diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,54651 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 9.0,
+  "eval_steps": 500,
+  "global_step": 7803,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.510638297872341e-08,
+      "loss": 8.42,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7021276595744683e-07,
+      "loss": 7.8116,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2.553191489361702e-07,
+      "loss": 7.3405,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4042553191489365e-07,
+      "loss": 7.9831,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 8.8292,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 5.106382978723404e-07,
+      "loss": 9.3832,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 5.957446808510639e-07,
+      "loss": 8.4834,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 6.808510638297873e-07,
+      "loss": 8.338,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 7.659574468085107e-07,
+      "loss": 7.8708,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 8.1726,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 9.361702127659575e-07,
+      "loss": 8.7641,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0212765957446809e-06,
+      "loss": 9.4236,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1063829787234042e-06,
+      "loss": 8.0038,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1914893617021278e-06,
+      "loss": 6.9279,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 7.1003,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3617021276595746e-06,
+      "loss": 7.8895,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4468085106382978e-06,
+      "loss": 7.5029,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5319148936170214e-06,
+      "loss": 8.1941,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.617021276595745e-06,
+      "loss": 6.7967,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 7.0943,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7872340425531918e-06,
+      "loss": 6.9398,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.872340425531915e-06,
+      "loss": 6.6345,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9574468085106385e-06,
+      "loss": 5.5552,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0425531914893617e-06,
+      "loss": 5.6271,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 5.9879,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2127659574468085e-06,
+      "loss": 5.2435,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.297872340425532e-06,
+      "loss": 6.1067,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3829787234042557e-06,
+      "loss": 5.6905,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.468085106382979e-06,
+      "loss": 5.4866,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 4.0927,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6382978723404256e-06,
+      "loss": 3.1946,
+      "step": 31
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7234042553191492e-06,
+      "loss": 3.323,
+      "step": 32
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.808510638297873e-06,
+      "loss": 3.615,
+      "step": 33
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8936170212765956e-06,
+      "loss": 3.2357,
+      "step": 34
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 2.8849,
+      "step": 35
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0638297872340428e-06,
+      "loss": 3.1966,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1489361702127664e-06,
+      "loss": 2.8122,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 3.23404255319149e-06,
+      "loss": 2.3877,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3191489361702127e-06,
+      "loss": 2.4174,
+      "step": 39
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 2.6009,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 3.48936170212766e-06,
+      "loss": 2.7382,
+      "step": 41
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5744680851063835e-06,
+      "loss": 2.5757,
+      "step": 42
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6595744680851063e-06,
+      "loss": 2.1993,
+      "step": 43
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 3.74468085106383e-06,
+      "loss": 2.6667,
+      "step": 44
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 2.055,
+      "step": 45
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 3.914893617021277e-06,
+      "loss": 2.4098,
+      "step": 46
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.4524,
+      "step": 47
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.085106382978723e-06,
+      "loss": 2.8391,
+      "step": 48
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.170212765957447e-06,
+      "loss": 2.6667,
+      "step": 49
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 2.2407,
+      "step": 50
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.340425531914894e-06,
+      "loss": 1.7787,
+      "step": 51
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.425531914893617e-06,
+      "loss": 2.8719,
+      "step": 52
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5106382978723406e-06,
+      "loss": 2.6247,
+      "step": 53
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.595744680851064e-06,
+      "loss": 2.2297,
+      "step": 54
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 2.1361,
+      "step": 55
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0,
+      "learning_rate": 4.765957446808511e-06,
+      "loss": 1.8228,
+      "step": 56
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.851063829787234e-06,
+      "loss": 2.651,
+      "step": 57
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 4.936170212765958e-06,
+      "loss": 1.9376,
+      "step": 58
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0212765957446805e-06,
+      "loss": 2.1441,
+      "step": 59
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 2.5828,
+      "step": 60
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.191489361702128e-06,
+      "loss": 1.6916,
+      "step": 61
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.276595744680851e-06,
+      "loss": 1.7109,
+      "step": 62
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.361702127659575e-06,
+      "loss": 2.2316,
+      "step": 63
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4468085106382985e-06,
+      "loss": 2.2679,
+      "step": 64
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 2.2825,
+      "step": 65
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.617021276595746e-06,
+      "loss": 1.9097,
+      "step": 66
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.702127659574469e-06,
+      "loss": 2.2865,
+      "step": 67
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.787234042553191e-06,
+      "loss": 2.6797,
+      "step": 68
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.872340425531915e-06,
+      "loss": 1.9614,
+      "step": 69
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 1.838,
+      "step": 70
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 6.042553191489362e-06,
+      "loss": 2.1008,
+      "step": 71
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 6.1276595744680855e-06,
+      "loss": 1.9877,
+      "step": 72
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0,
+      "learning_rate": 6.212765957446809e-06,
+      "loss": 2.0037,
+      "step": 73
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 6.297872340425533e-06,
+      "loss": 1.7601,
+      "step": 74
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 2.4598,
+      "step": 75
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 6.46808510638298e-06,
+      "loss": 1.4717,
+      "step": 76
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 6.553191489361702e-06,
+      "loss": 1.7133,
+      "step": 77
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 6.6382978723404254e-06,
+      "loss": 2.097,
+      "step": 78
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 6.723404255319149e-06,
+      "loss": 1.9819,
+      "step": 79
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 1.755,
+      "step": 80
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 6.893617021276596e-06,
+      "loss": 2.5593,
+      "step": 81
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0,
+      "learning_rate": 6.97872340425532e-06,
+      "loss": 1.3687,
+      "step": 82
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 7.0638297872340434e-06,
+      "loss": 1.637,
+      "step": 83
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 7.148936170212767e-06,
+      "loss": 2.5012,
+      "step": 84
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 2.1997,
+      "step": 85
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 7.3191489361702125e-06,
+      "loss": 1.897,
+      "step": 86
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 7.404255319148936e-06,
+      "loss": 1.3258,
+      "step": 87
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 7.48936170212766e-06,
+      "loss": 1.9583,
+      "step": 88
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 7.574468085106383e-06,
+      "loss": 1.4425,
+      "step": 89
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 1.9072,
+      "step": 90
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0,
+      "learning_rate": 7.74468085106383e-06,
+      "loss": 1.5434,
+      "step": 91
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 7.829787234042554e-06,
+      "loss": 1.3598,
+      "step": 92
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 7.914893617021278e-06,
+      "loss": 1.3292,
+      "step": 93
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.7655,
+      "step": 94
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 2.3734,
+      "step": 95
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.170212765957447e-06,
+      "loss": 2.3653,
+      "step": 96
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.25531914893617e-06,
+      "loss": 2.0333,
+      "step": 97
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.340425531914894e-06,
+      "loss": 1.5327,
+      "step": 98
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.425531914893618e-06,
+      "loss": 2.2848,
+      "step": 99
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 2.0775,
+      "step": 100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.595744680851065e-06,
+      "loss": 1.9294,
+      "step": 101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.680851063829788e-06,
+      "loss": 1.5409,
+      "step": 102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.765957446808512e-06,
+      "loss": 1.7937,
+      "step": 103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.851063829787234e-06,
+      "loss": 1.7683,
+      "step": 104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 1.7788,
+      "step": 105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 9.021276595744681e-06,
+      "loss": 2.0132,
+      "step": 106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 9.106382978723405e-06,
+      "loss": 2.4136,
+      "step": 107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0,
+      "learning_rate": 9.191489361702128e-06,
+      "loss": 1.56,
+      "step": 108
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 9.276595744680852e-06,
+      "loss": 1.6401,
+      "step": 109
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 2.573,
+      "step": 110
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 9.446808510638299e-06,
+      "loss": 1.4836,
+      "step": 111
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 9.531914893617023e-06,
+      "loss": 1.5127,
+      "step": 112
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 9.617021276595745e-06,
+      "loss": 1.5648,
+      "step": 113
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 9.702127659574468e-06,
+      "loss": 2.0786,
+      "step": 114
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 2.1154,
+      "step": 115
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 9.872340425531915e-06,
+      "loss": 1.9178,
+      "step": 116
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0,
+      "learning_rate": 9.957446808510639e-06,
+      "loss": 2.1499,
+      "step": 117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0042553191489361e-05,
+      "loss": 1.2971,
+      "step": 118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0127659574468085e-05,
+      "loss": 1.4621,
+      "step": 119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0212765957446808e-05,
+      "loss": 2.2563,
+      "step": 120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0297872340425532e-05,
+      "loss": 1.3251,
+      "step": 121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0382978723404255e-05,
+      "loss": 2.2923,
+      "step": 122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0468085106382979e-05,
+      "loss": 1.4845,
+      "step": 123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0553191489361703e-05,
+      "loss": 1.5756,
+      "step": 124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0638297872340426e-05,
+      "loss": 1.4902,
+      "step": 125
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.072340425531915e-05,
+      "loss": 1.4426,
+      "step": 126
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0808510638297873e-05,
+      "loss": 1.921,
+      "step": 127
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0893617021276597e-05,
+      "loss": 2.1011,
+      "step": 128
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.097872340425532e-05,
+      "loss": 2.0556,
+      "step": 129
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1063829787234044e-05,
+      "loss": 2.0671,
+      "step": 130
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1148936170212768e-05,
+      "loss": 1.2012,
+      "step": 131
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1234042553191491e-05,
+      "loss": 1.75,
+      "step": 132
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1319148936170215e-05,
+      "loss": 1.2119,
+      "step": 133
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1404255319148939e-05,
+      "loss": 1.5734,
+      "step": 134
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1489361702127662e-05,
+      "loss": 1.6814,
+      "step": 135
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1574468085106382e-05,
+      "loss": 2.0887,
+      "step": 136
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1659574468085106e-05,
+      "loss": 1.3118,
+      "step": 137
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.174468085106383e-05,
+      "loss": 1.397,
+      "step": 138
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1829787234042553e-05,
+      "loss": 1.7979,
+      "step": 139
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1914893617021277e-05,
+      "loss": 1.8423,
+      "step": 140
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2e-05,
+      "loss": 2.1667,
+      "step": 141
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2085106382978724e-05,
+      "loss": 1.955,
+      "step": 142
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2170212765957448e-05,
+      "loss": 1.7638,
+      "step": 143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2255319148936171e-05,
+      "loss": 1.3332,
+      "step": 144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2340425531914895e-05,
+      "loss": 1.602,
+      "step": 145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2425531914893618e-05,
+      "loss": 1.72,
+      "step": 146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2510638297872342e-05,
+      "loss": 1.2205,
+      "step": 147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2595744680851065e-05,
+      "loss": 1.9567,
+      "step": 148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2680851063829789e-05,
+      "loss": 1.5785,
+      "step": 149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2765957446808513e-05,
+      "loss": 1.58,
+      "step": 150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2851063829787236e-05,
+      "loss": 1.8026,
+      "step": 151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.293617021276596e-05,
+      "loss": 1.8456,
+      "step": 152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3021276595744683e-05,
+      "loss": 1.7062,
+      "step": 153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3106382978723404e-05,
+      "loss": 2.0579,
+      "step": 154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3191489361702127e-05,
+      "loss": 1.9036,
+      "step": 155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3276595744680851e-05,
+      "loss": 1.3548,
+      "step": 156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3361702127659574e-05,
+      "loss": 1.212,
+      "step": 157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3446808510638298e-05,
+      "loss": 1.9522,
+      "step": 158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3531914893617022e-05,
+      "loss": 1.2715,
+      "step": 159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3617021276595745e-05,
+      "loss": 1.8048,
+      "step": 160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3702127659574469e-05,
+      "loss": 1.4929,
+      "step": 161
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3787234042553192e-05,
+      "loss": 1.8582,
+      "step": 162
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3872340425531916e-05,
+      "loss": 1.1246,
+      "step": 163
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.395744680851064e-05,
+      "loss": 1.5867,
+      "step": 164
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4042553191489363e-05,
+      "loss": 1.1973,
+      "step": 165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4127659574468087e-05,
+      "loss": 1.5066,
+      "step": 166
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.421276595744681e-05,
+      "loss": 1.782,
+      "step": 167
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4297872340425534e-05,
+      "loss": 1.2071,
+      "step": 168
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4382978723404258e-05,
+      "loss": 1.9611,
+      "step": 169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4468085106382981e-05,
+      "loss": 1.7584,
+      "step": 170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4553191489361705e-05,
+      "loss": 1.521,
+      "step": 171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4638297872340425e-05,
+      "loss": 1.6986,
+      "step": 172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4723404255319149e-05,
+      "loss": 1.2474,
+      "step": 173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4808510638297872e-05,
+      "loss": 1.3479,
+      "step": 174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4893617021276596e-05,
+      "loss": 1.3141,
+      "step": 175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.497872340425532e-05,
+      "loss": 2.3654,
+      "step": 176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5063829787234043e-05,
+      "loss": 1.9257,
+      "step": 177
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5148936170212767e-05,
+      "loss": 1.6664,
+      "step": 178
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.523404255319149e-05,
+      "loss": 1.9523,
+      "step": 179
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5319148936170214e-05,
+      "loss": 1.4715,
+      "step": 180
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5404255319148937e-05,
+      "loss": 1.0523,
+      "step": 181
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.548936170212766e-05,
+      "loss": 1.7935,
+      "step": 182
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5574468085106385e-05,
+      "loss": 1.818,
+      "step": 183
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5659574468085108e-05,
+      "loss": 1.5303,
+      "step": 184
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5744680851063832e-05,
+      "loss": 1.4181,
+      "step": 185
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5829787234042555e-05,
+      "loss": 1.4132,
+      "step": 186
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.591489361702128e-05,
+      "loss": 1.2468,
+      "step": 187
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.3493,
+      "step": 188
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6085106382978726e-05,
+      "loss": 1.7436,
+      "step": 189
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6170212765957446e-05,
+      "loss": 2.1027,
+      "step": 190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.625531914893617e-05,
+      "loss": 1.6055,
+      "step": 191
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6340425531914894e-05,
+      "loss": 1.417,
+      "step": 192
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6425531914893617e-05,
+      "loss": 1.2369,
+      "step": 193
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.651063829787234e-05,
+      "loss": 1.3705,
+      "step": 194
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6595744680851064e-05,
+      "loss": 1.5408,
+      "step": 195
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6680851063829788e-05,
+      "loss": 0.8064,
+      "step": 196
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.676595744680851e-05,
+      "loss": 1.2991,
+      "step": 197
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6851063829787235e-05,
+      "loss": 2.015,
+      "step": 198
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.693617021276596e-05,
+      "loss": 1.8201,
+      "step": 199
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7021276595744682e-05,
+      "loss": 2.1389,
+      "step": 200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7106382978723406e-05,
+      "loss": 1.0459,
+      "step": 201
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.719148936170213e-05,
+      "loss": 1.6487,
+      "step": 202
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7276595744680853e-05,
+      "loss": 1.6784,
+      "step": 203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7361702127659577e-05,
+      "loss": 1.6428,
