       "epoch": 4.07,
       "learning_rate": 7.231213872832369e-05,
-      "loss": 0.0723,
+      "loss": 0.0642,
       "step": 1817
     },
     {
       "epoch": 4.08,
       "learning_rate": 7.213872832369941e-05,
-      "loss": 0.0706,
+      "loss": 0.0728,
       "step": 1818
     },
     {
       "epoch": 4.08,
       "learning_rate": 7.196531791907513e-05,
-      "loss": 0.0583,
+      "loss": 0.0582,
       "step": 1819
     },
     {
       "epoch": 4.08,
       "learning_rate": 7.179190751445085e-05,
-      "loss": 0.0719,
+      "loss": 0.0681,
       "step": 1820
     },
     {
       "epoch": 4.08,
       "learning_rate": 7.161849710982659e-05,
-      "loss": 0.0874,
+      "loss": 0.055,
       "step": 1821
     },
     {
       "epoch": 4.09,
       "learning_rate": 7.144508670520231e-05,
-      "loss": 0.0596,
+      "loss": 0.053,
       "step": 1822
     },
     {
       "epoch": 4.09,
       "learning_rate": 7.127167630057803e-05,
-      "loss": 0.0784,
+      "loss": 0.0802,
       "step": 1823
     },
     {
       "epoch": 4.09,
       "learning_rate": 7.109826589595375e-05,
-      "loss": 0.0541,
+      "loss": 0.0759,
       "step": 1824
     },
     {
       "epoch": 4.09,
       "learning_rate": 7.092485549132947e-05,
-      "loss": 0.0844,
+      "loss": 0.0842,
       "step": 1825
     },
     {
       "epoch": 4.09,
       "learning_rate": 7.075144508670519e-05,
-      "loss": 0.0824,
+      "loss": 0.0695,
       "step": 1826
     },
     {
       "epoch": 4.1,
       "learning_rate": 7.057803468208091e-05,
-      "loss": 0.0689,
+      "loss": 0.0674,
       "step": 1827
     },
     {
       "epoch": 4.1,
       "learning_rate": 7.040462427745665e-05,
-      "loss": 0.064,
+      "loss": 0.522,
       "step": 1828
     },
     {
       "epoch": 4.1,
       "learning_rate": 7.023121387283237e-05,
-      "loss": 0.0773,
+      "loss": 0.0617,
       "step": 1829
     },
     {
       "epoch": 4.1,
       "learning_rate": 7.005780346820809e-05,
-      "loss": 0.0569,
+      "loss": 0.074,
       "step": 1830
     },
     {
       "epoch": 4.11,
       "learning_rate": 6.988439306358381e-05,
-      "loss": 0.0629,
+      "loss": 0.0635,
       "step": 1831
     },
     {
       "epoch": 4.11,
       "learning_rate": 6.971098265895953e-05,
-      "loss": 0.0536,
+      "loss": 0.0933,
       "step": 1832
     },
     {
       "epoch": 4.11,
       "learning_rate": 6.953757225433525e-05,
-      "loss": 0.0645,
+      "loss": 0.0421,
       "step": 1833
     },
     {
       "epoch": 4.11,
       "learning_rate": 6.936416184971097e-05,
-      "loss": 0.091,
+      "loss": 0.1101,
       "step": 1834
     },
     {
       "epoch": 4.11,
       "learning_rate": 6.91907514450867e-05,
-      "loss": 0.1786,
+      "loss": 0.1471,
       "step": 1835
     },
     {
       "epoch": 4.12,
       "learning_rate": 6.901734104046242e-05,
-      "loss": 0.1782,
+      "loss": 0.1476,
       "step": 1836
     },
     {
       "epoch": 4.12,
       "learning_rate": 6.884393063583815e-05,
-      "loss": 0.1291,
+      "loss": 0.0971,
       "step": 1837
     },
     {
       "epoch": 4.12,
       "learning_rate": 6.867052023121387e-05,
-      "loss": 0.1287,
+      "loss": 0.092,
       "step": 1838
     },
     {
       "epoch": 4.12,
       "learning_rate": 6.849710982658959e-05,
-      "loss": 0.1288,
+      "loss": 0.1004,
       "step": 1839
     },
     {
       "epoch": 4.13,
       "learning_rate": 6.832369942196531e-05,
-      "loss": 0.1246,
+      "loss": 0.1173,
       "step": 1840
     },
     {
       "epoch": 4.13,
       "learning_rate": 6.815028901734103e-05,
-      "loss": 0.1106,
+      "loss": 0.091,
       "step": 1841
     },
     {
       "epoch": 4.13,
       "learning_rate": 6.797687861271676e-05,
-      "loss": 0.1123,
+      "loss": 0.1089,
       "step": 1842
     },
     {
       "epoch": 4.13,
       "learning_rate": 6.780346820809248e-05,
-      "loss": 0.0932,
+      "loss": 0.0859,
       "step": 1843
     },
     {
       "epoch": 4.13,
       "learning_rate": 6.76300578034682e-05,
-      "loss": 0.0964,
+      "loss": 0.0726,
       "step": 1844
     },
     {
       "epoch": 4.14,
       "learning_rate": 6.745664739884392e-05,
-      "loss": 0.0843,
+      "loss": 0.084,
       "step": 1845
     },
     {
       "epoch": 4.14,
       "learning_rate": 6.728323699421964e-05,
-      "loss": 0.0865,
+      "loss": 0.0684,
       "step": 1846
     },
     {
       "epoch": 4.14,
       "learning_rate": 6.710982658959537e-05,
-      "loss": 0.0682,
+      "loss": 0.0713,
       "step": 1847
     },
     {
       "epoch": 4.14,
       "learning_rate": 6.693641618497109e-05,
-      "loss": 0.0936,
+      "loss": 0.0874,
       "step": 1848
     },
     {
       "epoch": 4.15,
       "learning_rate": 6.676300578034682e-05,
-      "loss": 0.0974,
+      "loss": 0.0691,
       "step": 1849
     },
     {
       "epoch": 4.15,
       "learning_rate": 6.658959537572254e-05,
-      "loss": 0.077,
+      "loss": 0.0745,
       "step": 1850
     },
     {
       "epoch": 4.15,
       "learning_rate": 6.641618497109826e-05,
-      "loss": 0.0933,
+      "loss": 0.0953,
       "step": 1851
     },
     {
       "epoch": 4.15,
       "learning_rate": 6.624277456647398e-05,
-      "loss": 0.0889,
+      "loss": 0.0935,
       "step": 1852
     },
     {
       "epoch": 4.15,
       "learning_rate": 6.60693641618497e-05,
-      "loss": 0.087,
+      "loss": 0.0775,
       "step": 1853
     },
     {
       "epoch": 4.16,
       "learning_rate": 6.589595375722542e-05,
-      "loss": 0.1003,
+      "loss": 0.0828,
       "step": 1854
     },
     {
       "epoch": 4.16,
       "learning_rate": 6.572254335260114e-05,
-      "loss": 0.1036,
+      "loss": 0.1012,
       "step": 1855
     },
     {
       "epoch": 4.16,
       "learning_rate": 6.554913294797688e-05,
-      "loss": 0.0973,
+      "loss": 0.1145,
       "step": 1856
     },
     {
       "epoch": 4.16,
       "learning_rate": 6.53757225433526e-05,
-      "loss": 0.0868,
+      "loss": 0.0819,
       "step": 1857
     },
     {
       "epoch": 4.17,
       "learning_rate": 6.520231213872832e-05,
-      "loss": 0.0673,
+      "loss": 0.0653,
       "step": 1858
     },
     {
       "epoch": 4.17,
       "learning_rate": 6.502890173410404e-05,
-      "loss": 0.0833,
+      "loss": 0.095,
       "step": 1859
     },
     {
       "epoch": 4.17,
       "learning_rate": 6.485549132947976e-05,
-      "loss": 0.0597,
+      "loss": 0.0649,
       "step": 1860
     },
     {
       "epoch": 4.17,
       "learning_rate": 6.468208092485548e-05,
-      "loss": 0.0666,
+      "loss": 0.0629,
       "step": 1861
     },
     {
       "epoch": 4.17,
       "learning_rate": 6.45086705202312e-05,
-      "loss": 0.0615,
+      "loss": 0.0555,
       "step": 1862
     },
     {
       "epoch": 4.18,
       "learning_rate": 6.433526011560694e-05,
-      "loss": 0.0832,
+      "loss": 0.0947,
       "step": 1863
     },
     {
       "epoch": 4.18,
       "learning_rate": 6.416184971098266e-05,
-      "loss": 0.0922,
+      "loss": 0.0661,
       "step": 1864
     },
     {
       "epoch": 4.18,
       "learning_rate": 6.398843930635838e-05,
-      "loss": 0.0847,
+      "loss": 0.0607,
       "step": 1865
     },
     {
       "epoch": 4.18,
       "learning_rate": 6.38150289017341e-05,
-      "loss": 0.067,
+      "loss": 0.0694,
       "step": 1866
     },
     {
       "epoch": 4.19,
       "learning_rate": 6.364161849710982e-05,
-      "loss": 0.0962,
+      "loss": 0.0924,
       "step": 1867
     },
     {
       "epoch": 4.19,
       "learning_rate": 6.346820809248554e-05,
-      "loss": 0.0805,
+      "loss": 0.089,
       "step": 1868
     },
     {
       "epoch": 4.19,
       "learning_rate": 6.329479768786126e-05,