+      "step": 204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.74468085106383e-05,
+      "loss": 1.2589,
+      "step": 205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7531914893617024e-05,
+      "loss": 1.5843,
+      "step": 206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7617021276595748e-05,
+      "loss": 1.0254,
+      "step": 207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7702127659574468e-05,
+      "loss": 2.0064,
+      "step": 208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.778723404255319e-05,
+      "loss": 1.6273,
+      "step": 209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7872340425531915e-05,
+      "loss": 1.2627,
+      "step": 210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.795744680851064e-05,
+      "loss": 1.7395,
+      "step": 211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8042553191489362e-05,
+      "loss": 2.056,
+      "step": 212
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8127659574468086e-05,
+      "loss": 1.7548,
+      "step": 213
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.821276595744681e-05,
+      "loss": 1.0854,
+      "step": 214
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8297872340425533e-05,
+      "loss": 1.5037,
+      "step": 215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8382978723404257e-05,
+      "loss": 1.3276,
+      "step": 216
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.846808510638298e-05,
+      "loss": 1.1777,
+      "step": 217
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8553191489361704e-05,
+      "loss": 1.7991,
+      "step": 218
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8638297872340427e-05,
+      "loss": 1.7696,
+      "step": 219
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.872340425531915e-05,
+      "loss": 1.7687,
+      "step": 220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8808510638297875e-05,
+      "loss": 0.9113,
+      "step": 221
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8893617021276598e-05,
+      "loss": 1.703,
+      "step": 222
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8978723404255322e-05,
+      "loss": 1.2613,
+      "step": 223
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9063829787234045e-05,
+      "loss": 1.0527,
+      "step": 224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.914893617021277e-05,
+      "loss": 1.2803,
+      "step": 225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.923404255319149e-05,
+      "loss": 1.5915,
+      "step": 226
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9319148936170213e-05,
+      "loss": 1.4361,
+      "step": 227
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9404255319148936e-05,
+      "loss": 1.8447,
+      "step": 228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.948936170212766e-05,
+      "loss": 1.1079,
+      "step": 229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9574468085106384e-05,
+      "loss": 1.5217,
+      "step": 230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9659574468085107e-05,
+      "loss": 1.8054,
+      "step": 231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.974468085106383e-05,
+      "loss": 1.7943,
+      "step": 232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9829787234042554e-05,
+      "loss": 1.4356,
+      "step": 233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9914893617021278e-05,
+      "loss": 1.382,
+      "step": 234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 2e-05,
+      "loss": 1.3556,
+      "step": 235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999999138396417e-05,
+      "loss": 1.885,
+      "step": 236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999996553585817e-05,
+      "loss": 1.494,
+      "step": 237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999992245568637e-05,
+      "loss": 1.2221,
+      "step": 238
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999998621434563e-05,
+      "loss": 1.6283,
+      "step": 239
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999997845991783e-05,
+      "loss": 1.3918,
+      "step": 240
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999968982286576e-05,
+      "loss": 2.0266,
+      "step": 241
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99999577814535e-05,
+      "loss": 2.1751,
+      "step": 242
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999944857420527e-05,
+      "loss": 1.5996,
+      "step": 243
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999930210189894e-05,
+      "loss": 2.0769,
+      "step": 244
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999913839764117e-05,
+      "loss": 1.9959,
+      "step": 245
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999989574614602e-05,
+      "loss": 1.4403,
+      "step": 246
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999875929338718e-05,
+      "loss": 1.646,
+      "step": 247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999854389345632e-05,
+      "loss": 0.9902,
+      "step": 248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999983112617047e-05,
+      "loss": 1.5002,
+      "step": 249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999806139817236e-05,
+      "loss": 1.1182,
+      "step": 250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999779430290247e-05,
+      "loss": 1.5147,
+      "step": 251
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999750997594095e-05,
+      "loss": 2.1442,
+      "step": 252
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999720841733688e-05,
+      "loss": 1.7654,
+      "step": 253
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999688962714217e-05,
+      "loss": 2.5233,
+      "step": 254
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999965536054117e-05,
+      "loss": 1.7948,
+      "step": 255
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999620035220354e-05,
+      "loss": 1.2109,
+      "step": 256
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999958298675784e-05,
+      "loss": 1.3188,
+      "step": 257
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999954421516002e-05,
+      "loss": 1.6432,
+      "step": 258
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999503720433575e-05,
+      "loss": 0.8701,
+      "step": 259
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999946150258548e-05,
+      "loss": 1.3864,
+      "step": 260
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999941756162301e-05,
+      "loss": 1.8293,
+      "step": 261
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999371897553744e-05,
+      "loss": 1.2499,
+      "step": 262
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999324510385543e-05,
+      "loss": 1.7406,
+      "step": 263
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999275400126575e-05,
+      "loss": 1.581,
+      "step": 264
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999224566785302e-05,
+      "loss": 1.6417,
+      "step": 265
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999172010370485e-05,
+      "loss": 1.1926,
+      "step": 266
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999911773089118e-05,
+      "loss": 1.6034,
+      "step": 267
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999061728356743e-05,
+      "loss": 1.0279,
+      "step": 268
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999004002776822e-05,
+      "loss": 1.8958,
+      "step": 269
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9998944554161365e-05,
+      "loss": 1.2075,
+      "step": 270
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999888338252061e-05,
+      "loss": 1.7594,
+      "step": 271
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999882048786511e-05,
+      "loss": 1.6885,
+      "step": 272
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9998755870205694e-05,
+      "loss": 1.6296,
+      "step": 273
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99986895295535e-05,
+      "loss": 1.456,
+      "step": 274
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999862146591996e-05,
+      "loss": 1.208,
+      "step": 275
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9998551679316803e-05,
+      "loss": 2.0674,
+      "step": 276
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9998480169756052e-05,
+      "loss": 1.5543,
+      "step": 277
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9998406937250035e-05,
+      "loss": 1.3525,
+      "step": 278
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999833198181137e-05,
+      "loss": 1.1776,
+      "step": 279
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9998255303452963e-05,
+      "loss": 2.0884,
+      "step": 280
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9998176902188042e-05,
+      "loss": 1.1698,
+      "step": 281
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9998096778030113e-05,
+      "loss": 1.2396,
+      "step": 282
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9998014930992976e-05,
+      "loss": 1.3916,
+      "step": 283
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999793136109074e-05,
+      "loss": 0.9712,
+      "step": 284
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999784606833781e-05,
+      "loss": 1.1341,
+      "step": 285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9997759052748877e-05,
+      "loss": 1.0984,
+      "step": 286
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9997670314338936e-05,
+      "loss": 1.3076,
+      "step": 287
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9997579853123286e-05,
+      "loss": 1.0348,
+      "step": 288
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9997487669117506e-05,
+      "loss": 1.7866,
+      "step": 289
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9997393762337486e-05,
+      "loss": 1.9494,
+      "step": 290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9997298132799408e-05,
+      "loss": 1.535,
+      "step": 291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999720078051975e-05,
+      "loss": 1.5591,
+      "step": 292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999710170551529e-05,
+      "loss": 1.347,
+      "step": 293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9997000907803094e-05,
+      "loss": 1.3367,
+      "step": 294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999689838740054e-05,
+      "loss": 1.5453,
+      "step": 295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9996794144325288e-05,
+      "loss": 1.3461,
+      "step": 296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9996688178595303e-05,
+      "loss": 2.0435,
+      "step": 297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9996580490228846e-05,
+      "loss": 1.3144,
+      "step": 298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9996471079244477e-05,
+      "loss": 2.4926,
+      "step": 299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999635994566105e-05,
+      "loss": 1.7137,
+      "step": 300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9996247089497703e-05,
+      "loss": 1.6097,
+      "step": 301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99961325107739e-05,
+      "loss": 1.0879,
+      "step": 302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999601620950937e-05,
+      "loss": 1.4578,
+      "step": 303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9995898185724168e-05,
+      "loss": 1.5959,
+      "step": 304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999577843943863e-05,
+      "loss": 1.4501,
+      "step": 305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999565697067338e-05,
+      "loss": 1.4152,
+      "step": 306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999553377944936e-05,
+      "loss": 2.0109,
+      "step": 307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9995408865787792e-05,
+      "loss": 1.4868,
+      "step": 308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9995282229710208e-05,
+      "loss": 1.9465,
+      "step": 309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999515387123842e-05,
+      "loss": 1.0985,
+      "step": 310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9995023790394558e-05,
+      "loss": 1.9274,
+      "step": 311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9994891987201033e-05,
+      "loss": 1.8448,
+      "step": 312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999475846168055e-05,
+      "loss": 1.1021,
+      "step": 313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9994623213856134e-05,
+      "loss": 1.3757,
+      "step": 314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9994486243751076e-05,
+      "loss": 2.0586,
+      "step": 315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999434755138899e-05,
+      "loss": 1.4009,
+      "step": 316
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9994207136793763e-05,
+      "loss": 1.5505,
+      "step": 317
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9994064999989605e-05,
+      "loss": 1.1985,
+      "step": 318
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9993921141001003e-05,
+      "loss": 1.0187,
+      "step": 319
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9993775559852745e-05,
+      "loss": 1.3077,
+      "step": 320
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999362825656992e-05,
+      "loss": 2.1929,
+      "step": 321
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999347923117791e-05,
+      "loss": 0.9928,
+      "step": 322
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9993328483702393e-05,
+      "loss": 1.1146,
+      "step": 323
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9993176014169355e-05,
+      "loss": 0.8433,
+      "step": 324
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999302182260506e-05,
+      "loss": 1.1088,
+      "step": 325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9992865909036085e-05,
+      "loss": 0.9518,
+      "step": 326
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999270827348929e-05,
+      "loss": 1.8801,
+      "step": 327
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9992548915991844e-05,
+      "loss": 1.4719,
+      "step": 328
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999238783657121e-05,
+      "loss": 1.4299,
+      "step": 329
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999222503525514e-05,
+      "loss": 1.4219,
+      "step": 330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999206051207169e-05,
+      "loss": 1.4356,
+      "step": 331
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999189426704921e-05,
+      "loss": 1.3987,
+      "step": 332
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999172630021635e-05,
+      "loss": 1.4225,
+      "step": 333
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999155661160205e-05,
+      "loss": 1.519,
+      "step": 334
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9991385201235552e-05,
+      "loss": 1.2867,
+      "step": 335
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99912120691464e-05,
+      "loss": 1.5244,
+      "step": 336
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999103721536442e-05,
+      "loss": 1.342,
+      "step": 337
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9990860639919743e-05,
+      "loss": 0.8987,
+      "step": 338
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9990682342842805e-05,
+      "loss": 1.2,
+      "step": 339
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999050232416432e-05,
+      "loss": 1.9059,
+      "step": 340
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999032058391532e-05,
+      "loss": 0.9323,
+      "step": 341
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999013712212711e-05,
+      "loss": 1.6117,
+      "step": 342
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9989951938831315e-05,
+      "loss": 1.609,
+      "step": 343
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998976503405984e-05,
+      "loss": 1.3835,
+      "step": 344
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9989576407844894e-05,
+      "loss": 1.3594,
+      "step": 345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998938606021898e-05,
+      "loss": 1.0871,
+      "step": 346
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99891939912149e-05,
+      "loss": 1.7345,
+      "step": 347
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9989000200865757e-05,
+      "loss": 1.2132,
+      "step": 348
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9988804689204934e-05,
+      "loss": 1.4724,
+      "step": 349
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9988607456266128e-05,
+      "loss": 0.9844,
+      "step": 350
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9988408502083328e-05,
+      "loss": 1.7254,
+      "step": 351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9988207826690816e-05,
+      "loss": 1.6512,
+      "step": 352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998800543012317e-05,
+      "loss": 0.9675,
+      "step": 353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9987801312415273e-05,
+      "loss": 1.59,
+      "step": 354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9987595473602292e-05,
+      "loss": 1.7345,
+      "step": 355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9987387913719698e-05,
+      "loss": 1.6734,
+      "step": 356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9987178632803265e-05,
+      "loss": 1.5462,
+      "step": 357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9986967630889048e-05,
+      "loss": 1.473,
+      "step": 358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9986754908013415e-05,
+      "loss": 1.8486,
+      "step": 359
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9986540464213014e-05,
+      "loss": 1.7605,
+      "step": 360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9986324299524807e-05,
+      "loss": 0.7707,
+      "step": 361
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9986106413986036e-05,
+      "loss": 1.3303,
+      "step": 362
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9985886807634246e-05,
+      "loss": 1.7571,
+      "step": 363
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998566548050729e-05,
+      "loss": 1.8942,
+      "step": 364
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99854424326433e-05,
+      "loss": 1.0248,
+      "step": 365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998521766408071e-05,
+      "loss": 1.6188,
+      "step": 366
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998499117485826e-05,
+      "loss": 1.1917,
+      "step": 367
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998476296501497e-05,
+      "loss": 1.3675,
+      "step": 368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998453303459017e-05,
+      "loss": 2.3889,
+      "step": 369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9984301383623482e-05,
+      "loss": 1.6476,
+      "step": 370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9984068012154824e-05,
+      "loss": 1.444,
+      "step": 371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998383292022441e-05,
+      "loss": 1.7824,
+      "step": 372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9983596107872748e-05,
+      "loss": 2.0302,
+      "step": 373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9983357575140655e-05,
+      "loss": 1.5296,
+      "step": 374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9983117322069226e-05,
+      "loss": 1.4001,
+      "step": 375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9982875348699862e-05,
+      "loss": 0.8302,
+      "step": 376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9982631655074264e-05,
+      "loss": 1.6376,
+      "step": 377
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9982386241234426e-05,
+      "loss": 1.229,
+      "step": 378
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9982139107222634e-05,
+      "loss": 1.2652,
+      "step": 379
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9981890253081476e-05,
+      "loss": 1.742,
+      "step": 380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998163967885384e-05,
+      "loss": 0.9768,
+      "step": 381
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9981387384582895e-05,
+      "loss": 1.5925,
+      "step": 382
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9981133370312123e-05,
+      "loss": 1.7362,
+      "step": 383
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9980877636085295e-05,
+      "loss": 1.3239,
+      "step": 384
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9980620181946476e-05,
+      "loss": 1.236,
+      "step": 385
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998036100794004e-05,
+      "loss": 1.1723,
+      "step": 386
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9980100114110637e-05,
+      "loss": 1.4548,
+      "step": 387
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997983750050323e-05,
+      "loss": 1.3649,
+      "step": 388
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997957316716307e-05,
+      "loss": 1.6942,
+      "step": 389
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997930711413571e-05,
+      "loss": 1.2849,
+      "step": 390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9979039341466997e-05,
+      "loss": 1.5228,
+      "step": 391
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9978769849203074e-05,
+      "loss": 1.2276,
+      "step": 392
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9978498637390375e-05,
+      "loss": 1.4343,
+      "step": 393
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997822570607564e-05,
+      "loss": 1.3938,
+      "step": 394
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99779510553059e-05,
+      "loss": 1.3809,
+      "step": 395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997767468512848e-05,
+      "loss": 1.096,
+      "step": 396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997739659559101e-05,
+      "loss": 1.6983,
+      "step": 397
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9977116786741406e-05,
+      "loss": 1.3872,
+      "step": 398
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9976835258627884e-05,
+      "loss": 1.2209,
+      "step": 399
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997655201129896e-05,
+      "loss": 1.171,
+      "step": 400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9976267044803442e-05,
+      "loss": 1.4436,
+      "step": 401
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997598035919044e-05,
+      "loss": 1.0627,
+      "step": 402
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9975691954509347e-05,
+      "loss": 1.4267,
+      "step": 403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9975401830809872e-05,
+      "loss": 1.3074,
+      "step": 404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9975109988142e-05,
+      "loss": 1.0252,
+      "step": 405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997481642655603e-05,
+      "loss": 1.1064,
+      "step": 406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9974521146102535e-05,
+      "loss": 2.0766,
+      "step": 407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9974224146832414e-05,
+      "loss": 1.4027,
+      "step": 408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9973925428796837e-05,
+      "loss": 1.5382,
+      "step": 409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9973624992047282e-05,
+      "loss": 2.0154,
+      "step": 410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9973322836635517e-05,
+      "loss": 1.1344,
+      "step": 411
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9973018962613614e-05,
+      "loss": 1.9179,
+      "step": 412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9972713370033937e-05,
+      "loss": 1.2309,
+      "step": 413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9972406058949145e-05,
+      "loss": 1.2499,
+      "step": 414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997209702941219e-05,
+      "loss": 1.5624,
+      "step": 415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9971786281476325e-05,
+      "loss": 1.4585,
+      "step": 416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9971473815195106e-05,
+      "loss": 1.3946,
+      "step": 417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997115963062237e-05,
+      "loss": 2.0863,
+      "step": 418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997084372781226e-05,
+      "loss": 1.421,
+      "step": 419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9970526106819212e-05,
+      "loss": 1.5456,
+      "step": 420
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9970206767697958e-05,
+      "loss": 1.3417,
+      "step": 421
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996988571050353e-05,
+      "loss": 1.8652,
+      "step": 422
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9969562935291248e-05,
+      "loss": 1.0431,
+      "step": 423
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9969238442116737e-05,
+      "loss": 1.0137,
+      "step": 424
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996891223103591e-05,
+      "loss": 1.5943,
+      "step": 425
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9968584302104987e-05,
+      "loss": 1.6267,
+      "step": 426
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9968254655380465e-05,
+      "loss": 1.5033,
+      "step": 427
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996792329091916e-05,
+      "loss": 0.8204,
+      "step": 428
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996759020877817e-05,
+      "loss": 1.4775,
+      "step": 429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9967255409014888e-05,
+      "loss": 1.6336,
+      "step": 430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996691889168701e-05,
+      "loss": 1.8339,
+      "step": 431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9966580656852524e-05,
+      "loss": 1.0135,
+      "step": 432
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9966240704569722e-05,
+      "loss": 1.0952,
+      "step": 433
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996589903489717e-05,
+      "loss": 1.2929,
+      "step": 434
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996555564789376e-05,
+      "loss": 1.3729,
+      "step": 435
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9965210543618655e-05,
+      "loss": 1.6443,
+      "step": 436
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996486372213133e-05,
+      "loss": 1.2645,
+      "step": 437
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996451518349154e-05,
+      "loss": 1.3424,
+      "step": 438
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9964164927759354e-05,
+      "loss": 1.6375,
+      "step": 439
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9963812954995126e-05,
+      "loss": 1.4862,
+      "step": 440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9963459265259512e-05,
+      "loss": 1.2269,
+      "step": 441
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996310385861345e-05,
+      "loss": 1.4741,
+      "step": 442
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996274673511819e-05,
+      "loss": 1.2935,
+      "step": 443
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9962387894835277e-05,
+      "loss": 1.1065,
+      "step": 444
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9962027337826538e-05,
+      "loss": 1.6466,
+      "step": 445
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996166506415411e-05,
+      "loss": 0.8027,
+      "step": 446
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9961301073880413e-05,
+      "loss": 1.5649,
+      "step": 447
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996093536706818e-05,
+      "loss": 1.2821,
+      "step": 448
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9960567943780423e-05,
+      "loss": 1.0627,
+      "step": 449
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9960198804080462e-05,
+      "loss": 0.8905,
+      "step": 450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99598279480319e-05,
+      "loss": 0.9157,
+      "step": 451
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.995945537569865e-05,
+      "loss": 1.4639,
+      "step": 452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.995908108714491e-05,
+      "loss": 1.219,
+      "step": 453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9958705082435177e-05,
+      "loss": 1.4041,
+      "step": 454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9958327361634248e-05,
+      "loss": 1.3449,
+      "step": 455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9957947924807214e-05,
+      "loss": 1.1713,
+      "step": 456
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.995756677201945e-05,
+      "loss": 1.0898,
+      "step": 457
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9957183903336646e-05,
+      "loss": 1.4032,
+      "step": 458
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9956799318824776e-05,
+      "loss": 1.5517,
+      "step": 459
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9956413018550106e-05,
+      "loss": 1.3011,
+      "step": 460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9956025002579214e-05,
+      "loss": 1.4016,
+      "step": 461
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9955635270978953e-05,
+      "loss": 1.8784,
+      "step": 462
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.995524382381649e-05,
+      "loss": 1.6828,
+      "step": 463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9954850661159276e-05,
+      "loss": 1.7758,
+      "step": 464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.995445578307506e-05,
+      "loss": 0.873,
+      "step": 465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9954059189631883e-05,
+      "loss": 1.1641,
+      "step": 466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99536608808981e-05,
+      "loss": 1.1853,
+      "step": 467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9953260856942336e-05,
+      "loss": 1.0598,
+      "step": 468
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9952859117833526e-05,
+      "loss": 0.8125,
+      "step": 469
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99524556636409e-05,
+      "loss": 1.4485,
+      "step": 470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9952050494433983e-05,
+      "loss": 1.4083,
+      "step": 471
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9951643610282588e-05,
+      "loss": 1.3276,
+      "step": 472
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9951235011256836e-05,
+      "loss": 1.2714,
+      "step": 473
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9950824697427133e-05,
+      "loss": 1.0991,
+      "step": 474
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.995041266886419e-05,
+      "loss": 0.9617,
+      "step": 475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9949998925639e-05,
+      "loss": 1.0211,
+      "step": 476
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9949583467822863e-05,
+      "loss": 1.0131,
+      "step": 477
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994916629548737e-05,
+      "loss": 1.0687,
+      "step": 478
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9948747408704415e-05,
+      "loss": 1.6558,
+      "step": 479
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994832680754617e-05,
+      "loss": 1.6618,
+      "step": 480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9947904492085122e-05,
+      "loss": 1.4461,
+      "step": 481
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9947480462394043e-05,
+      "loss": 1.7258,
+      "step": 482
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9947054718545996e-05,
+      "loss": 0.985,
+      "step": 483
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9946627260614355e-05,
+      "loss": 1.2363,
+      "step": 484
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9946198088672776e-05,
+      "loss": 1.3736,
+      "step": 485
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994576720279521e-05,
+      "loss": 2.0391,
+      "step": 486
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994533460305591e-05,
+      "loss": 1.3745,
+      "step": 487
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9944900289529427e-05,
+      "loss": 0.7183,
+      "step": 488
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9944464262290596e-05,
+      "loss": 0.6403,
+      "step": 489
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9944026521414558e-05,
+      "loss": 1.7751,
+      "step": 490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994358706697674e-05,
+      "loss": 1.3566,
+      "step": 491
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994314589905287e-05,
+      "loss": 0.8968,
+      "step": 492
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9942703017718977e-05,
+      "loss": 1.3227,
+      "step": 493
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9942258423051367e-05,
+      "loss": 1.58,
+      "step": 494
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9941812115126664e-05,
+      "loss": 1.2345,
+      "step": 495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994136409402177e-05,
+      "loss": 0.985,
+      "step": 496
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994091435981389e-05,
+      "loss": 1.8588,
+      "step": 497
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9940462912580523e-05,
+      "loss": 0.8572,
+      "step": 498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9940009752399462e-05,
+      "loss": 1.3648,
+      "step": 499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9939554879348795e-05,
+      "loss": 1.8837,
+      "step": 500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9939098293506906e-05,
+      "loss": 1.375,
+      "step": 501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993863999495248e-05,
+      "loss": 1.5786,
+      "step": 502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993817998376448e-05,
+      "loss": 1.0434,
+      "step": 503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9937718260022184e-05,
+      "loss": 1.5309,
+      "step": 504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9937254823805156e-05,
+      "loss": 1.2764,
+      "step": 505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9936789675193254e-05,
+      "loss": 1.4037,
+      "step": 506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9936322814266634e-05,
+      "loss": 1.3393,
+      "step": 507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9935854241105744e-05,
+      "loss": 0.9479,
+      "step": 508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9935383955791326e-05,
+      "loss": 0.7509,
+      "step": 509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993491195840443e-05,
+      "loss": 1.1176,
+      "step": 510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993443824902638e-05,
+      "loss": 1.5917,
+      "step": 511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993396282773881e-05,
+      "loss": 1.2238,
+      "step": 512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993348569462365e-05,
+      "loss": 1.4436,
+      "step": 513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993300684976311e-05,
+      "loss": 0.969,
+      "step": 514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9932526293239713e-05,
+      "loss": 1.8801,
+      "step": 515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9932044025136267e-05,
+      "loss": 1.08,
+      "step": 516
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9931560045535873e-05,
+      "loss": 1.264,
+      "step": 517
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9931074354521935e-05,
+      "loss": 1.0251,
+      "step": 518
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9930586952178146e-05,
+      "loss": 1.1116,
+      "step": 519
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9930097838588496e-05,
+      "loss": 1.0868,
+      "step": 520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9929607013837268e-05,
+      "loss": 0.9113,
+      "step": 521
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9929114478009044e-05,
+      "loss": 1.2994,
+      "step": 522
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9928620231188694e-05,
+      "loss": 1.1154,
+      "step": 523
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9928124273461387e-05,
+      "loss": 1.0764,
+      "step": 524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9927626604912594e-05,
+      "loss": 1.223,
+      "step": 525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.992712722562807e-05,
+      "loss": 1.3821,
+      "step": 526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9926626135693866e-05,
+      "loss": 1.1291,
+      "step": 527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9926123335196326e-05,
+      "loss": 1.0556,
+      "step": 528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9925618824222103e-05,
+      "loss": 1.1173,
+      "step": 529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9925112602858132e-05,
+      "loss": 1.276,
+      "step": 530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.992460467119164e-05,
+      "loss": 1.4803,
+      "step": 531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.992409502931016e-05,
+      "loss": 1.3561,
+      "step": 532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9923583677301507e-05,
+      "loss": 1.6169,
+      "step": 533
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9923070615253806e-05,
+      "loss": 1.1208,
+      "step": 534
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9922555843255463e-05,
+      "loss": 1.3236,
+      "step": 535
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9922039361395186e-05,
+      "loss": 1.0343,
+      "step": 536
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9921521169761974e-05,
+      "loss": 1.2838,
+      "step": 537
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.992100126844512e-05,
+      "loss": 0.7448,
+      "step": 538
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.992047965753422e-05,
+      "loss": 1.4302,
+      "step": 539
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9919956337119156e-05,
+      "loss": 1.3162,
+      "step": 540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9919431307290105e-05,
+      "loss": 1.6937,
+      "step": 541
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9918904568137538e-05,
+      "loss": 1.2365,
+      "step": 542
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.991837611975223e-05,
+      "loss": 0.9456,
+      "step": 543
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.991784596222524e-05,
+      "loss": 1.6073,
+      "step": 544
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.991731409564792e-05,
+      "loss": 0.863,
+      "step": 545
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9916780520111932e-05,
+      "loss": 1.1324,
+      "step": 546
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.991624523570922e-05,
+      "loss": 0.9465,
+      "step": 547
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9915708242532013e-05,
+      "loss": 1.8668,
+      "step": 548
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9915169540672857e-05,
+      "loss": 1.504,
+      "step": 549
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.991462913022458e-05,
+      "loss": 1.8263,
+      "step": 550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9914087011280308e-05,
+      "loss": 1.5122,
+      "step": 551
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9913543183933455e-05,
+      "loss": 1.1238,
+      "step": 552
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9912997648277736e-05,
+      "loss": 1.5432,
+      "step": 553
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9912450404407152e-05,
+      "loss": 1.3462,
+      "step": 554
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9911901452416012e-05,
+      "loss": 0.6962,
+      "step": 555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9911350792398913e-05,
+      "loss": 0.7934,
+      "step": 556
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9910798424450734e-05,
+      "loss": 2.0264,
+      "step": 557
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.991024434866667e-05,
+      "loss": 0.8436,
+      "step": 558
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99096885651422e-05,
+      "loss": 1.4275,
+      "step": 559
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9909131073973092e-05,
+      "loss": 1.5881,
+      "step": 560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9908571875255416e-05,
+      "loss": 1.3194,
+      "step": 561
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.990801096908553e-05,
+      "loss": 1.4428,
+      "step": 562
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9907448355560094e-05,
+      "loss": 0.9304,
+      "step": 563
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9906884034776053e-05,
+      "loss": 1.4775,
+      "step": 564
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.990631800683066e-05,
+      "loss": 1.1986,
+      "step": 565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.990575027182144e-05,
+      "loss": 1.0014,
+      "step": 566
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.990518082984624e-05,
+      "loss": 1.0092,
+      "step": 567
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9904609681003174e-05,
+      "loss": 0.9542,
+      "step": 568
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9904036825390672e-05,
+      "loss": 1.1505,
+      "step": 569
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9903462263107447e-05,
+      "loss": 1.6982,
+      "step": 570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9902885994252506e-05,
+      "loss": 1.1678,
+      "step": 571
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.990230801892515e-05,
+      "loss": 1.855,
+      "step": 572
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.990172833722498e-05,
+      "loss": 0.8169,
+      "step": 573
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9901146949251888e-05,
+      "loss": 0.8464,
+      "step": 574
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9900563855106055e-05,
+      "loss": 0.8798,
+      "step": 575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9899979054887964e-05,
+      "loss": 1.2129,
+      "step": 576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9899392548698385e-05,
+      "loss": 0.9906,
+      "step": 577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9898804336638385e-05,
+      "loss": 1.4536,
+      "step": 578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 2.1857,
+      "step": 579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.989762279531287e-05,
+      "loss": 1.0796,
+      "step": 580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9897029466250955e-05,
+      "loss": 1.356,
+      "step": 581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.989643443172583e-05,
+      "loss": 1.2721,
+      "step": 582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.989583769184003e-05,
+      "loss": 1.4162,
+      "step": 583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9895239246696386e-05,
+      "loss": 1.469,
+      "step": 584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9894639096398022e-05,
+      "loss": 1.2074,
+      "step": 585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9894037241048356e-05,
+      "loss": 1.4794,
+      "step": 586
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9893433680751105e-05,
+      "loss": 0.9805,
+      "step": 587
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9892828415610265e-05,
+      "loss": 1.0221,
+      "step": 588
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.989222144573015e-05,
+      "loss": 1.6823,
+      "step": 589
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9891612771215337e-05,
+      "loss": 1.4907,
+      "step": 590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9891002392170727e-05,
+      "loss": 0.8602,
+      "step": 591
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9890390308701494e-05,
+      "loss": 0.9021,
+      "step": 592
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.988977652091311e-05,