-      "loss": 0.0747,
+      "loss": 0.0783,
       "step": 1869
     },
     {
       "epoch": 4.19,
       "learning_rate": 6.3121387283237e-05,
-      "loss": 0.0708,
+      "loss": 0.0567,
       "step": 1870
     },
     {
       "epoch": 4.2,
       "learning_rate": 6.294797687861272e-05,
-      "loss": 0.0769,
+      "loss": 0.0759,
       "step": 1871
     },
     {
       "epoch": 4.2,
       "learning_rate": 6.277456647398844e-05,
-      "loss": 0.0616,
+      "loss": 0.0725,
       "step": 1872
     },
     {
       "epoch": 4.2,
       "learning_rate": 6.260115606936416e-05,
-      "loss": 0.1026,
+      "loss": 0.0885,
       "step": 1873
     },
     {
       "epoch": 4.2,
       "learning_rate": 6.242774566473988e-05,
-      "loss": 0.07,
+      "loss": 0.0624,
       "step": 1874
     },
     {
       "epoch": 4.2,
       "learning_rate": 6.22543352601156e-05,
-      "loss": 0.075,
+      "loss": 0.0758,
       "step": 1875
     },
     {
       "epoch": 4.21,
       "learning_rate": 6.208092485549133e-05,
-      "loss": 0.1154,
+      "loss": 0.0935,
       "step": 1876
     },
     {
       "epoch": 4.21,
       "learning_rate": 6.190751445086705e-05,
-      "loss": 0.0768,
+      "loss": 0.0601,
       "step": 1877
     },
     {
       "epoch": 4.21,
       "learning_rate": 6.173410404624277e-05,
-      "loss": 0.0694,
+      "loss": 0.0447,
       "step": 1878
     },
     {
       "epoch": 4.21,
       "learning_rate": 6.15606936416185e-05,
-      "loss": 0.0634,
+      "loss": 0.0657,
       "step": 1879
     },
     {
       "epoch": 4.22,
       "learning_rate": 6.138728323699422e-05,
-      "loss": 0.0644,
+      "loss": 0.06,
       "step": 1880
     },
     {
       "epoch": 4.22,
       "learning_rate": 6.121387283236994e-05,
-      "loss": 0.0722,
+      "loss": 0.0806,
       "step": 1881
     },
     {
       "epoch": 4.22,
       "learning_rate": 6.104046242774566e-05,
-      "loss": 0.1028,
+      "loss": 0.096,
       "step": 1882
     },
     {
       "epoch": 4.22,
       "learning_rate": 6.0867052023121385e-05,
-      "loss": 0.0566,
+      "loss": 0.0729,
       "step": 1883
     },
     {
       "epoch": 4.22,
       "learning_rate": 6.0693641618497105e-05,
-      "loss": 0.1007,
+      "loss": 0.1086,
       "step": 1884
     },
     {
       "epoch": 4.23,
       "learning_rate": 6.052023121387283e-05,
-      "loss": 0.2018,
+      "loss": 0.1859,
       "step": 1885
     },
     {
       "epoch": 4.23,
       "learning_rate": 6.034682080924855e-05,
-      "loss": 0.1629,
+      "loss": 0.1361,
       "step": 1886
     },
     {
       "epoch": 4.23,
       "learning_rate": 6.0173410404624274e-05,
-      "loss": 0.122,
+      "loss": 0.1013,
       "step": 1887
     },
     {
       "epoch": 4.23,
       "learning_rate": 5.9999999999999995e-05,
-      "loss": 0.1055,
+      "loss": 0.0918,
       "step": 1888
     },
     {
       "epoch": 4.24,
       "learning_rate": 5.982658959537572e-05,
-      "loss": 0.1079,
+      "loss": 0.0916,
       "step": 1889
     },
     {
       "epoch": 4.24,
       "learning_rate": 5.965317919075144e-05,
-      "loss": 0.0997,
+      "loss": 0.0794,
       "step": 1890
     },
     {
       "epoch": 4.24,
       "learning_rate": 5.9479768786127164e-05,
-      "loss": 0.1041,
+      "loss": 0.1023,
       "step": 1891
     },
     {
       "epoch": 4.24,
       "learning_rate": 5.930635838150289e-05,
-      "loss": 0.1212,
+      "loss": 0.1047,
       "step": 1892
     },
     {
       "epoch": 4.24,
       "learning_rate": 5.913294797687861e-05,
-      "loss": 0.0923,
+      "loss": 0.0789,
       "step": 1893
     },
     {
       "epoch": 4.25,
       "learning_rate": 5.895953757225433e-05,
-      "loss": 0.1055,
+      "loss": 0.0798,
       "step": 1894
     },
     {
       "epoch": 4.25,
       "learning_rate": 5.878612716763006e-05,
-      "loss": 0.0925,
+      "loss": 0.0895,
       "step": 1895
     },
     {
       "epoch": 4.25,
       "learning_rate": 5.861271676300578e-05,
-      "loss": 0.0758,
+      "loss": 0.0754,
       "step": 1896
     },
     {
       "epoch": 4.25,
       "learning_rate": 5.84393063583815e-05,
-      "loss": 0.1113,
+      "loss": 0.1032,
       "step": 1897
     },
     {
       "epoch": 4.26,
       "learning_rate": 5.8265895953757215e-05,
-      "loss": 0.1055,
+      "loss": 0.0945,
       "step": 1898
     },
     {
       "epoch": 4.26,
       "learning_rate": 5.8092485549132936e-05,
-      "loss": 0.0753,
+      "loss": 0.0874,
       "step": 1899
     },
     {
       "epoch": 4.26,
       "learning_rate": 5.791907514450866e-05,
-      "loss": 0.0915,
+      "loss": 0.0859,
       "step": 1900
     },
     {
       "epoch": 4.26,
       "learning_rate": 5.7745664739884384e-05,
-      "loss": 0.0812,
+      "loss": 0.0794,
       "step": 1901
     },
     {
       "epoch": 4.26,
       "learning_rate": 5.7572254335260105e-05,
-      "loss": 0.0886,
+      "loss": 0.0629,
       "step": 1902
     },
     {
       "epoch": 4.27,
       "learning_rate": 5.739884393063583e-05,
-      "loss": 0.0959,
+      "loss": 0.087,
       "step": 1903
     },
     {
       "epoch": 4.27,
       "learning_rate": 5.722543352601155e-05,
-      "loss": 0.0911,
+      "loss": 0.0803,
       "step": 1904
     },
     {
       "epoch": 4.27,
       "learning_rate": 5.705202312138727e-05,
-      "loss": 0.0982,
+      "loss": 0.0901,
       "step": 1905
     },
     {
       "epoch": 4.27,
       "learning_rate": 5.6878612716762994e-05,
-      "loss": 0.1004,
+      "loss": 0.1085,
       "step": 1906
     },
     {
       "epoch": 4.28,
       "learning_rate": 5.670520231213872e-05,
-      "loss": 0.0696,
+      "loss": 0.0724,
       "step": 1907
     },
     {
       "epoch": 4.28,
       "learning_rate": 5.653179190751444e-05,
-      "loss": 0.0809,
+      "loss": 0.092,
       "step": 1908
     },
     {
       "epoch": 4.28,
       "learning_rate": 5.635838150289016e-05,
-      "loss": 0.0669,
+      "loss": 0.0678,
       "step": 1909
     },
     {
       "epoch": 4.28,
       "learning_rate": 5.618497109826589e-05,
-      "loss": 0.0628,
+      "loss": 0.0715,
       "step": 1910
     },
     {
       "epoch": 4.28,
       "learning_rate": 5.601156069364161e-05,
-      "loss": 0.0613,
+      "loss": 0.0755,
       "step": 1911
     },
     {
       "epoch": 4.29,
       "learning_rate": 5.583815028901733e-05,
-      "loss": 0.0931,
+      "loss": 0.0718,
       "step": 1912
     },
     {
       "epoch": 4.29,
       "learning_rate": 5.566473988439306e-05,
-      "loss": 0.0753,
+      "loss": 0.0771,
       "step": 1913
     },
     {
       "epoch": 4.29,
       "learning_rate": 5.549132947976878e-05,
-      "loss": 0.0692,
+      "loss": 0.0682,
       "step": 1914
     },
     {
       "epoch": 4.29,
       "learning_rate": 5.53179190751445e-05,
-      "loss": 0.1025,
+      "loss": 0.087,
       "step": 1915
     },
     {
       "epoch": 4.3,
       "learning_rate": 5.514450867052022e-05,
-      "loss": 0.0699,
+      "loss": 0.0566,
       "step": 1916
     },
     {
       "epoch": 4.3,
       "learning_rate": 5.497109826589595e-05,
-      "loss": 0.085,
+      "loss": 0.0778,
       "step": 1917
     },
     {
       "epoch": 4.3,
       "learning_rate": 5.479768786127167e-05,
-      "loss": 0.0596,
+      "loss": 0.0692,
       "step": 1918
     },
     {
       "epoch": 4.3,
       "learning_rate": 5.462427745664739e-05,
-      "loss": 0.0638,
+      "loss": 0.0559,
       "step": 1919
     },
     {
       "epoch": 4.3,
       "learning_rate": 5.445086705202312e-05,
-      "loss": 0.0701,
+      "loss": 0.0472,
       "step": 1920
     },
     {
       "epoch": 4.31,
       "learning_rate": 5.427745664739884e-05,
-      "loss": 0.0571,
+      "loss": 0.0515,
       "step": 1921
     },
     {
       "epoch": 4.31,
       "learning_rate": 5.410404624277456e-05,
-      "loss": 0.084,
+      "loss": 0.0699,
       "step": 1922
     },
     {
       "epoch": 4.31,
       "learning_rate": 5.393063583815028e-05,
-      "loss": 0.053,
+      "loss": 0.0601,
       "step": 1923
     },
     {
       "epoch": 4.31,
       "learning_rate": 5.375722543352601e-05,
-      "loss": 0.0747,
+      "loss": 0.0811,
       "step": 1924
     },
     {
       "epoch": 4.32,
       "learning_rate": 5.358381502890173e-05,
-      "loss": 0.0491,
+      "loss": 0.0418,
       "step": 1925
     },
     {
       "epoch": 4.32,
       "learning_rate": 5.341040462427745e-05,
-      "loss": 0.0774,
+      "loss": 0.0683,
       "step": 1926
     },
     {
       "epoch": 4.32,
       "learning_rate": 5.3236994219653176e-05,
-      "loss": 0.1053,
+      "loss": 0.0661,
       "step": 1927
     },
     {
       "epoch": 4.32,
       "learning_rate": 5.3063583815028896e-05,
-      "loss": 0.0966,
+      "loss": 0.098,
       "step": 1928
     },
     {
       "epoch": 4.33,
       "learning_rate": 5.289017341040462e-05,
-      "loss": 0.0522,
+      "loss": 0.0497,
       "step": 1929
     },
     {
       "epoch": 4.33,
       "learning_rate": 5.271676300578034e-05,
-      "loss": 0.0609,
+      "loss": 0.0389,
       "step": 1930
     },
     {
       "epoch": 4.33,
       "learning_rate": 5.2543352601156065e-05,
-      "loss": 0.055,
+      "loss": 0.041,
       "step": 1931
     },
     {
       "epoch": 4.33,
       "learning_rate": 5.2369942196531786e-05,
-      "loss": 0.1006,
+      "loss": 0.0961,
       "step": 1932
     },
     {
       "epoch": 4.33,
       "learning_rate": 5.2196531791907507e-05,
-      "loss": 0.1089,
+      "loss": 0.0785,
       "step": 1933
     },
     {
       "epoch": 4.34,
       "learning_rate": 5.2023121387283234e-05,
-      "loss": 0.0967,
+      "loss": 0.1104,
       "step": 1934
     },
     {
       "epoch": 4.34,
       "learning_rate": 5.1849710982658955e-05,
-      "loss": 0.1576,
+      "loss": 0.124,
       "step": 1935
     },
     {
       "epoch": 4.34,
       "learning_rate": 5.1676300578034675e-05,
-      "loss": 0.144,
+      "loss": 0.1314,
       "step": 1936
     },
     {
       "epoch": 4.34,
       "learning_rate": 5.1502890173410396e-05,
-      "loss": 0.1126,
+      "loss": 0.094,
       "step": 1937
     },
     {
       "epoch": 4.35,
       "learning_rate": 5.1329479768786124e-05,
-      "loss": 0.1134,
+      "loss": 0.0923,
       "step": 1938
     },
     {
       "epoch": 4.35,
       "learning_rate": 5.1156069364161844e-05,
-      "loss": 0.1297,
+      "loss": 0.1036,
       "step": 1939
     },
     {
       "epoch": 4.35,
       "learning_rate": 5.0982658959537565e-05,
-      "loss": 0.1048,
+      "loss": 0.0725,
       "step": 1940
     },
     {
       "epoch": 4.35,
       "learning_rate": 5.080924855491329e-05,
-      "loss": 0.0935,
+      "loss": 0.0853,
       "step": 1941
     },
     {
       "epoch": 4.35,
       "learning_rate": 5.063583815028901e-05,
-      "loss": 0.1055,
+      "loss": 0.1104,
       "step": 1942
     },
     {
       "epoch": 4.36,
       "learning_rate": 5.0462427745664734e-05,
-      "loss": 0.0977,
+      "loss": 0.0764,
       "step": 1943
     },
     {
       "epoch": 4.36,
       "learning_rate": 5.028901734104046e-05,
-      "loss": 0.0831,
+      "loss": 0.0781,
       "step": 1944
     },
     {
       "epoch": 4.36,
       "learning_rate": 5.011560693641618e-05,
-      "loss": 0.0901,
+      "loss": 0.0802,
       "step": 1945
     },
     {
       "epoch": 4.36,
       "learning_rate": 4.99421965317919e-05,
-      "loss": 0.1007,
+      "loss": 0.0705,
       "step": 1946
     },
     {
       "epoch": 4.37,
       "learning_rate": 4.976878612716762e-05,
-      "loss": 0.0779,
+      "loss": 0.0725,
       "step": 1947
     },
     {
       "epoch": 4.37,
       "learning_rate": 4.959537572254335e-05,
-      "loss": 0.0516,
+      "loss": 0.054,
       "step": 1948
     },
     {
       "epoch": 4.37,
       "learning_rate": 4.942196531791907e-05,
-      "loss": 0.0659,
+      "loss": 0.0575,
       "step": 1949
     },
     {
       "epoch": 4.37,
       "learning_rate": 4.924855491329479e-05,
-      "loss": 0.0775,
+      "loss": 0.0801,
       "step": 1950
     },
     {
       "epoch": 4.37,
       "learning_rate": 4.907514450867052e-05,
-      "loss": 0.0825,
+      "loss": 0.0763,
       "step": 1951
     },
     {
       "epoch": 4.38,
       "learning_rate": 4.890173410404624e-05,
-      "loss": 0.0651,
+      "loss": 0.0724,
       "step": 1952
     },
     {
       "epoch": 4.38,
       "learning_rate": 4.872832369942196e-05,
-      "loss": 0.0806,
+      "loss": 0.0776,
       "step": 1953
     },
     {
       "epoch": 4.38,
       "learning_rate": 4.855491329479768e-05,
-      "loss": 0.0581,
+      "loss": 0.0493,
       "step": 1954
     },
     {
       "epoch": 4.38,
       "learning_rate": 4.838150289017341e-05,
-      "loss": 0.0721,
+      "loss": 0.0714,
       "step": 1955
     },
     {
       "epoch": 4.39,
       "learning_rate": 4.820809248554913e-05,
-      "loss": 0.0684,
+      "loss": 0.0701,
       "step": 1956
     },
     {
       "epoch": 4.39,
       "learning_rate": 4.803468208092485e-05,
-      "loss": 0.0702,
+      "loss": 0.0742,
       "step": 1957
     },
     {
       "epoch": 4.39,
       "learning_rate": 4.786127167630058e-05,
-      "loss": 0.0604,
+      "loss": 0.0693,
       "step": 1958
     },
     {
       "epoch": 4.39,
       "learning_rate": 4.76878612716763e-05,
-      "loss": 0.0662,
+      "loss": 0.0687,
       "step": 1959
     },
     {
       "epoch": 4.39,
       "learning_rate": 4.751445086705202e-05,
-      "loss": 0.0537,
+      "loss": 0.0831,
       "step": 1960
     },
     {
       "epoch": 4.4,
       "learning_rate": 4.734104046242774e-05,
-      "loss": 0.0625,
+      "loss": 0.0627,
       "step": 1961
     },
     {
       "epoch": 4.4,
       "learning_rate": 4.716763005780347e-05,
-      "loss": 0.0728,
+      "loss": 0.0741,
       "step": 1962
     },
     {
       "epoch": 4.4,
       "learning_rate": 4.699421965317919e-05,
-      "loss": 0.0644,
+      "loss": 0.0666,
       "step": 1963
     },
     {
       "epoch": 4.4,
       "learning_rate": 4.682080924855491e-05,
-      "loss": 0.0851,
+      "loss": 0.0782,
       "step": 1964
     },
     {
       "epoch": 4.41,
       "learning_rate": 4.6647398843930636e-05,
-      "loss": 0.0825,
+      "loss": 0.0725,
       "step": 1965
     },
     {
       "epoch": 4.41,
       "learning_rate": 4.647398843930636e-05,
-      "loss": 0.0469,
+      "loss": 0.044,
       "step": 1966
     },
     {
       "epoch": 4.41,
       "learning_rate": 4.630057803468208e-05,
-      "loss": 0.0861,
+      "loss": 0.0721,
       "step": 1967
     },
     {
       "epoch": 4.41,
       "learning_rate": 4.61271676300578e-05,
-      "loss": 0.0768,
+      "loss": 0.0641,
       "step": 1968
     },
     {
       "epoch": 4.41,
       "learning_rate": 4.5953757225433526e-05,
-      "loss": 0.084,
+      "loss": 0.0749,
       "step": 1969
     },
     {
       "epoch": 4.42,
       "learning_rate": 4.5780346820809246e-05,
-      "loss": 0.072,
+      "loss": 0.0604,
       "step": 1970
     },
     {
       "epoch": 4.42,
       "learning_rate": 4.560693641618497e-05,
-      "loss": 0.0608,
+      "loss": 0.0612,
       "step": 1971
     },
     {
       "epoch": 4.42,
       "learning_rate": 4.5433526011560694e-05,
-      "loss": 0.0594,
+      "loss": 0.0562,
       "step": 1972
     },
     {
       "epoch": 4.42,
       "learning_rate": 4.5260115606936415e-05,
-      "loss": 0.0618,
+      "loss": 0.0648,
       "step": 1973
     },
     {
       "epoch": 4.43,
       "learning_rate": 4.5086705202312136e-05,