+      "loss": 1.1349,
+      "step": 593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.988916102891135e-05,
+      "loss": 1.3298,
+      "step": 594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9888543832802277e-05,
+      "loss": 1.2523,
+      "step": 595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.988792493269224e-05,
+      "loss": 1.4586,
+      "step": 596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9887304328687892e-05,
+      "loss": 1.0384,
+      "step": 597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9886682020896172e-05,
+      "loss": 1.5003,
+      "step": 598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9886058009424323e-05,
+      "loss": 0.8044,
+      "step": 599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9885432294379875e-05,
+      "loss": 1.2986,
+      "step": 600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9884804875870645e-05,
+      "loss": 1.4983,
+      "step": 601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9884175754004752e-05,
+      "loss": 1.3151,
+      "step": 602
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9883544928890612e-05,
+      "loss": 1.6239,
+      "step": 603
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9882912400636923e-05,
+      "loss": 1.1561,
+      "step": 604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.988227816935269e-05,
+      "loss": 1.1495,
+      "step": 605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9881642235147196e-05,
+      "loss": 1.796,
+      "step": 606
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.988100459813003e-05,
+      "loss": 0.8251,
+      "step": 607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9880365258411068e-05,
+      "loss": 1.6336,
+      "step": 608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9879724216100488e-05,
+      "loss": 0.9891,
+      "step": 609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9879081471308745e-05,
+      "loss": 1.064,
+      "step": 610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9878437024146603e-05,
+      "loss": 1.1524,
+      "step": 611
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9877790874725115e-05,
+      "loss": 1.2384,
+      "step": 612
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.987714302315562e-05,
+      "loss": 1.8204,
+      "step": 613
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9876493469549763e-05,
+      "loss": 1.2056,
+      "step": 614
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.987584221401947e-05,
+      "loss": 0.9622,
+      "step": 615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.987518925667697e-05,
+      "loss": 1.1638,
+      "step": 616
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9874534597634776e-05,
+      "loss": 1.1512,
+      "step": 617
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9873878237005707e-05,
+      "loss": 1.036,
+      "step": 618
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9873220174902857e-05,
+      "loss": 1.2118,
+      "step": 619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9872560411439634e-05,
+      "loss": 1.7049,
+      "step": 620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9871898946729728e-05,
+      "loss": 1.3069,
+      "step": 621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9871235780887114e-05,
+      "loss": 1.0782,
+      "step": 622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9870570914026074e-05,
+      "loss": 1.3206,
+      "step": 623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9869904346261186e-05,
+      "loss": 1.2649,
+      "step": 624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9869236077707302e-05,
+      "loss": 1.4265,
+      "step": 625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9868566108479586e-05,
+      "loss": 1.2742,
+      "step": 626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.986789443869348e-05,
+      "loss": 1.3799,
+      "step": 627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9867221068464734e-05,
+      "loss": 1.2979,
+      "step": 628
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9866545997909387e-05,
+      "loss": 1.315,
+      "step": 629
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.986586922714375e-05,
+      "loss": 1.4441,
+      "step": 630
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9865190756284467e-05,
+      "loss": 1.2974,
+      "step": 631
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.986451058544844e-05,
+      "loss": 1.4692,
+      "step": 632
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9863828714752877e-05,
+      "loss": 1.4709,
+      "step": 633
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9863145144315277e-05,
+      "loss": 0.87,
+      "step": 634
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9862459874253438e-05,
+      "loss": 2.052,
+      "step": 635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9861772904685445e-05,
+      "loss": 1.3218,
+      "step": 636
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.986108423572968e-05,
+      "loss": 1.0394,
+      "step": 637
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9860393867504807e-05,
+      "loss": 1.2499,
+      "step": 638
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9859701800129797e-05,
+      "loss": 1.2412,
+      "step": 639
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9859008033723905e-05,
+      "loss": 1.1362,
+      "step": 640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9858312568406686e-05,
+      "loss": 1.3069,
+      "step": 641
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9857615404297972e-05,
+      "loss": 1.2163,
+      "step": 642
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.985691654151791e-05,
+      "loss": 1.2776,
+      "step": 643
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.985621598018693e-05,
+      "loss": 1.3718,
+      "step": 644
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9855513720425742e-05,
+      "loss": 1.4831,
+      "step": 645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9854809762355365e-05,
+      "loss": 1.7005,
+      "step": 646
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.985410410609711e-05,
+      "loss": 0.6984,
+      "step": 647
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9853396751772567e-05,
+      "loss": 0.9428,
+      "step": 648
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9852687699503638e-05,
+      "loss": 0.9488,
+      "step": 649
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9851976949412498e-05,
+      "loss": 0.8388,
+      "step": 650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9851264501621635e-05,
+      "loss": 1.3112,
+      "step": 651
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9850550356253807e-05,
+      "loss": 1.2058,
+      "step": 652
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9849834513432084e-05,
+      "loss": 1.7492,
+      "step": 653
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9849116973279816e-05,
+      "loss": 1.4652,
+      "step": 654
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9848397735920656e-05,
+      "loss": 1.0287,
+      "step": 655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9847676801478534e-05,
+      "loss": 1.4465,
+      "step": 656
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.984695417007769e-05,
+      "loss": 1.0131,
+      "step": 657
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9846229841842644e-05,
+      "loss": 1.2865,
+      "step": 658
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.984550381689822e-05,
+      "loss": 1.3034,
+      "step": 659
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9844776095369514e-05,
+      "loss": 1.2804,
+      "step": 660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9844046677381936e-05,
+      "loss": 1.0181,
+      "step": 661
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9843315563061182e-05,
+      "loss": 1.5036,
+      "step": 662
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.984258275253323e-05,
+      "loss": 1.2672,
+      "step": 663
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.984184824592437e-05,
+      "loss": 1.7136,
+      "step": 664
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.984111204336116e-05,
+      "loss": 1.1917,
+      "step": 665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9840374144970474e-05,
+      "loss": 1.2652,
+      "step": 666
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.983963455087946e-05,
+      "loss": 0.7652,
+      "step": 667
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.983889326121557e-05,
+      "loss": 2.0113,
+      "step": 668
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.983815027610654e-05,
+      "loss": 1.4948,
+      "step": 669
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.98374055956804e-05,
+      "loss": 0.824,
+      "step": 670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.983665922006548e-05,
+      "loss": 1.4404,
+      "step": 671
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.983591114939039e-05,
+      "loss": 1.5903,
+      "step": 672
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.983516138378404e-05,
+      "loss": 1.7157,
+      "step": 673
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9834409923375636e-05,
+      "loss": 1.3506,
+      "step": 674
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.983365676829466e-05,
+      "loss": 1.5133,
+      "step": 675
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9832901918670905e-05,
+      "loss": 0.9215,
+      "step": 676
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9832145374634445e-05,
+      "loss": 1.7756,
+      "step": 677
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9831387136315644e-05,
+      "loss": 1.7432,
+      "step": 678
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9830627203845164e-05,
+      "loss": 0.9513,
+      "step": 679
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9829865577353963e-05,
+      "loss": 1.2244,
+      "step": 680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9829102256973274e-05,
+      "loss": 1.3483,
+      "step": 681
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9828337242834645e-05,
+      "loss": 1.2747,
+      "step": 682
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.982757053506989e-05,
+      "loss": 1.6066,
+      "step": 683
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.982680213381114e-05,
+      "loss": 1.7302,
+      "step": 684
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9826032039190805e-05,
+      "loss": 1.1372,
+      "step": 685
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9825260251341586e-05,
+      "loss": 1.328,
+      "step": 686
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9824486770396477e-05,
+      "loss": 0.936,
+      "step": 687
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9823711596488765e-05,
+      "loss": 0.9657,
+      "step": 688
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9822934729752036e-05,
+      "loss": 1.6658,
+      "step": 689
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9822156170320146e-05,
+      "loss": 1.4504,
+      "step": 690
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9821375918327268e-05,
+      "loss": 1.7168,
+      "step": 691
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.982059397390785e-05,
+      "loss": 1.0605,
+      "step": 692
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9819810337196644e-05,
+      "loss": 1.2567,
+      "step": 693
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9819025008328675e-05,
+      "loss": 1.2252,
+      "step": 694
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9818237987439287e-05,
+      "loss": 1.8643,
+      "step": 695
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9817449274664086e-05,
+      "loss": 1.3644,
+      "step": 696
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.981665887013899e-05,
+      "loss": 0.7114,
+      "step": 697
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9815866774000203e-05,
+      "loss": 1.3484,
+      "step": 698
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.981507298638422e-05,
+      "loss": 1.1183,
+      "step": 699
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.981427750742782e-05,
+      "loss": 1.4091,
+      "step": 700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.981348033726809e-05,
+      "loss": 1.3266,
+      "step": 701
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.981268147604239e-05,
+      "loss": 1.435,
+      "step": 702
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9811880923888388e-05,
+      "loss": 1.2752,
+      "step": 703
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9811078680944034e-05,
+      "loss": 0.9502,
+      "step": 704
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9810274747347565e-05,
+      "loss": 1.4204,
+      "step": 705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9809469123237526e-05,
+      "loss": 1.3061,
+      "step": 706
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9808661808752735e-05,
+      "loss": 1.4052,
+      "step": 707
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9807852804032306e-05,
+      "loss": 1.5284,
+      "step": 708
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9807042109215656e-05,
+      "loss": 1.696,
+      "step": 709
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9806229724442478e-05,
+      "loss": 0.9284,
+      "step": 710
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9805415649852767e-05,
+      "loss": 1.5188,
+      "step": 711
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9804599885586805e-05,
+      "loss": 0.8658,
+      "step": 712
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.980378243178516e-05,
+      "loss": 1.4977,
+      "step": 713
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.98029632885887e-05,
+      "loss": 1.3693,
+      "step": 714
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.980214245613858e-05,
+      "loss": 1.2128,
+      "step": 715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.980131993457625e-05,
+      "loss": 1.3567,
+      "step": 716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.980049572404344e-05,
+      "loss": 1.1863,
+      "step": 717
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9799669824682182e-05,
+      "loss": 1.1051,
+      "step": 718
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9798842236634797e-05,
+      "loss": 1.5907,
+      "step": 719
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9798012960043892e-05,
+      "loss": 1.4873,
+      "step": 720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9797181995052374e-05,
+      "loss": 1.3488,
+      "step": 721
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9796349341803432e-05,
+      "loss": 1.0626,
+      "step": 722
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.979551500044055e-05,
+      "loss": 1.1456,
+      "step": 723
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.97946789711075e-05,
+      "loss": 1.1562,
+      "step": 724
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.979384125394835e-05,
+      "loss": 1.2386,
+      "step": 725
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9793001849107457e-05,
+      "loss": 1.0105,
+      "step": 726
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9792160756729468e-05,
+      "loss": 1.2727,
+      "step": 727
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9791317976959315e-05,
+      "loss": 1.2772,
+      "step": 728
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9790473509942233e-05,
+      "loss": 1.3136,
+      "step": 729
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9789627355823737e-05,
+      "loss": 1.2681,
+      "step": 730
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9788779514749635e-05,
+      "loss": 1.0773,
+      "step": 731
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9787929986866035e-05,
+      "loss": 1.4486,
+      "step": 732
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9787078772319328e-05,
+      "loss": 1.5167,
+      "step": 733
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9786225871256188e-05,
+      "loss": 1.1093,
+      "step": 734
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.978537128382359e-05,
+      "loss": 1.5701,
+      "step": 735
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9784515010168798e-05,
+      "loss": 1.0813,
+      "step": 736
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.978365705043937e-05,
+      "loss": 1.4243,
+      "step": 737
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9782797404783142e-05,
+      "loss": 1.3941,
+      "step": 738
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.978193607334826e-05,
+      "loss": 1.1715,
+      "step": 739
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9781073056283136e-05,
+      "loss": 0.8964,
+      "step": 740
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9780208353736493e-05,
+      "loss": 1.6408,
+      "step": 741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.977934196585734e-05,
+      "loss": 1.363,
+      "step": 742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.977847389279497e-05,
+      "loss": 0.9124,
+      "step": 743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9777604134698967e-05,
+      "loss": 0.8885,
+      "step": 744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9776732691719215e-05,
+      "loss": 1.0986,
+      "step": 745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9775859564005876e-05,
+      "loss": 1.4716,
+      "step": 746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.977498475170941e-05,
+      "loss": 1.5743,
+      "step": 747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9774108254980566e-05,
+      "loss": 1.8857,
+      "step": 748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.977323007397038e-05,
+      "loss": 1.2778,
+      "step": 749
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9772350208830183e-05,
+      "loss": 1.7133,
+      "step": 750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9771468659711595e-05,
+      "loss": 0.8516,
+      "step": 751
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9770585426766528e-05,
+      "loss": 1.3035,
+      "step": 752
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9769700510147173e-05,
+      "loss": 2.0382,
+      "step": 753
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.976881391000602e-05,
+      "loss": 1.0794,
+      "step": 754
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9767925626495857e-05,
+      "loss": 1.2882,
+      "step": 755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.976703565976975e-05,
+      "loss": 1.0495,
+      "step": 756
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.976614400998105e-05,
+      "loss": 0.9139,
+      "step": 757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9765250677283423e-05,
+      "loss": 1.7564,
+      "step": 758
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9764355661830796e-05,
+      "loss": 0.9047,
+      "step": 759
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9763458963777404e-05,
+      "loss": 1.2472,
+      "step": 760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9762560583277763e-05,
+      "loss": 1.5501,
+      "step": 761
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9761660520486687e-05,
+      "loss": 1.0141,
+      "step": 762
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9760758775559275e-05,
+      "loss": 0.9573,
+      "step": 763
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975985534865091e-05,
+      "loss": 0.9637,
+      "step": 764
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975895023991728e-05,
+      "loss": 1.019,
+      "step": 765
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9758043449514348e-05,
+      "loss": 1.4875,
+      "step": 766
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9757134977598374e-05,
+      "loss": 1.5112,
+      "step": 767
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975622482432591e-05,
+      "loss": 1.2182,
+      "step": 768
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975531298985379e-05,
+      "loss": 1.1471,
+      "step": 769
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9754399474339144e-05,
+      "loss": 1.0198,
+      "step": 770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975348427793939e-05,
+      "loss": 0.8533,
+      "step": 771
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975256740081223e-05,
+      "loss": 1.1433,
+      "step": 772
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975164884311567e-05,
+      "loss": 1.2707,
+      "step": 773
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975072860500799e-05,
+      "loss": 1.2291,
+      "step": 774
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.974980668664777e-05,
+      "loss": 1.6827,
+      "step": 775
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.974888308819387e-05,
+      "loss": 1.1572,
+      "step": 776
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.974795780980545e-05,
+      "loss": 1.0579,
+      "step": 777
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9747030851641953e-05,
+      "loss": 1.283,
+      "step": 778
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9746102213863113e-05,
+      "loss": 1.7813,
+      "step": 779
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9745171896628955e-05,
+      "loss": 1.3122,
+      "step": 780
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.974423990009979e-05,
+      "loss": 0.9952,
+      "step": 781
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9743306224436224e-05,
+      "loss": 0.6918,
+      "step": 782
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9742370869799147e-05,
+      "loss": 0.9505,
+      "step": 783
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9741433836349735e-05,
+      "loss": 1.0609,
+      "step": 784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9740495124249462e-05,
+      "loss": 1.8976,
+      "step": 785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.973955473366009e-05,