-      "loss": 0.0636,
+      "loss": 0.0514,
       "step": 1974
     },
     {
       "epoch": 4.43,
       "learning_rate": 4.491329479768786e-05,
-      "loss": 0.0764,
+      "loss": 0.0888,
       "step": 1975
     },
     {
       "epoch": 4.43,
       "learning_rate": 4.4739884393063584e-05,
-      "loss": 0.0675,
+      "loss": 0.0814,
       "step": 1976
     },
     {
       "epoch": 4.43,
       "learning_rate": 4.4566473988439305e-05,
-      "loss": 0.0907,
+      "loss": 0.0913,
       "step": 1977
     },
     {
       "epoch": 4.43,
       "learning_rate": 4.4393063583815025e-05,
-      "loss": 0.0753,
+      "loss": 0.0764,
       "step": 1978
     },
     {
       "epoch": 4.44,
       "learning_rate": 4.421965317919075e-05,
-      "loss": 0.1134,
+      "loss": 0.0806,
       "step": 1979
     },
     {
       "epoch": 4.44,
       "learning_rate": 4.404624277456647e-05,
-      "loss": 0.0515,
+      "loss": 0.0493,
       "step": 1980
     },
     {
       "epoch": 4.44,
       "learning_rate": 4.3872832369942194e-05,
-      "loss": 0.0489,
+      "loss": 0.0454,
       "step": 1981
     },
     {
       "epoch": 4.44,
       "learning_rate": 4.369942196531792e-05,
-      "loss": 0.0706,
+      "loss": 0.0552,
       "step": 1982
     },
     {
       "epoch": 4.45,
       "learning_rate": 4.352601156069364e-05,
-      "loss": 0.0806,
+      "loss": 0.0891,
       "step": 1983
     },
     {
       "epoch": 4.45,
       "learning_rate": 4.335260115606936e-05,
-      "loss": 0.1286,
+      "loss": 0.1034,
       "step": 1984
     },
     {
       "epoch": 4.45,
       "learning_rate": 4.3179190751445084e-05,
-      "loss": 0.1647,
+      "loss": 0.1403,
       "step": 1985
     },
     {
       "epoch": 4.45,
       "learning_rate": 4.300578034682081e-05,
-      "loss": 0.1068,
+      "loss": 0.0965,
       "step": 1986
     },
     {
       "epoch": 4.46,
       "learning_rate": 4.283236994219653e-05,
-      "loss": 0.1227,
+      "loss": 0.1017,
       "step": 1987
     },
     {
       "epoch": 4.46,
       "learning_rate": 4.265895953757225e-05,
-      "loss": 0.1077,
+      "loss": 0.1015,
       "step": 1988
     },
     {
       "epoch": 4.46,
       "learning_rate": 4.248554913294798e-05,
-      "loss": 0.098,
+      "loss": 0.0803,
       "step": 1989
     },
     {
       "epoch": 4.46,
       "learning_rate": 4.23121387283237e-05,
-      "loss": 0.1061,
+      "loss": 0.0917,
       "step": 1990
     },
     {
       "epoch": 4.46,
       "learning_rate": 4.213872832369942e-05,
-      "loss": 0.1334,
+      "loss": 0.1152,
       "step": 1991
     },
     {
       "epoch": 4.47,
       "learning_rate": 4.196531791907514e-05,
-      "loss": 0.0944,
+      "loss": 0.0697,
       "step": 1992
     },
     {
       "epoch": 4.47,
       "learning_rate": 4.179190751445087e-05,
-      "loss": 0.1024,
+      "loss": 0.0904,
       "step": 1993
     },
     {
       "epoch": 4.47,
       "learning_rate": 4.161849710982658e-05,
-      "loss": 0.0711,
+      "loss": 0.0707,
       "step": 1994
     },
     {
       "epoch": 4.47,
       "learning_rate": 4.1445086705202304e-05,
-      "loss": 0.0685,
+      "loss": 0.0799,
       "step": 1995
     },
     {
       "epoch": 4.48,
       "learning_rate": 4.1271676300578025e-05,
-      "loss": 0.0949,
+      "loss": 0.0883,
       "step": 1996
     },
     {
       "epoch": 4.48,
       "learning_rate": 4.109826589595375e-05,
-      "loss": 0.0936,
+      "loss": 0.1033,
       "step": 1997
     },
     {
       "epoch": 4.48,
       "learning_rate": 4.092485549132947e-05,
-      "loss": 0.0813,
+      "loss": 0.0797,
       "step": 1998
     },
     {
       "epoch": 4.48,
       "learning_rate": 4.075144508670519e-05,
-      "loss": 0.0909,
+      "loss": 0.0984,
       "step": 1999
     },
     {
       "epoch": 4.48,
       "learning_rate": 4.057803468208092e-05,
-      "loss": 0.0826,
+      "loss": 0.0753,
       "step": 2000
     },
     {
       "epoch": 4.48,
-      "eval_loss": 0.309664785861969,
-      "eval_runtime": 583.4014,
-      "eval_samples_per_second": 4.529,
-      "eval_steps_per_second": 0.567,
-      "eval_wer": 0.09321697738992463,
+      "eval_loss": 0.32974740862846375,
+      "eval_runtime": 555.2371,
+      "eval_samples_per_second": 4.758,
+      "eval_steps_per_second": 0.596,
+      "eval_wer": 0.09492264974216581,
       "step": 2000
     },
     {
       "epoch": 4.49,
       "learning_rate": 4.040462427745664e-05,
-      "loss": 0.0722,
+      "loss": 0.0764,
       "step": 2001
     },
     {
       "epoch": 4.49,
       "learning_rate": 4.023121387283236e-05,
-      "loss": 0.0647,
+      "loss": 0.0685,
       "step": 2002
     },
     {
@@ -12064,1373 +12064,1373 @@
     {
       "epoch": 4.49,
       "learning_rate": 3.988439306358381e-05,
-      "loss": 0.0915,
+      "loss": 0.0908,
       "step": 2004
     },
     {
       "epoch": 4.5,
       "learning_rate": 3.971098265895953e-05,
-      "loss": 0.0799,
+      "loss": 0.0857,
       "step": 2005
     },
     {
       "epoch": 4.5,
       "learning_rate": 3.953757225433525e-05,
-      "loss": 0.0812,
+      "loss": 0.0824,
       "step": 2006
     },
     {
       "epoch": 4.5,
       "learning_rate": 3.936416184971098e-05,
-      "loss": 0.0717,
+      "loss": 0.0638,
       "step": 2007
     },
     {
       "epoch": 4.5,
       "learning_rate": 3.91907514450867e-05,
-      "loss": 0.0719,
+      "loss": 0.0703,
       "step": 2008
     },
     {
       "epoch": 4.5,
       "learning_rate": 3.901734104046242e-05,
-      "loss": 0.0865,
+      "loss": 0.0662,
       "step": 2009
     },
     {
       "epoch": 4.51,
       "learning_rate": 3.884393063583814e-05,
-      "loss": 0.0561,
+      "loss": 0.0581,
       "step": 2010
     },
     {
       "epoch": 4.51,
       "learning_rate": 3.867052023121387e-05,
-      "loss": 0.067,
+      "loss": 0.0755,
       "step": 2011
     },
     {
       "epoch": 4.51,
       "learning_rate": 3.849710982658959e-05,
-      "loss": 0.0446,
+      "loss": 0.0439,
       "step": 2012
     },
     {
       "epoch": 4.51,
       "learning_rate": 3.832369942196531e-05,
-      "loss": 0.053,
+      "loss": 0.0588,
       "step": 2013
     },
     {
       "epoch": 4.52,
       "learning_rate": 3.815028901734104e-05,
-      "loss": 0.0593,
+      "loss": 0.0649,
       "step": 2014
     },
     {
       "epoch": 4.52,
       "learning_rate": 3.797687861271676e-05,
-      "loss": 0.0753,
+      "loss": 0.0867,
       "step": 2015
     },
     {
       "epoch": 4.52,
       "learning_rate": 3.780346820809248e-05,
-      "loss": 0.0629,
+      "loss": 0.071,
       "step": 2016
     },
     {
       "epoch": 4.52,
       "learning_rate": 3.76300578034682e-05,
-      "loss": 0.0816,
+      "loss": 0.09,
       "step": 2017
     },
     {
       "epoch": 4.52,
       "learning_rate": 3.745664739884393e-05,
-      "loss": 0.0547,
+      "loss": 0.0448,
       "step": 2018
     },
     {
       "epoch": 4.53,
       "learning_rate": 3.728323699421965e-05,
-      "loss": 0.0735,
+      "loss": 0.0613,
       "step": 2019
     },
     {
       "epoch": 4.53,
       "learning_rate": 3.710982658959537e-05,
-      "loss": 0.0501,
+      "loss": 0.0503,
       "step": 2020
     },
     {
       "epoch": 4.53,
       "learning_rate": 3.6936416184971096e-05,
-      "loss": 0.0683,
+      "loss": 0.0799,
       "step": 2021
     },
     {
       "epoch": 4.53,
       "learning_rate": 3.6763005780346816e-05,
-      "loss": 0.0832,
+      "loss": 0.0628,
       "step": 2022
     },
     {
       "epoch": 4.54,
       "learning_rate": 3.658959537572254e-05,
-      "loss": 0.0498,
+      "loss": 0.0486,
       "step": 2023
     },
     {
       "epoch": 4.54,
       "learning_rate": 3.6416184971098265e-05,