+      "loss": 1.1224,
+      "step": 786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.973861266474366e-05,
+      "loss": 1.1525,
+      "step": 787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9737668917662523e-05,
+      "loss": 1.2866,
+      "step": 788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9736723492579295e-05,
+      "loss": 1.2957,
+      "step": 789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9735776389656897e-05,
+      "loss": 2.0182,
+      "step": 790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9734827609058533e-05,
+      "loss": 1.1365,
+      "step": 791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9733877150947696e-05,
+      "loss": 1.6181,
+      "step": 792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9732925015488174e-05,
+      "loss": 1.022,
+      "step": 793
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9731971202844036e-05,
+      "loss": 1.7042,
+      "step": 794
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9731015713179643e-05,
+      "loss": 0.9524,
+      "step": 795
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9730058546659653e-05,
+      "loss": 1.2281,
+      "step": 796
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9729099703449e-05,
+      "loss": 0.9549,
+      "step": 797
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9728139183712907e-05,
+      "loss": 1.427,
+      "step": 798
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9727176987616897e-05,
+      "loss": 1.6195,
+      "step": 799
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.972621311532678e-05,
+      "loss": 1.6572,
+      "step": 800
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.972524756700865e-05,
+      "loss": 1.5861,
+      "step": 801
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9724280342828886e-05,
+      "loss": 1.6712,
+      "step": 802
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9723311442954163e-05,
+      "loss": 1.1843,
+      "step": 803
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9722340867551445e-05,
+      "loss": 1.472,
+      "step": 804
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9721368616787976e-05,
+      "loss": 1.426,
+      "step": 805
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9720394690831305e-05,
+      "loss": 1.153,
+      "step": 806
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.971941908984925e-05,
+      "loss": 0.9503,
+      "step": 807
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.971844181400993e-05,
+      "loss": 1.5463,
+      "step": 808
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9717462863481753e-05,
+      "loss": 1.4975,
+      "step": 809
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.971648223843341e-05,
+      "loss": 0.7411,
+      "step": 810
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9715499939033883e-05,
+      "loss": 1.6359,
+      "step": 811
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.971451596545244e-05,
+      "loss": 0.7033,
+      "step": 812
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9713530317858647e-05,
+      "loss": 1.3728,
+      "step": 813
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.971254299642235e-05,
+      "loss": 1.2521,
+      "step": 814
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9711554001313674e-05,
+      "loss": 0.837,
+      "step": 815
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9710563332703055e-05,
+      "loss": 1.8633,
+      "step": 816
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9709570990761207e-05,
+      "loss": 1.0392,
+      "step": 817
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9708576975659123e-05,
+      "loss": 1.0017,
+      "step": 818
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9707581287568094e-05,
+      "loss": 1.4359,
+      "step": 819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9706583926659705e-05,
+      "loss": 1.448,
+      "step": 820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.970558489310581e-05,
+      "loss": 1.0107,
+      "step": 821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9704584187078577e-05,
+      "loss": 1.4913,
+      "step": 822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9703581808750436e-05,
+      "loss": 1.1024,
+      "step": 823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9702577758294127e-05,
+      "loss": 1.3709,
+      "step": 824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9701572035882658e-05,
+      "loss": 1.2554,
+      "step": 825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9700564641689346e-05,
+      "loss": 1.4732,
+      "step": 826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.969955557588778e-05,
+      "loss": 1.2257,
+      "step": 827
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9698544838651846e-05,
+      "loss": 1.0955,
+      "step": 828
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9697532430155718e-05,
+      "loss": 1.1561,
+      "step": 829
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9696518350573845e-05,
+      "loss": 0.9592,
+      "step": 830
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9695502600080983e-05,
+      "loss": 0.9793,
+      "step": 831
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9694485178852164e-05,
+      "loss": 0.6372,
+      "step": 832
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.969346608706271e-05,
+      "loss": 1.2617,
+      "step": 833
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9692445324888233e-05,
+      "loss": 1.1457,
+      "step": 834
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9691422892504626e-05,
+      "loss": 0.7962,
+      "step": 835
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9690398790088083e-05,
+      "loss": 1.2675,
+      "step": 836
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9689373017815076e-05,
+      "loss": 0.9127,
+      "step": 837
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.968834557586236e-05,
+      "loss": 0.8452,
+      "step": 838
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9687316464406997e-05,
+      "loss": 1.2247,
+      "step": 839
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9686285683626316e-05,
+      "loss": 1.116,
+      "step": 840
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9685253233697943e-05,
+      "loss": 1.0401,
+      "step": 841
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.968421911479979e-05,
+      "loss": 0.7475,
+      "step": 842
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.968318332711006e-05,
+      "loss": 1.2017,
+      "step": 843
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9682145870807233e-05,
+      "loss": 1.1856,
+      "step": 844
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9681106746070096e-05,
+      "loss": 1.248,
+      "step": 845
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9680065953077704e-05,
+      "loss": 0.8945,
+      "step": 846
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.967902349200941e-05,
+      "loss": 0.8405,
+      "step": 847
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9677979363044846e-05,
+      "loss": 1.3114,
+      "step": 848
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9676933566363943e-05,
+      "loss": 0.7763,
+      "step": 849
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9675886102146912e-05,
+      "loss": 0.9711,
+      "step": 850
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9674836970574253e-05,
+      "loss": 1.6413,
+      "step": 851
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9673786171826755e-05,
+      "loss": 1.3824,
+      "step": 852
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9672733706085488e-05,
+      "loss": 0.9458,
+      "step": 853
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9671679573531815e-05,
+      "loss": 0.928,
+      "step": 854
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9670623774347387e-05,
+      "loss": 0.8847,
+      "step": 855
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.966956630871414e-05,
+      "loss": 1.5227,
+      "step": 856
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9668507176814295e-05,
+      "loss": 1.4648,
+      "step": 857
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9667446378830363e-05,
+      "loss": 0.9323,
+      "step": 858
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.966638391494514e-05,
+      "loss": 1.5896,
+      "step": 859
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9665319785341713e-05,
+      "loss": 1.6798,
+      "step": 860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9664253990203456e-05,
+      "loss": 1.1682,
+      "step": 861
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9663186529714022e-05,
+      "loss": 0.804,
+      "step": 862
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.966211740405736e-05,
+      "loss": 1.5944,
+      "step": 863
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9661046613417702e-05,
+      "loss": 1.1328,
+      "step": 864
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9659974157979566e-05,
+      "loss": 0.9464,
+      "step": 865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.965890003792776e-05,
+      "loss": 1.1277,
+      "step": 866
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9657824253447378e-05,
+      "loss": 0.9706,
+      "step": 867
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9656746804723794e-05,
+      "loss": 0.8101,
+      "step": 868
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9655667691942684e-05,
+      "loss": 1.1957,
+      "step": 869
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9654586915289994e-05,
+      "loss": 1.0528,
+      "step": 870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9653504474951965e-05,
+      "loss": 1.2829,
+      "step": 871
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.965242037111513e-05,
+      "loss": 0.6818,
+      "step": 872
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9651334603966298e-05,
+      "loss": 1.1286,
+      "step": 873
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9650247173692565e-05,
+      "loss": 0.7951,
+      "step": 874
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9649158080481327e-05,
+      "loss": 0.8535,
+      "step": 875
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.964806732452025e-05,
+      "loss": 0.7372,
+      "step": 876
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9646974905997295e-05,
+      "loss": 0.8706,
+      "step": 877
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.964588082510071e-05,
+      "loss": 0.8035,
+      "step": 878
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.964478508201903e-05,
+      "loss": 1.3082,
+      "step": 879
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.964368767694107e-05,
+      "loss": 0.7903,
+      "step": 880
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.964258861005594e-05,
+      "loss": 1.1664,
+      "step": 881
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9641487881553026e-05,
+      "loss": 0.7454,
+      "step": 882
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.964038549162201e-05,
+      "loss": 1.2365,
+      "step": 883
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9639281440452858e-05,
+      "loss": 1.2786,
+      "step": 884
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9638175728235817e-05,
+      "loss": 1.0413,
+      "step": 885
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9637068355161428e-05,
+      "loss": 0.7057,
+      "step": 886
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9635959321420514e-05,
+      "loss": 0.9939,
+      "step": 887
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9634848627204182e-05,
+      "loss": 0.8481,
+      "step": 888
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9633736272703826e-05,
+      "loss": 0.782,
+      "step": 889
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9632622258111135e-05,
+      "loss": 0.9992,
+      "step": 890
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.963150658361807e-05,
+      "loss": 0.9911,
+      "step": 891
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.963038924941689e-05,
+      "loss": 1.0151,
+      "step": 892
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9629270255700127e-05,
+      "loss": 1.3492,
+      "step": 893
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9628149602660615e-05,
+      "loss": 1.027,
+      "step": 894
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.962702729049146e-05,
+      "loss": 0.5747,
+      "step": 895
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9625903319386064e-05,
+      "loss": 1.0845,
+      "step": 896
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9624777689538107e-05,
+      "loss": 0.8127,
+      "step": 897
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9623650401141563e-05,
+      "loss": 0.7639,
+      "step": 898
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.962252145439068e-05,
+      "loss": 1.1167,
+      "step": 899
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.962139084948001e-05,
+      "loss": 0.8443,
+      "step": 900
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9620258586604364e-05,
+      "loss": 1.0628,
+      "step": 901
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9619124665958867e-05,
+      "loss": 1.3133,
+      "step": 902
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9617989087738913e-05,
+      "loss": 1.3344,
+      "step": 903
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9616851852140188e-05,
+      "loss": 0.8093,
+      "step": 904
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9615712959358656e-05,
+      "loss": 0.922,
+      "step": 905
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9614572409590575e-05,
+      "loss": 1.0419,
+      "step": 906
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9613430203032486e-05,
+      "loss": 1.0293,
+      "step": 907
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9612286339881213e-05,
+      "loss": 1.1019,
+      "step": 908
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9611140820333868e-05,
+      "loss": 0.689,
+      "step": 909
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.960999364458785e-05,
+      "loss": 0.9873,
+      "step": 910
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.960884481284084e-05,
+      "loss": 1.1895,
+      "step": 911
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.96076943252908e-05,
+      "loss": 1.2913,
+      "step": 912
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.960654218213599e-05,
+      "loss": 1.208,
+      "step": 913
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9605388383574948e-05,
+      "loss": 0.8498,
+      "step": 914
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9604232929806493e-05,
+      "loss": 0.8538,
+      "step": 915
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9603075821029734e-05,
+      "loss": 0.9295,
+      "step": 916
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.960191705744407e-05,
+      "loss": 1.0465,
+      "step": 917
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9600756639249175e-05,
+      "loss": 0.8313,
+      "step": 918
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9599594566645016e-05,
+      "loss": 0.8451,
+      "step": 919
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9598430839831836e-05,
+      "loss": 1.0658,
+      "step": 920
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9597265459010177e-05,
+      "loss": 0.9519,
+      "step": 921
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9596098424380855e-05,
+      "loss": 0.8551,
+      "step": 922
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.8863,
+      "step": 923
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9593759394503925e-05,
+      "loss": 1.0073,
+      "step": 924
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.959258739965938e-05,
+      "loss": 0.959,
+      "step": 925
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.95914137518133e-05,
+      "loss": 0.8418,
+      "step": 926
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9590238451167927e-05,
+      "loss": 1.2574,
+      "step": 927
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.958906149792579e-05,
+      "loss": 0.8694,
+      "step": 928
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.95878828922897e-05,
+      "loss": 0.9615,
+      "step": 929
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.958670263446276e-05,
+      "loss": 1.0167,
+      "step": 930
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9585520724648354e-05,
+      "loss": 1.3627,
+      "step": 931
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9584337163050145e-05,
+      "loss": 1.1647,
+      "step": 932
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9583151949872083e-05,
+      "loss": 0.8245,
+      "step": 933
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.958196508531841e-05,
+      "loss": 0.7713,
+      "step": 934
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9580776569593646e-05,
+      "loss": 0.997,
+      "step": 935
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.95795864029026e-05,
+      "loss": 1.2893,
+      "step": 936
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9578394585450354e-05,
+      "loss": 1.009,
+      "step": 937
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.957720111744229e-05,
+      "loss": 1.0886,
+      "step": 938
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.957600599908406e-05,
+      "loss": 1.0557,
+      "step": 939
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.957480923058162e-05,
+      "loss": 0.7284,
+      "step": 940
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9573610812141183e-05,
+      "loss": 1.3707,
+      "step": 941
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9572410743969275e-05,
+      "loss": 0.8368,
+      "step": 942
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9571209026272683e-05,
+      "loss": 1.0802,
+      "step": 943
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9570005659258493e-05,
+      "loss": 1.2621,
+      "step": 944
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9568800643134073e-05,
+      "loss": 0.9765,
+      "step": 945
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.956759397810706e-05,
+      "loss": 0.9658,
+      "step": 946
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.95663856643854e-05,
+      "loss": 0.7247,
+      "step": 947
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9565175702177305e-05,
+      "loss": 0.9484,
+      "step": 948
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9563964091691275e-05,
+      "loss": 0.738,
+      "step": 949
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9562750833136096e-05,
+      "loss": 0.7181,
+      "step": 950
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9561535926720846e-05,
+      "loss": 0.7743,
+      "step": 951
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9560319372654864e-05,
+      "loss": 0.5682,
+      "step": 952
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.95591011711478e-05,
+      "loss": 1.2538,
+      "step": 953
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9557881322409572e-05,
+      "loss": 1.2674,
+      "step": 954
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.955665982665038e-05,
+      "loss": 1.2156,
+      "step": 955
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.955543668408072e-05,
+      "loss": 1.0532,
+      "step": 956
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9554211894911363e-05,
+      "loss": 0.6557,
+      "step": 957
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.955298545935336e-05,
+      "loss": 0.7071,
+      "step": 958
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9551757377618056e-05,
+      "loss": 1.2263,
+      "step": 959
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9550527649917077e-05,
+      "loss": 1.126,
+      "step": 960
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9549296276462326e-05,
+      "loss": 1.2136,
+      "step": 961
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9548063257465997e-05,
+      "loss": 1.0141,
+      "step": 962
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9546828593140565e-05,
+      "loss": 0.7208,
+      "step": 963
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9545592283698784e-05,
+      "loss": 0.9963,
+      "step": 964
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.95443543293537e-05,
+      "loss": 1.1406,
+      "step": 965
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.954311473031864e-05,
+      "loss": 0.8551,
+      "step": 966
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.954187348680721e-05,
+      "loss": 1.4571,
+      "step": 967
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9540630599033297e-05,
+      "loss": 0.6772,
+      "step": 968
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.953938606721108e-05,
+      "loss": 0.7642,
+      "step": 969
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9538139891555024e-05,
+      "loss": 1.0363,
+      "step": 970
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9536892072279863e-05,
+      "loss": 0.8562,
+      "step": 971
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9535642609600625e-05,
+      "loss": 1.5801,
+      "step": 972
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9534391503732617e-05,
+      "loss": 1.0238,
+      "step": 973
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.953313875489143e-05,
+      "loss": 0.9156,
+      "step": 974
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9531884363292944e-05,
+      "loss": 1.0633,
+      "step": 975
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.953062832915331e-05,
+      "loss": 1.0503,
+      "step": 976
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.952937065268897e-05,
+      "loss": 0.5454,
+      "step": 977
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.952811133411665e-05,
+      "loss": 0.6742,
+      "step": 978
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9526850373653356e-05,
+      "loss": 0.8818,