-      "loss": 0.0779,
+      "loss": 0.0715,
       "step": 2024
     },
     {
       "epoch": 4.54,
       "learning_rate": 3.6242774566473985e-05,
-      "loss": 0.0607,
+      "loss": 0.0677,
       "step": 2025
     },
     {
       "epoch": 4.54,
       "learning_rate": 3.6069364161849706e-05,
-      "loss": 0.0659,
+      "loss": 0.063,
       "step": 2026
     },
     {
       "epoch": 4.54,
       "learning_rate": 3.5895953757225427e-05,
-      "loss": 0.0821,
+      "loss": 0.0659,
       "step": 2027
     },
     {
       "epoch": 4.55,
       "learning_rate": 3.5722543352601154e-05,
-      "loss": 0.0873,
+      "loss": 0.7377,
       "step": 2028
     },
     {
       "epoch": 4.55,
       "learning_rate": 3.5549132947976875e-05,
-      "loss": 0.1007,
+      "loss": 0.093,
       "step": 2029
     },
     {
       "epoch": 4.55,
       "learning_rate": 3.5375722543352595e-05,
-      "loss": 0.0521,
+      "loss": 0.0447,
       "step": 2030
     },
     {
       "epoch": 4.55,
       "learning_rate": 3.520231213872832e-05,
-      "loss": 0.0381,
+      "loss": 0.0416,
       "step": 2031
     },
     {
       "epoch": 4.56,
       "learning_rate": 3.5028901734104043e-05,
-      "loss": 0.0719,
+      "loss": 0.0708,
       "step": 2032
     },
     {
       "epoch": 4.56,
       "learning_rate": 3.4855491329479764e-05,
-      "loss": 0.067,
+      "loss": 0.0784,
       "step": 2033
     },
     {
       "epoch": 4.56,
       "learning_rate": 3.4682080924855485e-05,
-      "loss": 0.1192,
+      "loss": 0.108,
       "step": 2034
     },
     {
       "epoch": 4.56,
       "learning_rate": 3.450867052023121e-05,
-      "loss": 0.1441,
+      "loss": 0.1217,
       "step": 2035
     },
     {
       "epoch": 4.57,
       "learning_rate": 3.433526011560693e-05,
-      "loss": 0.1257,
+      "loss": 0.1173,
       "step": 2036
     },
     {
       "epoch": 4.57,
       "learning_rate": 3.4161849710982654e-05,
-      "loss": 0.142,
+      "loss": 0.1258,
       "step": 2037
     },
     {
       "epoch": 4.57,
       "learning_rate": 3.398843930635838e-05,
-      "loss": 0.1018,
+      "loss": 0.0868,
       "step": 2038
     },
     {
       "epoch": 4.57,
       "learning_rate": 3.38150289017341e-05,
-      "loss": 0.1209,
+      "loss": 0.1049,
       "step": 2039
     },
     {
       "epoch": 4.57,
       "learning_rate": 3.364161849710982e-05,
-      "loss": 0.0767,
+      "loss": 0.0621,
       "step": 2040
     },
     {
       "epoch": 4.58,
       "learning_rate": 3.346820809248554e-05,
-      "loss": 0.0638,
+      "loss": 0.0679,
       "step": 2041
     },
     {
       "epoch": 4.58,
       "learning_rate": 3.329479768786127e-05,
-      "loss": 0.0763,
+      "loss": 0.0786,
       "step": 2042
     },
     {
       "epoch": 4.58,
       "learning_rate": 3.312138728323699e-05,
-      "loss": 0.1027,
+      "loss": 0.0961,
       "step": 2043
     },
     {
       "epoch": 4.58,
       "learning_rate": 3.294797687861271e-05,
-      "loss": 0.0879,
+      "loss": 0.098,
       "step": 2044
     },
     {
       "epoch": 4.59,
       "learning_rate": 3.277456647398844e-05,
-      "loss": 0.072,
+      "loss": 0.0582,
       "step": 2045
     },
     {
       "epoch": 4.59,
       "learning_rate": 3.260115606936416e-05,
-      "loss": 0.0753,
+      "loss": 0.0676,
       "step": 2046
     },
     {
       "epoch": 4.59,
       "learning_rate": 3.242774566473988e-05,
-      "loss": 0.0689,
+      "loss": 0.0756,
       "step": 2047
     },
     {
       "epoch": 4.59,
       "learning_rate": 3.22543352601156e-05,
-      "loss": 0.071,
+      "loss": 0.0821,
       "step": 2048
     },
     {
       "epoch": 4.59,
       "learning_rate": 3.208092485549133e-05,
-      "loss": 0.0758,
+      "loss": 0.0867,
       "step": 2049
     },
     {
       "epoch": 4.6,
       "learning_rate": 3.190751445086705e-05,
-      "loss": 0.0574,
+      "loss": 0.0538,
       "step": 2050
     },
     {
       "epoch": 4.6,
       "learning_rate": 3.173410404624277e-05,
-      "loss": 0.0876,
+      "loss": 0.0831,
       "step": 2051
     },
     {
       "epoch": 4.6,
       "learning_rate": 3.15606936416185e-05,
-      "loss": 0.0567,
+      "loss": 0.0608,
       "step": 2052
     },
     {
       "epoch": 4.6,
       "learning_rate": 3.138728323699422e-05,
-      "loss": 0.0697,
+      "loss": 0.067,
       "step": 2053
     },
     {
       "epoch": 4.61,
       "learning_rate": 3.121387283236994e-05,
-      "loss": 0.0634,
+      "loss": 0.0615,
       "step": 2054
     },
     {
       "epoch": 4.61,
       "learning_rate": 3.1040462427745667e-05,
-      "loss": 0.067,
+      "loss": 0.0613,
       "step": 2055
     },
     {
       "epoch": 4.61,
       "learning_rate": 3.086705202312139e-05,
-      "loss": 0.0614,
+      "loss": 0.0709,
       "step": 2056
     },
     {
       "epoch": 4.61,
       "learning_rate": 3.069364161849711e-05,
-      "loss": 0.0644,
+      "loss": 0.0792,
       "step": 2057
     },
     {
       "epoch": 4.61,
       "learning_rate": 3.052023121387283e-05,
-      "loss": 0.0702,
+      "loss": 0.0584,
       "step": 2058
     },
     {
       "epoch": 4.62,
       "learning_rate": 3.0346820809248553e-05,
-      "loss": 0.0629,
+      "loss": 0.071,
       "step": 2059
     },
     {
       "epoch": 4.62,
       "learning_rate": 3.0173410404624277e-05,
-      "loss": 0.0638,
+      "loss": 0.0718,
       "step": 2060
     },
     {
       "epoch": 4.62,
       "learning_rate": 2.9999999999999997e-05,
-      "loss": 0.0759,
+      "loss": 0.0671,
       "step": 2061
     },
     {
       "epoch": 4.62,
       "learning_rate": 2.982658959537572e-05,
-      "loss": 0.0786,
+      "loss": 0.0882,
       "step": 2062
     },
     {
       "epoch": 4.63,
       "learning_rate": 2.9653179190751446e-05,
-      "loss": 0.0659,
+      "loss": 0.0676,
       "step": 2063
     },
     {
       "epoch": 4.63,
       "learning_rate": 2.9479768786127166e-05,
-      "loss": 0.0564,
+      "loss": 0.0508,
       "step": 2064
     },
     {
       "epoch": 4.63,
       "learning_rate": 2.930635838150289e-05,
-      "loss": 0.0729,
+      "loss": 0.0604,
       "step": 2065
     },
     {
       "epoch": 4.63,
       "learning_rate": 2.9132947976878608e-05,
-      "loss": 0.0761,
+      "loss": 0.0594,
       "step": 2066
     },
     {
       "epoch": 4.63,
       "learning_rate": 2.895953757225433e-05,
-      "loss": 0.0656,
+      "loss": 0.071,
       "step": 2067
     },
     {
       "epoch": 4.64,
       "learning_rate": 2.8786127167630052e-05,
-      "loss": 0.0555,
+      "loss": 0.053,
       "step": 2068
     },
     {
       "epoch": 4.64,
       "learning_rate": 2.8612716763005776e-05,
-      "loss": 0.0612,
+      "loss": 0.0668,
       "step": 2069
     },
     {
       "epoch": 4.64,
       "learning_rate": 2.8439306358381497e-05,
-      "loss": 0.0519,
+      "loss": 0.054,
       "step": 2070
     },
     {
       "epoch": 4.64,
       "learning_rate": 2.826589595375722e-05,
-      "loss": 0.0574,
+      "loss": 0.0587,
       "step": 2071
     },
     {
       "epoch": 4.65,
       "learning_rate": 2.8092485549132945e-05,
-      "loss": 0.0427,
+      "loss": 0.0616,
       "step": 2072
     },
     {
       "epoch": 4.65,
       "learning_rate": 2.7919075144508666e-05,
-      "loss": 0.0728,
+      "loss": 0.0662,
       "step": 2073
     },
     {
       "epoch": 4.65,
       "learning_rate": 2.774566473988439e-05,
-      "loss": 0.0571,
+      "loss": 0.0504,
       "step": 2074
     },
     {
       "epoch": 4.65,
       "learning_rate": 2.757225433526011e-05,
-      "loss": 0.0653,
+      "loss": 0.0569,
       "step": 2075
     },
     {
       "epoch": 4.65,
       "learning_rate": 2.7398843930635835e-05,
-      "loss": 0.0716,
+      "loss": 0.065,
       "step": 2076
     },
     {
       "epoch": 4.66,
       "learning_rate": 2.722543352601156e-05,
-      "loss": 0.0673,
+      "loss": 0.0687,
       "step": 2077
     },
     {
       "epoch": 4.66,
       "learning_rate": 2.705202312138728e-05,
-      "loss": 0.0506,
+      "loss": 0.049,
       "step": 2078
     },
     {
       "epoch": 4.66,
       "learning_rate": 2.6878612716763003e-05,
-      "loss": 0.0705,
+      "loss": 0.0833,
       "step": 2079
     },
     {
       "epoch": 4.66,
       "learning_rate": 2.6705202312138724e-05,
-      "loss": 0.0568,
+      "loss": 0.0515,
       "step": 2080
     },
     {
       "epoch": 4.67,
       "learning_rate": 2.6531791907514448e-05,
-      "loss": 0.089,
+      "loss": 0.0803,
       "step": 2081
     },
     {
       "epoch": 4.67,
       "learning_rate": 2.635838150289017e-05,
-      "loss": 0.0592,
+      "loss": 0.0367,
       "step": 2082
     },
     {
       "epoch": 4.67,
       "learning_rate": 2.6184971098265893e-05,
-      "loss": 0.0563,
+      "loss": 0.0588,
       "step": 2083
     },
     {
       "epoch": 4.67,
       "learning_rate": 2.6011560693641617e-05,
-      "loss": 0.0935,
+      "loss": 0.085,
       "step": 2084
     },
     {
       "epoch": 4.67,
       "learning_rate": 2.5838150289017338e-05,
-      "loss": 0.1174,
+      "loss": 0.1184,
       "step": 2085
     },
     {
       "epoch": 4.68,
       "learning_rate": 2.5664739884393062e-05,
-      "loss": 0.1055,
+      "loss": 0.1096,
       "step": 2086
     },
     {
       "epoch": 4.68,
       "learning_rate": 2.5491329479768782e-05,
-      "loss": 0.0938,
+      "loss": 0.0876,
       "step": 2087
     },
     {
       "epoch": 4.68,
       "learning_rate": 2.5317919075144507e-05,
-      "loss": 0.0872,
+      "loss": 0.0747,
       "step": 2088
     },
     {
       "epoch": 4.68,
       "learning_rate": 2.514450867052023e-05,
-      "loss": 0.1099,
+      "loss": 0.0849,
       "step": 2089
     },
     {
       "epoch": 4.69,
       "learning_rate": 2.497109826589595e-05,
-      "loss": 0.094,
+      "loss": 0.0875,
       "step": 2090
     },
     {
       "epoch": 4.69,
       "learning_rate": 2.4797687861271675e-05,
-      "loss": 0.1113,
+      "loss": 0.1031,
       "step": 2091
     },
     {
       "epoch": 4.69,
       "learning_rate": 2.4624277456647396e-05,
-      "loss": 0.0828,
+      "loss": 0.1113,
       "step": 2092
     },
     {
       "epoch": 4.69,
       "learning_rate": 2.445086705202312e-05,
-      "loss": 0.0799,
+      "loss": 0.0534,
       "step": 2093
     },
     {
       "epoch": 4.7,
       "learning_rate": 2.427745664739884e-05,
-      "loss": 0.0837,
+      "loss": 0.0803,
       "step": 2094
     },
     {
       "epoch": 4.7,
       "learning_rate": 2.4104046242774565e-05,
-      "loss": 0.0823,
+      "loss": 0.0649,
       "step": 2095
     },
     {
       "epoch": 4.7,
       "learning_rate": 2.393063583815029e-05,
-      "loss": 0.0878,
+      "loss": 0.0756,
       "step": 2096
     },
     {
       "epoch": 4.7,
       "learning_rate": 2.375722543352601e-05,
-      "loss": 0.0632,
+      "loss": 0.0671,
       "step": 2097
     },
     {
       "epoch": 4.7,
       "learning_rate": 2.3583815028901734e-05,
-      "loss": 0.0509,
+      "loss": 0.0649,
       "step": 2098
     },
     {
       "epoch": 4.71,
       "learning_rate": 2.3410404624277454e-05,
-      "loss": 0.0624,
+      "loss": 0.0595,
       "step": 2099
     },
     {
       "epoch": 4.71,
       "learning_rate": 2.323699421965318e-05,
-      "loss": 0.0719,
+      "loss": 0.0871,
       "step": 2100
     },
     {
       "epoch": 4.71,
       "learning_rate": 2.30635838150289e-05,
-      "loss": 0.0763,
+      "loss": 0.0608,
       "step": 2101
     },
     {
       "epoch": 4.71,
       "learning_rate": 2.2890173410404623e-05,
-      "loss": 0.0815,
+      "loss": 0.0828,
       "step": 2102
     },
     {
       "epoch": 4.72,
       "learning_rate": 2.2716763005780347e-05,
-      "loss": 0.0548,
+      "loss": 0.0578,
       "step": 2103
     },
     {
       "epoch": 4.72,
       "learning_rate": 2.2543352601156068e-05,
-      "loss": 0.0742,
+      "loss": 0.0769,
       "step": 2104
     },
     {
       "epoch": 4.72,
       "learning_rate": 2.2369942196531792e-05,
-      "loss": 0.0509,
+      "loss": 0.0526,
       "step": 2105
     },
     {
       "epoch": 4.72,
       "learning_rate": 2.2196531791907513e-05,
-      "loss": 0.0468,
+      "loss": 0.0487,
       "step": 2106
     },
     {
       "epoch": 4.72,
       "learning_rate": 2.2023121387283237e-05,
-      "loss": 0.0499,
+      "loss": 0.0588,
       "step": 2107
     },
     {
       "epoch": 4.73,
       "learning_rate": 2.184971098265896e-05,
-      "loss": 0.0682,
+      "loss": 0.0664,
       "step": 2108
     },
     {
       "epoch": 4.73,
       "learning_rate": 2.167630057803468e-05,
-      "loss": 0.054,
+      "loss": 0.0583,
       "step": 2109
     },
     {
       "epoch": 4.73,
       "learning_rate": 2.1502890173410405e-05,
-      "loss": 0.0559,
+      "loss": 0.0539,
       "step": 2110
     },
     {
       "epoch": 4.73,
       "learning_rate": 2.1329479768786126e-05,
-      "loss": 0.0495,
+      "loss": 0.0558,
       "step": 2111
     },
     {
       "epoch": 4.74,
       "learning_rate": 2.115606936416185e-05,
-      "loss": 0.0674,
+      "loss": 0.0781,
       "step": 2112
     },
     {
       "epoch": 4.74,
       "learning_rate": 2.098265895953757e-05,
-      "loss": 0.0673,
+      "loss": 0.0674,
       "step": 2113
     },
     {
       "epoch": 4.74,
       "learning_rate": 2.080924855491329e-05,
-      "loss": 0.0707,
+      "loss": 0.0528,
       "step": 2114
     },
     {
       "epoch": 4.74,
       "learning_rate": 2.0635838150289012e-05,
-      "loss": 0.0635,
+      "loss": 0.0598,
       "step": 2115
     },
     {
       "epoch": 4.74,
       "learning_rate": 2.0462427745664736e-05,
-      "loss": 0.0778,
+      "loss": 0.0864,
       "step": 2116
     },
     {
       "epoch": 4.75,
       "learning_rate": 2.028901734104046e-05,
-      "loss": 0.0424,
+      "loss": 0.0483,
       "step": 2117
     },
     {
       "epoch": 4.75,
       "learning_rate": 2.011560693641618e-05,
-      "loss": 0.0546,
+      "loss": 0.0676,
       "step": 2118
     },
     {
       "epoch": 4.75,
       "learning_rate": 1.9942196531791905e-05,
-      "loss": 0.0524,
+      "loss": 0.0697,
       "step": 2119
     },
     {
       "epoch": 4.75,
       "learning_rate": 1.9768786127167626e-05,
-      "loss": 0.0984,
+      "loss": 0.0706,
       "step": 2120
     },
     {
       "epoch": 4.76,
       "learning_rate": 1.959537572254335e-05,
-      "loss": 0.0594,
+      "loss": 0.0506,
       "step": 2121
     },
     {
       "epoch": 4.76,
       "learning_rate": 1.942196531791907e-05,
-      "loss": 0.0632,
+      "loss": 0.0448,
       "step": 2122
     },
     {
       "epoch": 4.76,
       "learning_rate": 1.9248554913294795e-05,
-      "loss": 0.0498,
+      "loss": 0.0587,
       "step": 2123
     },
     {
       "epoch": 4.76,
       "learning_rate": 1.907514450867052e-05,
-      "loss": 0.0575,
+      "loss": 0.0716,
       "step": 2124
     },
     {
       "epoch": 4.76,
       "learning_rate": 1.890173410404624e-05,
-      "loss": 0.049,
+      "loss": 0.0411,
       "step": 2125
     },
     {
       "epoch": 4.77,
       "learning_rate": 1.8728323699421963e-05,
-      "loss": 0.0828,
+      "loss": 0.0973,
       "step": 2126
     },
     {
       "epoch": 4.77,
       "learning_rate": 1.8554913294797684e-05,
-      "loss": 0.0676,
+      "loss": 0.066,
       "step": 2127
     },
     {
       "epoch": 4.77,
       "learning_rate": 1.8381502890173408e-05,
-      "loss": 0.0341,
+      "loss": 0.039,
       "step": 2128
     },
     {
       "epoch": 4.77,