+      "step": 979
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9525587771516378e-05,
+      "loss": 0.7173,
+      "step": 980
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9524323527923284e-05,
+      "loss": 1.1061,
+      "step": 981
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.952305764309194e-05,
+      "loss": 0.716,
+      "step": 982
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9521790117240472e-05,
+      "loss": 0.8946,
+      "step": 983
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9520520950587304e-05,
+      "loss": 1.2197,
+      "step": 984
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9519250143351146e-05,
+      "loss": 0.9845,
+      "step": 985
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.951797769575098e-05,
+      "loss": 0.948,
+      "step": 986
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9516703608006074e-05,
+      "loss": 1.1609,
+      "step": 987
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9515427880335986e-05,
+      "loss": 1.0806,
+      "step": 988
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.951415051296054e-05,
+      "loss": 0.9029,
+      "step": 989
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.951287150609986e-05,
+      "loss": 0.8838,
+      "step": 990
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9511590859974344e-05,
+      "loss": 1.1779,
+      "step": 991
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9510308574804673e-05,
+      "loss": 1.0206,
+      "step": 992
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9509024650811813e-05,
+      "loss": 1.2156,
+      "step": 993
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9507739088217007e-05,
+      "loss": 1.1812,
+      "step": 994
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9506451887241787e-05,
+      "loss": 0.7558,
+      "step": 995
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9505163048107968e-05,
+      "loss": 1.2155,
+      "step": 996
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9503872571037637e-05,
+      "loss": 0.9786,
+      "step": 997
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9502580456253173e-05,
+      "loss": 1.3106,
+      "step": 998
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9501286703977232e-05,
+      "loss": 0.7946,
+      "step": 999
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9499991314432757e-05,
+      "loss": 1.4868,
+      "step": 1000
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9498694287842972e-05,
+      "loss": 0.823,
+      "step": 1001
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9497395624431378e-05,
+      "loss": 0.6461,
+      "step": 1002
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.949609532442176e-05,
+      "loss": 0.8271,
+      "step": 1003
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.949479338803819e-05,
+      "loss": 1.1006,
+      "step": 1004
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.949348981550502e-05,
+      "loss": 1.0635,
+      "step": 1005
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.949218460704688e-05,
+      "loss": 1.2877,
+      "step": 1006
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9490877762888685e-05,
+      "loss": 1.0851,
+      "step": 1007
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9489569283255633e-05,
+      "loss": 1.0516,
+      "step": 1008
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9488259168373198e-05,
+      "loss": 0.8952,
+      "step": 1009
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.948694741846714e-05,
+      "loss": 0.8495,
+      "step": 1010
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9485634033763507e-05,
+      "loss": 0.901,
+      "step": 1011
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9484319014488618e-05,
+      "loss": 1.1878,
+      "step": 1012
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9483002360869082e-05,
+      "loss": 0.9584,
+      "step": 1013
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9481684073131774e-05,
+      "loss": 0.6097,
+      "step": 1014
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9480364151503876e-05,
+      "loss": 0.8318,
+      "step": 1015
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9479042596212833e-05,
+      "loss": 0.8177,
+      "step": 1016
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9477719407486372e-05,
+      "loss": 1.0826,
+      "step": 1017
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.947639458555251e-05,
+      "loss": 0.9918,
+      "step": 1018
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9475068130639543e-05,
+      "loss": 0.939,
+      "step": 1019
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9473740042976045e-05,
+      "loss": 0.9573,
+      "step": 1020
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.947241032279087e-05,
+      "loss": 0.8464,
+      "step": 1021
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.947107897031316e-05,
+      "loss": 1.3508,
+      "step": 1022
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9469745985772333e-05,
+      "loss": 0.9948,
+      "step": 1023
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.946841136939809e-05,
+      "loss": 0.5097,
+      "step": 1024
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9467075121420415e-05,
+      "loss": 0.9076,
+      "step": 1025
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9465737242069565e-05,
+      "loss": 1.0428,
+      "step": 1026
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9464397731576093e-05,
+      "loss": 0.7069,
+      "step": 1027
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9463056590170822e-05,
+      "loss": 1.0542,
+      "step": 1028
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9461713818084857e-05,
+      "loss": 1.2891,
+      "step": 1029
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.946036941554958e-05,
+      "loss": 1.0626,
+      "step": 1030
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9459023382796667e-05,
+      "loss": 0.6815,
+      "step": 1031
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9457675720058063e-05,
+      "loss": 1.2258,
+      "step": 1032
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9456326427566002e-05,
+      "loss": 1.0482,
+      "step": 1033
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9454975505552994e-05,
+      "loss": 0.98,
+      "step": 1034
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.945362295425183e-05,
+      "loss": 0.6251,
+      "step": 1035
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9452268773895582e-05,
+      "loss": 0.9217,
+      "step": 1036
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9450912964717604e-05,
+      "loss": 0.9712,
+      "step": 1037
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.944955552695153e-05,
+      "loss": 1.2332,
+      "step": 1038
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9448196460831274e-05,
+      "loss": 0.8629,
+      "step": 1039
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9446835766591032e-05,
+      "loss": 0.4908,
+      "step": 1040
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.944547344446528e-05,
+      "loss": 0.8569,
+      "step": 1041
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9444109494688773e-05,
+      "loss": 1.4193,
+      "step": 1042
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.944274391749655e-05,
+      "loss": 0.7405,
+      "step": 1043
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9441376713123924e-05,
+      "loss": 0.9289,
+      "step": 1044
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9440007881806502e-05,
+      "loss": 0.995,
+      "step": 1045
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.943863742378015e-05,
+      "loss": 1.0817,
+      "step": 1046
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9437265339281034e-05,
+      "loss": 1.0899,
+      "step": 1047
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9435891628545593e-05,
+      "loss": 0.8082,
+      "step": 1048
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.943451629181054e-05,
+      "loss": 0.736,
+      "step": 1049
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9433139329312876e-05,
+      "loss": 1.1373,
+      "step": 1050
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9431760741289886e-05,
+      "loss": 0.7918,
+      "step": 1051
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9430380527979124e-05,
+      "loss": 0.6676,
+      "step": 1052
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9428998689618428e-05,
+      "loss": 1.1873,
+      "step": 1053
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9427615226445918e-05,
+      "loss": 0.9693,
+      "step": 1054
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9426230138699998e-05,
+      "loss": 0.732,
+      "step": 1055
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9424843426619343e-05,
+      "loss": 0.9839,
+      "step": 1056
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9423455090442917e-05,
+      "loss": 1.1055,
+      "step": 1057
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9422065130409954e-05,
+      "loss": 1.0089,
+      "step": 1058
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.942067354675997e-05,
+      "loss": 1.2319,
+      "step": 1059
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.941928033973277e-05,
+      "loss": 1.0473,
+      "step": 1060
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9417885509568433e-05,
+      "loss": 1.1752,
+      "step": 1061
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9416489056507312e-05,
+      "loss": 0.8267,
+      "step": 1062
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.941509098079005e-05,
+      "loss": 0.8377,
+      "step": 1063
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9413691282657563e-05,
+      "loss": 0.6613,
+      "step": 1064
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9412289962351042e-05,
+      "loss": 1.1661,
+      "step": 1065
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9410887020111973e-05,
+      "loss": 0.9211,
+      "step": 1066
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9409482456182105e-05,
+      "loss": 0.8762,
+      "step": 1067
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9408076270803476e-05,
+      "loss": 0.9251,
+      "step": 1068
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.94066684642184e-05,
+      "loss": 0.7659,
+      "step": 1069
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9405259036669476e-05,
+      "loss": 0.8459,
+      "step": 1070
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.940384798839957e-05,
+      "loss": 0.8723,
+      "step": 1071
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9402435319651838e-05,
+      "loss": 1.394,
+      "step": 1072
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9401021030669715e-05,
+      "loss": 0.9702,
+      "step": 1073
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.939960512169691e-05,
+      "loss": 0.9151,
+      "step": 1074
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.939818759297741e-05,
+      "loss": 1.0298,
+      "step": 1075
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.939676844475549e-05,
+      "loss": 1.0313,
+      "step": 1076
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9395347677275695e-05,
+      "loss": 0.8852,
+      "step": 1077
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9393925290782856e-05,
+      "loss": 0.878,
+      "step": 1078
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9392501285522078e-05,
+      "loss": 0.8891,
+      "step": 1079
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9391075661738747e-05,
+      "loss": 1.2454,
+      "step": 1080
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9389648419678523e-05,
+      "loss": 1.1052,
+      "step": 1081
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9388219559587353e-05,
+      "loss": 1.2967,
+      "step": 1082
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9386789081711465e-05,
+      "loss": 0.9364,
+      "step": 1083
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.938535698629735e-05,
+      "loss": 1.0871,
+      "step": 1084
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9383923273591793e-05,
+      "loss": 0.6089,
+      "step": 1085
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9382487943841855e-05,
+      "loss": 0.8669,
+      "step": 1086
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9381050997294864e-05,
+      "loss": 0.9008,
+      "step": 1087
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9379612434198446e-05,
+      "loss": 1.3479,
+      "step": 1088
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9378172254800487e-05,
+      "loss": 0.9746,
+      "step": 1089
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9376730459349165e-05,
+      "loss": 0.8877,
+      "step": 1090
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9375287048092927e-05,
+      "loss": 1.01,
+      "step": 1091
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.937384202128051e-05,
+      "loss": 0.5423,
+      "step": 1092
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9372395379160913e-05,
+      "loss": 1.1549,
+      "step": 1093
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9370947121983427e-05,
+      "loss": 1.0928,
+      "step": 1094
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.936949724999762e-05,
+      "loss": 1.1516,
+      "step": 1095
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9368045763453328e-05,
+      "loss": 0.7792,
+      "step": 1096
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9366592662600678e-05,
+      "loss": 0.77,
+      "step": 1097
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9365137947690063e-05,
+      "loss": 1.1121,
+      "step": 1098
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9363681618972166e-05,
+      "loss": 1.1635,
+      "step": 1099
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.936222367669794e-05,
+      "loss": 1.1591,
+      "step": 1100
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.936076412111862e-05,
+      "loss": 0.6275,
+      "step": 1101
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.935930295248572e-05,
+      "loss": 0.9429,
+      "step": 1102
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9357840171051025e-05,
+      "loss": 0.9767,
+      "step": 1103
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9356375777066604e-05,
+      "loss": 0.8115,
+      "step": 1104
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9354909770784808e-05,
+      "loss": 1.1492,
+      "step": 1105
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9353442152458244e-05,
+      "loss": 1.4193,
+      "step": 1106
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9351972922339835e-05,
+      "loss": 1.1824,
+      "step": 1107
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9350502080682744e-05,
+      "loss": 0.6456,
+      "step": 1108
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9349029627740434e-05,
+      "loss": 1.1632,
+      "step": 1109
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9347555563766635e-05,
+      "loss": 0.9042,
+      "step": 1110
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9346079889015366e-05,
+      "loss": 1.3507,
+      "step": 1111
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.934460260374091e-05,
+      "loss": 1.0867,
+      "step": 1112
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9343123708197834e-05,
+      "loss": 0.8163,
+      "step": 1113
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9341643202640987e-05,
+      "loss": 0.9019,
+      "step": 1114
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9340161087325483e-05,
+      "loss": 0.9791,
+      "step": 1115
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.933867736250673e-05,
+      "loss": 0.6435,
+      "step": 1116
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9337192028440395e-05,
+      "loss": 0.7911,
+      "step": 1117
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.933570508538244e-05,
+      "loss": 0.957,
+      "step": 1118
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9334216533589095e-05,
+      "loss": 1.2601,
+      "step": 1119
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9332726373316866e-05,
+      "loss": 1.1132,
+      "step": 1120
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.933123460482254e-05,
+      "loss": 0.9882,
+      "step": 1121
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9329741228363174e-05,
+      "loss": 1.3386,
+      "step": 1122
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9328246244196117e-05,
+      "loss": 0.5956,
+      "step": 1123
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.932674965257898e-05,
+      "loss": 0.7152,
+      "step": 1124
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9325251453769657e-05,
+      "loss": 1.15,
+      "step": 1125
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.932375164802632e-05,
+      "loss": 0.8366,
+      "step": 1126
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9322250235607418e-05,
+      "loss": 0.9542,
+      "step": 1127
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.932074721677167e-05,
+      "loss": 1.0656,
+      "step": 1128
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9319242591778086e-05,
+      "loss": 0.7784,
+      "step": 1129
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9317736360885937e-05,
+      "loss": 0.8202,
+      "step": 1130
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.931622852435478e-05,
+      "loss": 0.5428,
+      "step": 1131
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9314719082444443e-05,
+      "loss": 1.0818,
+      "step": 1132
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9313208035415042e-05,
+      "loss": 0.8158,
+      "step": 1133
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.931169538352695e-05,
+      "loss": 1.1454,
+      "step": 1134
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9310181127040842e-05,
+      "loss": 0.8106,
+      "step": 1135
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.930866526621765e-05,
+      "loss": 0.8883,
+      "step": 1136
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9307147801318585e-05,
+      "loss": 1.5005,
+      "step": 1137
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.930562873260514e-05,
+      "loss": 0.9591,
+      "step": 1138
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.930410806033908e-05,
+      "loss": 1.1762,
+      "step": 1139
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9302585784782452e-05,
+      "loss": 1.2755,
+      "step": 1140
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9301061906197572e-05,
+      "loss": 1.1519,
+      "step": 1141
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.929953642484704e-05,
+      "loss": 0.9465,
+      "step": 1142
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9298009340993725e-05,
+      "loss": 0.5935,
+      "step": 1143
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9296480654900778e-05,
+      "loss": 1.0974,
+      "step": 1144
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9294950366831617e-05,
+      "loss": 1.1291,
+      "step": 1145
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9293418477049952e-05,
+      "loss": 0.6172,
+      "step": 1146
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.929188498581975e-05,
+      "loss": 1.1306,
+      "step": 1147
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9290349893405268e-05,
+      "loss": 1.1638,
+      "step": 1148
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9288813200071035e-05,
+      "loss": 0.9797,
+      "step": 1149
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9287274906081853e-05,
+      "loss": 0.8172,
+      "step": 1150
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9285735011702803e-05,
+      "loss": 0.9998,
+      "step": 1151
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9284193517199237e-05,
+      "loss": 1.1224,
+      "step": 1152
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.928265042283679e-05,
+      "loss": 0.8443,
+      "step": 1153
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9281105728881373e-05,
+      "loss": 0.8591,
+      "step": 1154
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9279559435599164e-05,
+      "loss": 1.0501,
+      "step": 1155
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.927801154325662e-05,
+      "loss": 0.8019,
+      "step": 1156
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9276462052120475e-05,
+      "loss": 0.5942,
+      "step": 1157
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9274910962457742e-05,
+      "loss": 0.7414,
+      "step": 1158
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9273358274535703e-05,
+      "loss": 1.0736,
+      "step": 1159
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9271803988621923e-05,
+      "loss": 1.0634,
+      "step": 1160
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.927024810498423e-05,
+      "loss": 0.8871,
+      "step": 1161
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9268690623890743e-05,
+      "loss": 0.9835,
+      "step": 1162
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.926713154560984e-05,
+      "loss": 0.763,
+      "step": 1163
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.926557087041019e-05,
+      "loss": 1.2195,
+      "step": 1164
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9264008598560727e-05,
+      "loss": 0.7374,
+      "step": 1165
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.926244473033066e-05,
+      "loss": 1.2687,
+      "step": 1166
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9260879265989477e-05,
+      "loss": 0.9142,
+      "step": 1167
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9259312205806946e-05,
+      "loss": 0.851,
+      "step": 1168
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9257743550053097e-05,
+      "loss": 0.6506,
+      "step": 1169
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9256173298998243e-05,
+      "loss": 0.7926,
+      "step": 1170