       "learning_rate": 1.8208092485549132e-05,
-      "loss": 0.0705,
+      "loss": 0.0588,
       "step": 2129
     },
     {
       "epoch": 4.78,
       "learning_rate": 1.8034682080924853e-05,
-      "loss": 0.0598,
+      "loss": 0.0343,
       "step": 2130
     },
     {
       "epoch": 4.78,
       "learning_rate": 1.7861271676300577e-05,
-      "loss": 0.0948,
+      "loss": 0.0909,
       "step": 2131
     },
     {
       "epoch": 4.78,
       "learning_rate": 1.7687861271676298e-05,
-      "loss": 0.0519,
+      "loss": 0.0581,
       "step": 2132
     },
     {
       "epoch": 4.78,
       "learning_rate": 1.7514450867052022e-05,
-      "loss": 0.0651,
+      "loss": 0.0546,
       "step": 2133
     },
     {
       "epoch": 4.78,
       "learning_rate": 1.7341040462427742e-05,
-      "loss": 0.0618,
+      "loss": 0.057,
       "step": 2134
     },
     {
       "epoch": 4.79,
       "learning_rate": 1.7167630057803466e-05,
-      "loss": 0.1591,
+      "loss": 0.1371,
       "step": 2135
     },
     {
       "epoch": 4.79,
       "learning_rate": 1.699421965317919e-05,
-      "loss": 0.1144,
+      "loss": 0.1127,
       "step": 2136
     },
     {
       "epoch": 4.79,
       "learning_rate": 1.682080924855491e-05,
-      "loss": 0.1283,
+      "loss": 0.1117,
       "step": 2137
     },
     {
       "epoch": 4.79,
       "learning_rate": 1.6647398843930635e-05,
-      "loss": 0.102,
+      "loss": 0.0863,
       "step": 2138
     },
     {
       "epoch": 4.8,
       "learning_rate": 1.6473988439306356e-05,
-      "loss": 0.0936,
+      "loss": 0.085,
       "step": 2139
     },
     {
       "epoch": 4.8,
       "learning_rate": 1.630057803468208e-05,
-      "loss": 0.0992,
+      "loss": 0.0834,
       "step": 2140
     },
     {
       "epoch": 4.8,
       "learning_rate": 1.61271676300578e-05,
-      "loss": 0.0878,
+      "loss": 0.0748,
       "step": 2141
     },
     {
       "epoch": 4.8,
       "learning_rate": 1.5953757225433525e-05,
-      "loss": 0.0724,
+      "loss": 0.0827,
       "step": 2142
     },
     {
       "epoch": 4.8,
       "learning_rate": 1.578034682080925e-05,
-      "loss": 0.0741,
+      "loss": 0.0786,
       "step": 2143
     },
     {
       "epoch": 4.81,
       "learning_rate": 1.560693641618497e-05,
-      "loss": 0.0743,
+      "loss": 0.0794,
       "step": 2144
     },
     {
       "epoch": 4.81,
       "learning_rate": 1.5433526011560694e-05,
-      "loss": 0.0643,
+      "loss": 0.0692,
       "step": 2145
     },
     {
       "epoch": 4.81,
       "learning_rate": 1.5260115606936414e-05,
-      "loss": 0.087,
+      "loss": 0.091,
       "step": 2146
     },
     {
       "epoch": 4.81,
       "learning_rate": 1.5086705202312138e-05,
-      "loss": 0.0482,
+      "loss": 0.0572,
       "step": 2147
     },
     {
       "epoch": 4.82,
       "learning_rate": 1.491329479768786e-05,
-      "loss": 0.0634,
+      "loss": 0.0749,
       "step": 2148
     },
     {
       "epoch": 4.82,
       "learning_rate": 1.4739884393063583e-05,
-      "loss": 0.0908,
+      "loss": 0.0841,
       "step": 2149
     },
     {
       "epoch": 4.82,
       "learning_rate": 1.4566473988439304e-05,
-      "loss": 0.0669,
+      "loss": 0.07,
       "step": 2150
     },
     {
       "epoch": 4.82,
       "learning_rate": 1.4393063583815026e-05,
-      "loss": 0.0904,
+      "loss": 0.0935,
       "step": 2151
     },
     {
       "epoch": 4.83,
       "learning_rate": 1.4219653179190749e-05,
-      "loss": 0.0609,
+      "loss": 0.0645,
       "step": 2152
     },
     {
       "epoch": 4.83,
       "learning_rate": 1.4046242774566473e-05,
-      "loss": 0.0588,
+      "loss": 0.4197,
       "step": 2153
     },
     {
       "epoch": 4.83,
       "learning_rate": 1.3872832369942195e-05,
-      "loss": 0.082,
+      "loss": 0.0902,
       "step": 2154
     },
     {
       "epoch": 4.83,
       "learning_rate": 1.3699421965317917e-05,
-      "loss": 0.0537,
+      "loss": 0.0462,
       "step": 2155
     },
     {
       "epoch": 4.83,
       "learning_rate": 1.352601156069364e-05,
-      "loss": 0.0742,
+      "loss": 0.0749,
       "step": 2156
     },
     {
       "epoch": 4.84,
       "learning_rate": 1.3352601156069362e-05,
-      "loss": 0.0523,
+      "loss": 0.0589,
       "step": 2157
     },
     {
       "epoch": 4.84,
       "learning_rate": 1.3179190751445084e-05,
-      "loss": 0.0743,
+      "loss": 0.0734,
       "step": 2158
     },
     {
       "epoch": 4.84,
       "learning_rate": 1.3005780346820809e-05,
-      "loss": 0.0412,
+      "loss": 0.0546,
       "step": 2159
     },
     {
       "epoch": 4.84,
       "learning_rate": 1.2832369942196531e-05,
-      "loss": 0.0736,
+      "loss": 0.0643,
       "step": 2160
     },
     {
       "epoch": 4.85,
       "learning_rate": 1.2658959537572253e-05,
-      "loss": 0.0472,
+      "loss": 0.0613,
       "step": 2161
     },
     {
       "epoch": 4.85,
       "learning_rate": 1.2485549132947976e-05,
-      "loss": 0.0592,
+      "loss": 0.0652,
       "step": 2162
     },
     {
       "epoch": 4.85,
       "learning_rate": 1.2312138728323698e-05,
-      "loss": 0.066,
+      "loss": 0.0587,
       "step": 2163
     },
     {
       "epoch": 4.85,
       "learning_rate": 1.213872832369942e-05,
-      "loss": 0.061,
+      "loss": 0.0617,
       "step": 2164
     },
     {
       "epoch": 4.85,
       "learning_rate": 1.1965317919075144e-05,
-      "loss": 0.0519,
+      "loss": 0.0527,
       "step": 2165
     },
     {
       "epoch": 4.86,
       "learning_rate": 1.1791907514450867e-05,
-      "loss": 0.0337,
+      "loss": 0.0402,
       "step": 2166
     },
     {
       "epoch": 4.86,
       "learning_rate": 1.161849710982659e-05,
-      "loss": 0.0496,
+      "loss": 0.0552,
       "step": 2167
     },
     {
       "epoch": 4.86,
       "learning_rate": 1.1445086705202312e-05,
-      "loss": 0.0457,
+      "loss": 0.0607,
       "step": 2168
     },
     {
       "epoch": 4.86,
       "learning_rate": 1.1271676300578034e-05,
-      "loss": 0.048,
+      "loss": 0.056,
       "step": 2169
     },
     {
       "epoch": 4.87,
       "learning_rate": 1.1098265895953756e-05,
-      "loss": 0.085,
+      "loss": 0.0579,
       "step": 2170
     },
     {
       "epoch": 4.87,
       "learning_rate": 1.092485549132948e-05,
-      "loss": 0.0576,
+      "loss": 0.0532,
       "step": 2171
     },
     {
       "epoch": 4.87,
       "learning_rate": 1.0751445086705203e-05,
-      "loss": 0.0437,
+      "loss": 0.0433,
       "step": 2172
     },
     {
       "epoch": 4.87,
       "learning_rate": 1.0578034682080925e-05,
-      "loss": 0.0581,
+      "loss": 0.0471,
       "step": 2173
     },
     {
       "epoch": 4.87,
       "learning_rate": 1.0404624277456646e-05,
-      "loss": 0.0467,
+      "loss": 0.043,
       "step": 2174
     },
     {
       "epoch": 4.88,
       "learning_rate": 1.0231213872832368e-05,
-      "loss": 0.0702,
+      "loss": 0.0672,
       "step": 2175
     },
     {
       "epoch": 4.88,
       "learning_rate": 1.005780346820809e-05,
-      "loss": 0.0588,
+      "loss": 0.0645,
       "step": 2176
     },
     {
       "epoch": 4.88,
       "learning_rate": 9.884393063583813e-06,
-      "loss": 0.0374,
+      "loss": 0.0419,
       "step": 2177
     },
     {
       "epoch": 4.88,
       "learning_rate": 9.710982658959535e-06,
-      "loss": 0.0616,
+      "loss": 0.0506,
       "step": 2178
     },
     {
       "epoch": 4.89,
       "learning_rate": 9.53757225433526e-06,
-      "loss": 0.0325,
+      "loss": 0.0312,
       "step": 2179
     },
     {
       "epoch": 4.89,
       "learning_rate": 9.364161849710982e-06,
-      "loss": 0.0427,
+      "loss": 0.0385,
       "step": 2180
     },
     {
       "epoch": 4.89,
       "learning_rate": 9.190751445086704e-06,