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9254601452912972e-05,
+      "loss": 0.8086,
+      "step": 1171
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9253028012068146e-05,
+      "loss": 0.9848,
+      "step": 1172
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.92514529767349e-05,
+      "loss": 1.0288,
+      "step": 1173
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9249876347184647e-05,
+      "loss": 1.2112,
+      "step": 1174
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.924829812368907e-05,
+      "loss": 0.8669,
+      "step": 1175
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9246718306520135e-05,
+      "loss": 1.0486,
+      "step": 1176
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.924513689595007e-05,
+      "loss": 0.6792,
+      "step": 1177
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9243553892251393e-05,
+      "loss": 0.9367,
+      "step": 1178
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.924196929569688e-05,
+      "loss": 0.692,
+      "step": 1179
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9240383106559594e-05,
+      "loss": 0.91,
+      "step": 1180
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 0.9462,
+      "step": 1181
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9237205951630315e-05,
+      "loss": 0.915,
+      "step": 1182
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9235614986385806e-05,
+      "loss": 0.9478,
+      "step": 1183
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.92340224296535e-05,
+      "loss": 1.0252,
+      "step": 1184
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.923242828170784e-05,
+      "loss": 1.1693,
+      "step": 1185
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9230832542823514e-05,
+      "loss": 0.8354,
+      "step": 1186
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.922923521327551e-05,
+      "loss": 0.686,
+      "step": 1187
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9227636293339077e-05,
+      "loss": 0.8476,
+      "step": 1188
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9226035783289748e-05,
+      "loss": 0.9872,
+      "step": 1189
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9224433683403318e-05,
+      "loss": 0.5585,
+      "step": 1190
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.922282999395586e-05,
+      "loss": 1.0537,
+      "step": 1191
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9221224715223735e-05,
+      "loss": 0.9639,
+      "step": 1192
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9219617847483553e-05,
+      "loss": 1.0952,
+      "step": 1193
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9218009391012217e-05,
+      "loss": 0.7972,
+      "step": 1194
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9216399346086893e-05,
+      "loss": 0.825,
+      "step": 1195
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.921478771298503e-05,
+      "loss": 0.9467,
+      "step": 1196
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9213174491984342e-05,
+      "loss": 0.9054,
+      "step": 1197
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9211559683362823e-05,
+      "loss": 0.8332,
+      "step": 1198
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9209943287398738e-05,
+      "loss": 1.4368,
+      "step": 1199
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9208325304370623e-05,
+      "loss": 0.9664,
+      "step": 1200
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9206705734557293e-05,
+      "loss": 0.8721,
+      "step": 1201
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.920508457823783e-05,
+      "loss": 1.1227,
+      "step": 1202
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9203461835691596e-05,
+      "loss": 0.6857,
+      "step": 1203
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9201837507198224e-05,
+      "loss": 0.9412,
+      "step": 1204
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9200211593037612e-05,
+      "loss": 0.7618,
+      "step": 1205
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9198584093489946e-05,
+      "loss": 1.0026,
+      "step": 1206
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9196955008835678e-05,
+      "loss": 0.802,
+      "step": 1207
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.919532433935553e-05,
+      "loss": 0.9566,
+      "step": 1208
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9193692085330503e-05,
+      "loss": 1.0098,
+      "step": 1209
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9192058247041862e-05,
+      "loss": 0.5655,
+      "step": 1210
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9190422824771158e-05,
+      "loss": 1.172,
+      "step": 1211
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9188785818800205e-05,
+      "loss": 0.5212,
+      "step": 1212
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9187147229411094e-05,
+      "loss": 0.7883,
+      "step": 1213
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9185507056886184e-05,
+      "loss": 0.5871,
+      "step": 1214
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.918386530150812e-05,
+      "loss": 0.84,
+      "step": 1215
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9182221963559803e-05,
+      "loss": 0.8348,
+      "step": 1216
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9180577043324415e-05,
+      "loss": 0.8396,
+      "step": 1217
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9178930541085408e-05,
+      "loss": 1.2515,
+      "step": 1218
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9177282457126515e-05,
+      "loss": 0.8904,
+      "step": 1219
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9175632791731732e-05,
+      "loss": 0.8697,
+      "step": 1220
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9173981545185327e-05,
+      "loss": 1.0595,
+      "step": 1221
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9172328717771846e-05,
+      "loss": 0.8295,
+      "step": 1222
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.917067430977611e-05,
+      "loss": 0.8634,
+      "step": 1223
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9169018321483198e-05,
+      "loss": 1.4225,
+      "step": 1224
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.916736075317848e-05,
+      "loss": 1.1889,
+      "step": 1225
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9165701605147586e-05,
+      "loss": 0.8789,
+      "step": 1226
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9164040877676425e-05,
+      "loss": 1.3559,
+      "step": 1227
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.916237857105117e-05,
+      "loss": 0.5961,
+      "step": 1228
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9160714685558272e-05,
+      "loss": 0.7763,
+      "step": 1229
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.915904922148445e-05,
+      "loss": 0.99,
+      "step": 1230
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9157382179116705e-05,
+      "loss": 1.1272,
+      "step": 1231
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9155713558742298e-05,
+      "loss": 1.3021,
+      "step": 1232
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.915404336064877e-05,
+      "loss": 1.5989,
+      "step": 1233
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9152371585123926e-05,
+      "loss": 0.6444,
+      "step": 1234
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9150698232455853e-05,
+      "loss": 0.5814,
+      "step": 1235
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.91490233029329e-05,
+      "loss": 1.2726,
+      "step": 1236
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9147346796843695e-05,
+      "loss": 0.8199,
+      "step": 1237
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9145668714477134e-05,
+      "loss": 1.1923,
+      "step": 1238
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9143989056122385e-05,
+      "loss": 0.9207,
+      "step": 1239
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9142307822068888e-05,
+      "loss": 0.8837,
+      "step": 1240
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9140625012606353e-05,
+      "loss": 0.8879,
+      "step": 1241
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9138940628024766e-05,
+      "loss": 0.9021,
+      "step": 1242
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.913725466861438e-05,
+      "loss": 1.0032,
+      "step": 1243
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9135567134665716e-05,
+      "loss": 0.6035,
+      "step": 1244
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.913387802646958e-05,
+      "loss": 0.7501,
+      "step": 1245
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9132187344317034e-05,
+      "loss": 1.0618,
+      "step": 1246
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9130495088499417e-05,
+      "loss": 0.9703,
+      "step": 1247
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9128801259308347e-05,
+      "loss": 0.9945,
+      "step": 1248
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9127105857035697e-05,
+      "loss": 1.2012,
+      "step": 1249
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9125408881973622e-05,
+      "loss": 1.2705,
+      "step": 1250
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9123710334414552e-05,
+      "loss": 0.9003,
+      "step": 1251
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9122010214651178e-05,
+      "loss": 1.1295,
+      "step": 1252
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9120308522976464e-05,
+      "loss": 0.7773,
+      "step": 1253
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.911860525968365e-05,
+      "loss": 0.6686,
+      "step": 1254
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.911690042506624e-05,
+      "loss": 1.2248,
+      "step": 1255
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9115194019418017e-05,
+      "loss": 1.1336,
+      "step": 1256
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9113486043033025e-05,
+      "loss": 0.7693,
+      "step": 1257
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.911177649620558e-05,
+      "loss": 0.7565,
+      "step": 1258
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.911006537923029e-05,
+      "loss": 0.601,
+      "step": 1259
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9108352692401997e-05,
+      "loss": 1.4508,
+      "step": 1260
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9106638436015842e-05,
+      "loss": 0.9008,
+      "step": 1261
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9104922610367223e-05,
+      "loss": 1.1163,
+      "step": 1262
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9103205215751815e-05,
+      "loss": 0.8841,
+      "step": 1263
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9101486252465558e-05,
+      "loss": 0.7989,
+      "step": 1264
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9099765720804666e-05,
+      "loss": 0.9186,
+      "step": 1265
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9098043621065624e-05,
+      "loss": 0.9559,
+      "step": 1266
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.7524,
+      "step": 1267
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9094594718540372e-05,
+      "loss": 0.9876,
+      "step": 1268
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.909286791634848e-05,
+      "loss": 1.0193,
+      "step": 1269
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9091139547267067e-05,
+      "loss": 1.0126,
+      "step": 1270
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.908940961159397e-05,
+      "loss": 0.8733,
+      "step": 1271
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.90876781096273e-05,
+      "loss": 0.7198,
+      "step": 1272
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9085945041665418e-05,
+      "loss": 0.9788,
+      "step": 1273
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.908421040800698e-05,
+      "loss": 0.973,
+      "step": 1274
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.908247420895089e-05,
+      "loss": 0.809,
+      "step": 1275
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9080736444796337e-05,
+      "loss": 1.0709,
+      "step": 1276
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9078997115842767e-05,
+      "loss": 1.1362,
+      "step": 1277
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.907725622238991e-05,
+      "loss": 0.8961,
+      "step": 1278
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9075513764737753e-05,
+      "loss": 1.1579,
+      "step": 1279
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.907376974318656e-05,
+      "loss": 0.9172,
+      "step": 1280
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.907202415803686e-05,
+      "loss": 0.6848,
+      "step": 1281
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9070277009589456e-05,
+      "loss": 0.8529,
+      "step": 1282
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9068528298145418e-05,
+      "loss": 0.9369,
+      "step": 1283
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9066778024006084e-05,
+      "loss": 0.7271,
+      "step": 1284
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.906502618747306e-05,
+      "loss": 1.0358,
+      "step": 1285
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9063272788848226e-05,
+      "loss": 1.3083,
+      "step": 1286
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.906151782843373e-05,
+      "loss": 1.0471,
+      "step": 1287
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9059761306531984e-05,
+      "loss": 1.1671,
+      "step": 1288
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9058003223445676e-05,
+      "loss": 0.7031,
+      "step": 1289
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9056243579477763e-05,
+      "loss": 0.6819,
+      "step": 1290
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.905448237493147e-05,
+      "loss": 0.5948,
+      "step": 1291
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9052719610110276e-05,
+      "loss": 0.7454,
+      "step": 1292
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9050955285317957e-05,
+      "loss": 0.6821,
+      "step": 1293
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.904918940085853e-05,
+      "loss": 1.0945,
+      "step": 1294
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.90474219570363e-05,
+      "loss": 1.1046,
+      "step": 1295
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9045652954155837e-05,
+      "loss": 0.9988,
+      "step": 1296
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9043882392521973e-05,
+      "loss": 0.9127,
+      "step": 1297
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9042110272439812e-05,
+      "loss": 0.9328,
+      "step": 1298
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9040336594214727e-05,
+      "loss": 1.1247,
+      "step": 1299
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.903856135815236e-05,
+      "loss": 1.1213,
+      "step": 1300
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9036784564558623e-05,
+      "loss": 1.0548,
+      "step": 1301
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.903500621373969e-05,
+      "loss": 1.1471,
+      "step": 1302
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9033226306002016e-05,
+      "loss": 1.1428,
+      "step": 1303
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9031444841652305e-05,
+      "loss": 1.0545,
+      "step": 1304
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9029661820997548e-05,
+      "loss": 0.9696,
+      "step": 1305
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9027877244344992e-05,
+      "loss": 1.2398,
+      "step": 1306
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9026091112002163e-05,
+      "loss": 0.76,
+      "step": 1307
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.902430342427684e-05,
+      "loss": 0.7723,
+      "step": 1308
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9022514181477085e-05,
+      "loss": 1.0508,
+      "step": 1309
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9020723383911214e-05,
+      "loss": 0.9945,
+      "step": 1310
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.901893103188783e-05,
+      "loss": 0.8327,
+      "step": 1311
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9017137125715786e-05,
+      "loss": 1.147,
+      "step": 1312
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9015341665704206e-05,
+      "loss": 1.2929,
+      "step": 1313
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.901354465216249e-05,
+      "loss": 1.2952,
+      "step": 1314
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.90117460854003e-05,
+      "loss": 0.6588,
+      "step": 1315
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9009945965727565e-05,
+      "loss": 1.1684,
+      "step": 1316
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9008144293454482e-05,
+      "loss": 0.7759,
+      "step": 1317
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.900634106889152e-05,
+      "loss": 1.2039,
+      "step": 1318
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9004536292349406e-05,
+      "loss": 1.171,
+      "step": 1319
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9002729964139148e-05,
+      "loss": 0.9394,
+      "step": 1320
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9000922084572008e-05,
+      "loss": 0.6855,
+      "step": 1321
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8999112653959524e-05,
+      "loss": 0.7712,
+      "step": 1322
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8997301672613496e-05,
+      "loss": 1.2156,
+      "step": 1323
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8995489140845995e-05,
+      "loss": 0.9544,
+      "step": 1324
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.899367505896936e-05,
+      "loss": 0.6627,
+      "step": 1325
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.899185942729619e-05,
+      "loss": 0.9871,
+      "step": 1326
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.899004224613936e-05,
+      "loss": 1.273,
+      "step": 1327
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8988223515812004e-05,
+      "loss": 0.673,
+      "step": 1328
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.898640323662753e-05,
+      "loss": 0.9722,
+      "step": 1329
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8984581408899615e-05,
+      "loss": 1.0711,
+      "step": 1330
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8982758032942184e-05,
+      "loss": 1.2383,
+      "step": 1331
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8980933109069455e-05,
+      "loss": 0.9922,
+      "step": 1332
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8979106637595895e-05,
+      "loss": 1.1308,
+      "step": 1333
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8977278618836242e-05,
+      "loss": 0.877,
+      "step": 1334
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8975449053105505e-05,
+      "loss": 0.9533,
+      "step": 1335
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8973617940718952e-05,
+      "loss": 0.5231,
+      "step": 1336
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8971785281992124e-05,
+      "loss": 0.8968,
+      "step": 1337
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.896995107724083e-05,
+      "loss": 1.2747,
+      "step": 1338
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.896811532678113e-05,
+      "loss": 0.8126,
+      "step": 1339
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8966278030929372e-05,
+      "loss": 1.0192,
+      "step": 1340
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8964439190002156e-05,
+      "loss": 0.9176,
+      "step": 1341
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8962598804316356e-05,
+      "loss": 0.4917,
+      "step": 1342
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8960756874189104e-05,
+      "loss": 0.8168,
+      "step": 1343
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8958913399937805e-05,
+      "loss": 0.9373,
+      "step": 1344
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.895706838188013e-05,
+      "loss": 0.6206,
+      "step": 1345
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.895522182033401e-05,
+      "loss": 0.842,
+      "step": 1346
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8953373715617646e-05,
+      "loss": 0.8915,
+      "step": 1347
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8951524068049508e-05,
+      "loss": 0.6094,
+      "step": 1348
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8949672877948324e-05,
+      "loss": 0.9084,
+      "step": 1349
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8947820145633095e-05,
+      "loss": 1.1829,
+      "step": 1350
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.894596587142309e-05,
+      "loss": 0.7651,
+      "step": 1351
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.894411005563783e-05,
+      "loss": 0.7805,
+      "step": 1352
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8942252698597113e-05,
+      "loss": 1.2607,
+      "step": 1353
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8940393800621003e-05,
+      "loss": 0.8192,
+      "step": 1354
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.893853336202983e-05,
+      "loss": 1.1411,
+      "step": 1355
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8936671383144175e-05,
+      "loss": 1.2939,
+      "step": 1356
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8934807864284904e-05,
+      "loss": 0.7382,
+      "step": 1357
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.893294280577314e-05,
+      "loss": 1.5348,
+      "step": 1358
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8931076207930265e-05,
+      "loss": 1.1912,
+      "step": 1359
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8929208071077936e-05,
+      "loss": 0.7771,
+      "step": 1360
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.892733839553808e-05,
+      "loss": 0.8085,
+      "step": 1361
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8925467181632866e-05,