-      "loss": 0.0816,
+      "loss": 0.0981,
       "step": 2181
     },
     {
       "epoch": 4.89,
       "learning_rate": 9.017341040462426e-06,
-      "loss": 0.0509,
+      "loss": 0.0456,
       "step": 2182
     },
     {
       "epoch": 4.89,
       "learning_rate": 8.843930635838149e-06,
-      "loss": 0.0458,
+      "loss": 0.0773,
       "step": 2183
     },
     {
       "epoch": 4.9,
       "learning_rate": 8.670520231213871e-06,
-      "loss": 0.061,
+      "loss": 0.0392,
       "step": 2184
     },
     {
       "epoch": 4.9,
       "learning_rate": 8.497109826589595e-06,
-      "loss": 0.111,
+      "loss": 0.1151,
       "step": 2185
     },
     {
       "epoch": 4.9,
       "learning_rate": 8.323699421965318e-06,
-      "loss": 0.0942,
+      "loss": 0.0991,
       "step": 2186
     },
     {
       "epoch": 4.9,
       "learning_rate": 8.15028901734104e-06,
-      "loss": 0.0915,
+      "loss": 0.0904,
       "step": 2187
     },
     {
       "epoch": 4.91,
       "learning_rate": 7.976878612716762e-06,
-      "loss": 0.0627,
+      "loss": 0.0539,
       "step": 2188
     },
     {
       "epoch": 4.91,
       "learning_rate": 7.803468208092485e-06,
-      "loss": 0.0891,
+      "loss": 0.0813,
       "step": 2189
     },
     {
       "epoch": 4.91,
       "learning_rate": 7.630057803468207e-06,
-      "loss": 0.0673,
+      "loss": 0.0639,
       "step": 2190
     },
     {
       "epoch": 4.91,
       "learning_rate": 7.45664739884393e-06,
-      "loss": 0.087,
+      "loss": 0.0839,
       "step": 2191
     },
     {
       "epoch": 4.91,
       "learning_rate": 7.283236994219652e-06,
-      "loss": 0.0792,
+      "loss": 0.0821,
       "step": 2192
     },
     {
       "epoch": 4.92,
       "learning_rate": 7.109826589595374e-06,
-      "loss": 0.0868,
+      "loss": 0.085,
       "step": 2193
     },
     {
       "epoch": 4.92,
       "learning_rate": 6.9364161849710975e-06,
-      "loss": 0.0859,
+      "loss": 0.0905,
       "step": 2194
     },
     {
       "epoch": 4.92,
       "learning_rate": 6.76300578034682e-06,
-      "loss": 0.0631,
+      "loss": 0.0731,
       "step": 2195
     },
     {
       "epoch": 4.92,
       "learning_rate": 6.589595375722542e-06,
-      "loss": 0.0661,
+      "loss": 0.069,
       "step": 2196
     },
     {
       "epoch": 4.93,
       "learning_rate": 6.4161849710982654e-06,
-      "loss": 0.0472,
+      "loss": 0.0554,
       "step": 2197
     },
     {
       "epoch": 4.93,
       "learning_rate": 6.242774566473988e-06,
-      "loss": 0.0575,
+      "loss": 0.0552,
       "step": 2198
     },
     {
       "epoch": 4.93,
       "learning_rate": 6.06936416184971e-06,
-      "loss": 0.0572,
+      "loss": 0.0705,
       "step": 2199
     },
     {
       "epoch": 4.93,
       "learning_rate": 5.895953757225433e-06,
-      "loss": 0.0756,
+      "loss": 0.0627,
       "step": 2200
     },
     {
       "epoch": 4.93,
       "learning_rate": 5.722543352601156e-06,
-      "loss": 0.0709,
+      "loss": 0.0784,
       "step": 2201
     },
     {
       "epoch": 4.94,
       "learning_rate": 5.549132947976878e-06,
-      "loss": 0.0564,
+      "loss": 0.0665,
       "step": 2202
     },
     {
       "epoch": 4.94,
       "learning_rate": 5.375722543352601e-06,
-      "loss": 0.0701,
+      "loss": 0.0605,
       "step": 2203
     },
     {
       "epoch": 4.94,
       "learning_rate": 5.202312138728323e-06,
-      "loss": 0.0484,
+      "loss": 0.0544,
       "step": 2204
     },
     {
       "epoch": 4.94,
       "learning_rate": 5.028901734104045e-06,
-      "loss": 0.068,
+      "loss": 0.0643,
       "step": 2205
     },
     {
       "epoch": 4.95,
       "learning_rate": 4.855491329479768e-06,
-      "loss": 0.0546,
+      "loss": 0.0615,
       "step": 2206
     },
     {
       "epoch": 4.95,
       "learning_rate": 4.682080924855491e-06,
-      "loss": 0.0635,
+      "loss": 0.0734,
       "step": 2207
     },
     {
       "epoch": 4.95,
       "learning_rate": 4.508670520231213e-06,
-      "loss": 0.0693,
+      "loss": 0.0615,
       "step": 2208
     },
     {
       "epoch": 4.95,
       "learning_rate": 4.335260115606936e-06,
-      "loss": 0.0459,
+      "loss": 0.053,
       "step": 2209
     },
     {
       "epoch": 4.96,
       "learning_rate": 4.161849710982659e-06,
-      "loss": 0.0561,
+      "loss": 0.0534,
       "step": 2210
     },
     {
       "epoch": 4.96,
       "learning_rate": 3.988439306358381e-06,
-      "loss": 0.0614,
+      "loss": 0.0597,
       "step": 2211
     },
     {
       "epoch": 4.96,
       "learning_rate": 3.8150289017341036e-06,
-      "loss": 0.0694,
+      "loss": 0.0839,
       "step": 2212
     },
     {
       "epoch": 4.96,
       "learning_rate": 3.641618497109826e-06,
-      "loss": 0.0504,
+      "loss": 0.0639,
       "step": 2213
     },
     {
       "epoch": 4.96,
       "learning_rate": 3.4682080924855487e-06,
-      "loss": 0.0363,
+      "loss": 0.0344,
       "step": 2214
     },
     {
       "epoch": 4.97,
       "learning_rate": 3.294797687861271e-06,
-      "loss": 0.064,
+      "loss": 0.0682,
       "step": 2215
     },
     {
       "epoch": 4.97,
       "learning_rate": 3.121387283236994e-06,
-      "loss": 0.0734,
+      "loss": 0.0535,
       "step": 2216
     },
     {
       "epoch": 4.97,
       "learning_rate": 2.9479768786127167e-06,
-      "loss": 0.0597,
+      "loss": 0.06,
       "step": 2217
     },
     {
       "epoch": 4.97,
       "learning_rate": 2.774566473988439e-06,
-      "loss": 0.0746,
+      "loss": 0.0719,
       "step": 2218
     },
     {
       "epoch": 4.98,
       "learning_rate": 2.6011560693641614e-06,
-      "loss": 0.0282,
+      "loss": 0.0284,
       "step": 2219
     },
     {
       "epoch": 4.98,
       "learning_rate": 2.427745664739884e-06,
-      "loss": 0.0546,
+      "loss": 0.0614,
       "step": 2220
     },
     {
       "epoch": 4.98,
       "learning_rate": 2.2543352601156066e-06,
-      "loss": 0.0621,
+      "loss": 0.0555,
       "step": 2221
     },
     {
       "epoch": 4.98,
       "learning_rate": 2.0809248554913294e-06,
-      "loss": 0.0722,
+      "loss": 0.0706,
       "step": 2222
     },
     {
       "epoch": 4.98,
       "learning_rate": 1.9075144508670518e-06,
-      "loss": 0.0788,
+      "loss": 0.0718,
       "step": 2223
     },
     {
       "epoch": 4.99,
       "learning_rate": 1.7341040462427744e-06,
-      "loss": 0.0485,
+      "loss": 0.0511,
       "step": 2224
     },
     {
       "epoch": 4.99,
       "learning_rate": 1.560693641618497e-06,
-      "loss": 0.0575,
+      "loss": 0.0561,
       "step": 2225
     },
     {
       "epoch": 4.99,
       "learning_rate": 1.3872832369942195e-06,
-      "loss": 0.0417,
+      "loss": 0.0437,
       "step": 2226
     },
     {
       "epoch": 4.99,
       "learning_rate": 1.213872832369942e-06,
-      "loss": 0.0393,
+      "loss": 0.055,
       "step": 2227
     },
     {
       "epoch": 5.0,
       "learning_rate": 1.0404624277456647e-06,
-      "loss": 0.0428,
+      "loss": 0.0464,
       "step": 2228
     },
     {
       "epoch": 5.0,
       "learning_rate": 8.670520231213872e-07,
-      "loss": 0.0489,
+      "loss": 0.048,
       "step": 2229
     },
     {
       "epoch": 5.0,
       "learning_rate": 6.936416184971098e-07,
-      "loss": 0.1199,
+      "loss": 0.0607,
       "step": 2230
     },
     {
       "epoch": 5.0,
       "step": 2230,
       "total_flos": 0.0,
-      "train_loss": 1.4829948718221064,
-      "train_runtime": 27349.9818,
-      "train_samples_per_second": 5.217,
-      "train_steps_per_second": 0.082
+      "train_loss": 1.6266609489166015,
+      "train_runtime": 26664.6603,
+      "train_samples_per_second": 5.351,
+      "train_steps_per_second": 0.084
     }
   ],
   "max_steps": 2230,