+      "loss": 0.8066,
+      "step": 1362
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.892359442968475e-05,
+      "loss": 0.8708,
+      "step": 1363
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8921720140016447e-05,
+      "loss": 0.9223,
+      "step": 1364
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8919844312950937e-05,
+      "loss": 0.8496,
+      "step": 1365
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.891796694881146e-05,
+      "loss": 1.343,
+      "step": 1366
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8916088047921527e-05,
+      "loss": 1.0125,
+      "step": 1367
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8914207610604908e-05,
+      "loss": 0.7536,
+      "step": 1368
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8912325637185647e-05,
+      "loss": 0.85,
+      "step": 1369
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8910442127988043e-05,
+      "loss": 0.7431,
+      "step": 1370
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8908557083336668e-05,
+      "loss": 0.9318,
+      "step": 1371
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8906670503556348e-05,
+      "loss": 0.7626,
+      "step": 1372
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8904782388972186e-05,
+      "loss": 1.1896,
+      "step": 1373
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8902892739909537e-05,
+      "loss": 0.9777,
+      "step": 1374
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.890100155669403e-05,
+      "loss": 0.561,
+      "step": 1375
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8899108839651555e-05,
+      "loss": 0.6686,
+      "step": 1376
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.889721458910827e-05,
+      "loss": 0.8553,
+      "step": 1377
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8895318805390586e-05,
+      "loss": 0.5662,
+      "step": 1378
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.889342148882519e-05,
+      "loss": 1.3371,
+      "step": 1379
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8891522639739032e-05,
+      "loss": 0.8995,
+      "step": 1380
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8889622258459317e-05,
+      "loss": 0.787,
+      "step": 1381
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.888772034531352e-05,
+      "loss": 0.7946,
+      "step": 1382
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.888581690062939e-05,
+      "loss": 0.587,
+      "step": 1383
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8883911924734916e-05,
+      "loss": 1.3115,
+      "step": 1384
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8882005417958376e-05,
+      "loss": 0.9202,
+      "step": 1385
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8880097380628293e-05,
+      "loss": 0.8286,
+      "step": 1386
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8878187813073465e-05,
+      "loss": 1.4705,
+      "step": 1387
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.887627671562295e-05,
+      "loss": 0.9628,
+      "step": 1388
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.887436408860607e-05,
+      "loss": 0.8583,
+      "step": 1389
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.887244993235241e-05,
+      "loss": 1.0944,
+      "step": 1390
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8870534247191815e-05,
+      "loss": 0.8204,
+      "step": 1391
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8868617033454407e-05,
+      "loss": 1.0965,
+      "step": 1392
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.886669829147055e-05,
+      "loss": 0.8403,
+      "step": 1393
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8864778021570893e-05,
+      "loss": 1.3762,
+      "step": 1394
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.886285622408633e-05,
+      "loss": 1.051,
+      "step": 1395
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8860932899348028e-05,
+      "loss": 0.9983,
+      "step": 1396
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.885900804768742e-05,
+      "loss": 1.0954,
+      "step": 1397
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8857081669436194e-05,
+      "loss": 0.6496,
+      "step": 1398
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8855153764926307e-05,
+      "loss": 0.9932,
+      "step": 1399
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8853224334489977e-05,
+      "loss": 0.7226,
+      "step": 1400
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8851293378459685e-05,
+      "loss": 1.0407,
+      "step": 1401
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8849360897168174e-05,
+      "loss": 0.7477,
+      "step": 1402
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8847426890948447e-05,
+      "loss": 1.1394,
+      "step": 1403
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8845491360133777e-05,
+      "loss": 1.1813,
+      "step": 1404
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.88435543050577e-05,
+      "loss": 0.708,
+      "step": 1405
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8841615726054007e-05,
+      "loss": 0.999,
+      "step": 1406
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.883967562345675e-05,
+      "loss": 0.8443,
+      "step": 1407
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8837733997600258e-05,
+      "loss": 0.7017,
+      "step": 1408
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8835790848819105e-05,
+      "loss": 0.7899,
+      "step": 1409
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8833846177448142e-05,
+      "loss": 0.8381,
+      "step": 1410
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8831899983822475e-05,
+      "loss": 1.0577,
+      "step": 1411
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8829952268277472e-05,
+      "loss": 1.0643,
+      "step": 1412
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8828003031148764e-05,
+      "loss": 1.2376,
+      "step": 1413
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8826052272772247e-05,
+      "loss": 0.9268,
+      "step": 1414
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8824099993484075e-05,
+      "loss": 1.0386,
+      "step": 1415
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.882214619362067e-05,
+      "loss": 0.6675,
+      "step": 1416
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8820190873518708e-05,
+      "loss": 0.5123,
+      "step": 1417
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.881823403351513e-05,
+      "loss": 0.9878,
+      "step": 1418
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8816275673947148e-05,
+      "loss": 1.0922,
+      "step": 1419
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.881431579515222e-05,
+      "loss": 0.8532,
+      "step": 1420
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8812354397468076e-05,
+      "loss": 0.8147,
+      "step": 1421
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.881039148123271e-05,
+      "loss": 0.8008,
+      "step": 1422
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8808427046784365e-05,
+      "loss": 0.8072,
+      "step": 1423
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.880646109446156e-05,
+      "loss": 0.8374,
+      "step": 1424
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8804493624603064e-05,
+      "loss": 0.9066,
+      "step": 1425
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8802524637547917e-05,
+      "loss": 0.7158,
+      "step": 1426
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8800554133635417e-05,
+      "loss": 1.0102,
+      "step": 1427
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.879858211320512e-05,
+      "loss": 0.7802,
+      "step": 1428
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.879660857659685e-05,
+      "loss": 0.7074,
+      "step": 1429
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.879463352415068e-05,
+      "loss": 1.4759,
+      "step": 1430
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.879265695620696e-05,
+      "loss": 0.6972,
+      "step": 1431
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.879067887310629e-05,
+      "loss": 0.8322,
+      "step": 1432
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8788699275189538e-05,
+      "loss": 1.2763,
+      "step": 1433
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.878671816279783e-05,
+      "loss": 0.7146,
+      "step": 1434
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8784735536272543e-05,
+      "loss": 0.8005,
+      "step": 1435
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8782751395955335e-05,
+      "loss": 0.8848,
+      "step": 1436
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8780765742188113e-05,
+      "loss": 1.1667,
+      "step": 1437
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8778778575313045e-05,
+      "loss": 0.8795,
+      "step": 1438
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8776789895672557e-05,
+      "loss": 0.9408,
+      "step": 1439
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8774799703609347e-05,
+      "loss": 0.6532,
+      "step": 1440
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8772807999466362e-05,
+      "loss": 0.7584,
+      "step": 1441
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8770814783586813e-05,
+      "loss": 0.8925,
+      "step": 1442
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8768820056314173e-05,
+      "loss": 0.6608,
+      "step": 1443
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.876682381799218e-05,
+      "loss": 0.8696,
+      "step": 1444
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.876482606896482e-05,
+      "loss": 0.6308,
+      "step": 1445
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.876282680957635e-05,
+      "loss": 1.039,
+      "step": 1446
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8760826040171286e-05,
+      "loss": 0.8059,
+      "step": 1447
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8758823761094394e-05,
+      "loss": 1.0531,
+      "step": 1448
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.875681997269072e-05,
+      "loss": 0.8664,
+      "step": 1449
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.875481467530555e-05,
+      "loss": 1.0358,
+      "step": 1450
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.875280786928444e-05,
+      "loss": 0.5602,
+      "step": 1451
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.87507995549732e-05,
+      "loss": 0.8756,
+      "step": 1452
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.874878973271791e-05,
+      "loss": 1.1982,
+      "step": 1453
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8746778402864907e-05,
+      "loss": 0.7761,
+      "step": 1454
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8744765565760778e-05,
+      "loss": 0.9946,
+      "step": 1455
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8742751221752377e-05,
+      "loss": 0.9208,
+      "step": 1456
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8740735371186823e-05,
+      "loss": 0.8759,
+      "step": 1457
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.873871801441148e-05,
+      "loss": 0.9907,
+      "step": 1458
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.873669915177399e-05,
+      "loss": 1.0108,
+      "step": 1459
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8734678783622237e-05,
+      "loss": 0.9066,
+      "step": 1460
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8732656910304378e-05,
+      "loss": 0.9982,
+      "step": 1461
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.873063353216882e-05,
+      "loss": 0.8244,
+      "step": 1462
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.872860864956423e-05,
+      "loss": 0.7987,
+      "step": 1463
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8726582262839547e-05,
+      "loss": 0.9643,
+      "step": 1464
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.872455437234395e-05,
+      "loss": 0.7059,
+      "step": 1465
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.872252497842689e-05,
+      "loss": 1.1742,
+      "step": 1466
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.872049408143808e-05,
+      "loss": 0.8649,
+      "step": 1467
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.871846168172747e-05,
+      "loss": 0.9192,
+      "step": 1468
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8716427779645303e-05,
+      "loss": 0.7445,
+      "step": 1469
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.871439237554205e-05,
+      "loss": 0.8409,
+      "step": 1470
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8712355469768454e-05,
+      "loss": 1.1298,
+      "step": 1471
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.871031706267552e-05,
+      "loss": 1.646,
+      "step": 1472
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.870827715461451e-05,
+      "loss": 1.0634,
+      "step": 1473
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8706235745936935e-05,
+      "loss": 0.9249,
+      "step": 1474
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8704192836994578e-05,
+      "loss": 1.0535,
+      "step": 1475
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8702148428139474e-05,
+      "loss": 1.5523,
+      "step": 1476
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8700102519723912e-05,
+      "loss": 1.0361,
+      "step": 1477
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.869805511210045e-05,
+      "loss": 0.8436,
+      "step": 1478
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8696006205621898e-05,
+      "loss": 0.8947,
+      "step": 1479
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8693955800641323e-05,
+      "loss": 0.7865,
+      "step": 1480
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.869190389751205e-05,
+      "loss": 0.8118,
+      "step": 1481
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8689850496587674e-05,
+      "loss": 1.0939,
+      "step": 1482
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8687795598222024e-05,
+      "loss": 0.8651,
+      "step": 1483
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8685739202769215e-05,
+      "loss": 0.7579,
+      "step": 1484
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.86836813105836e-05,
+      "loss": 0.9007,
+      "step": 1485
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8681621922019797e-05,
+      "loss": 0.7295,
+      "step": 1486
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.867956103743268e-05,
+      "loss": 1.0987,
+      "step": 1487
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.867749865717739e-05,
+      "loss": 0.66,
+      "step": 1488
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8675434781609305e-05,
+      "loss": 0.8636,
+      "step": 1489
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.867336941108408e-05,
+      "loss": 0.7063,
+      "step": 1490
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8671302545957628e-05,
+      "loss": 0.8654,
+      "step": 1491
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8669234186586102e-05,
+      "loss": 0.7687,
+      "step": 1492
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8667164333325928e-05,
+      "loss": 0.9043,
+      "step": 1493
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8665092986533785e-05,
+      "loss": 0.8682,
+      "step": 1494
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8663020146566606e-05,
+      "loss": 0.9069,
+      "step": 1495
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8660945813781587e-05,
+      "loss": 0.8439,
+      "step": 1496
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8658869988536175e-05,
+      "loss": 0.698,
+      "step": 1497
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8656792671188083e-05,
+      "loss": 1.2938,
+      "step": 1498
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8654713862095272e-05,
+      "loss": 0.6007,
+      "step": 1499
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8652633561615965e-05,
+      "loss": 1.0152,
+      "step": 1500
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8650551770108644e-05,
+      "loss": 0.9309,
+      "step": 1501
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.864846848793204e-05,
+      "loss": 1.2961,
+      "step": 1502
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8646383715445144e-05,
+      "loss": 0.5843,
+      "step": 1503
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8644297453007212e-05,
+      "loss": 1.6724,
+      "step": 1504
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8642209700977745e-05,
+      "loss": 1.0209,
+      "step": 1505
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.864012045971651e-05,
+      "loss": 1.0614,
+      "step": 1506
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8638029729583524e-05,
+      "loss": 1.3302,
+      "step": 1507
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8635937510939063e-05,
+      "loss": 0.9098,
+      "step": 1508
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8633843804143663e-05,
+      "loss": 0.6856,
+      "step": 1509
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8631748609558107e-05,
+      "loss": 0.9819,
+      "step": 1510
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8629651927543447e-05,
+      "loss": 0.7985,
+      "step": 1511
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8627553758460978e-05,
+      "loss": 0.6917,
+      "step": 1512
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8625454102672266e-05,
+      "loss": 0.7204,
+      "step": 1513
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8623352960539117e-05,
+      "loss": 0.8353,
+      "step": 1514
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8621250332423603e-05,
+      "loss": 1.0792,
+      "step": 1515
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8619146218688056e-05,
+      "loss": 0.7521,
+      "step": 1516
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8617040619695055e-05,
+      "loss": 0.737,
+      "step": 1517
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.861493353580744e-05,
+      "loss": 0.9545,
+      "step": 1518
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8612824967388297e-05,
+      "loss": 0.8228,
+      "step": 1519
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8610714914800987e-05,
+      "loss": 0.6718,
+      "step": 1520
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8608603378409108e-05,
+      "loss": 0.5972,
+      "step": 1521
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.860649035857653e-05,
+      "loss": 0.9935,
+      "step": 1522
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.860437585566736e-05,
+      "loss": 0.9664,
+      "step": 1523
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8602259870045976e-05,
+      "loss": 1.0549,
+      "step": 1524
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8600142402077006e-05,
+      "loss": 0.9177,
+      "step": 1525
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8598023452125337e-05,
+      "loss": 0.618,
+      "step": 1526
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8595903020556098e-05,
+      "loss": 0.9634,
+      "step": 1527
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8593781107734693e-05,
+      "loss": 0.9926,
+      "step": 1528
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8591657714026764e-05,
+      "loss": 0.6989,
+      "step": 1529
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.858953283979822e-05,
+      "loss": 1.1652,
+      "step": 1530
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8587406485415226e-05,
+      "loss": 1.0129,
+      "step": 1531
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8585278651244188e-05,
+      "loss": 1.1173,
+      "step": 1532
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.858314933765178e-05,
+      "loss": 1.0074,
+      "step": 1533
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.858101854500493e-05,
+      "loss": 0.6768,
+      "step": 1534
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8578886273670807e-05,
+      "loss": 0.6371,
+      "step": 1535
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8576752524016855e-05,
+      "loss": 0.8583,
+      "step": 1536
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.857461729641076e-05,
+      "loss": 1.039,
+      "step": 1537
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8572480591220467e-05,
+      "loss": 0.9666,
+      "step": 1538
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8570342408814173e-05,
+      "loss": 0.7245,
+      "step": 1539
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8568202749560334e-05,
+      "loss": 0.9574,
+      "step": 1540
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.856606161382766e-05,
+      "loss": 1.1558,
+      "step": 1541
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.85639190019851e-05,
+      "loss": 0.9211,
+      "step": 1542
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8561774914401883e-05,
+      "loss": 1.0402,
+      "step": 1543
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.855962935144748e-05,
+      "loss": 1.2528,
+      "step": 1544
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8557482313491607e-05,
+      "loss": 1.2236,
+      "step": 1545
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8555333800904253e-05,
+      "loss": 0.9339,
+      "step": 1546
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.855318381405564e-05,
+      "loss": 0.9248,
+      "step": 1547
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8551032353316267e-05,
+      "loss": 0.4309,
+      "step": 1548
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.854887941905687e-05,
+      "loss": 0.7963,
+      "step": 1549
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.854672501164844e-05,
+      "loss": 0.9687,
+      "step": 1550
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8544569131462235e-05,
+      "loss": 1.2818,
+      "step": 1551
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.854241177886975e-05,
+      "loss": 0.9208,
+      "step": 1552
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8540252954242742e-05,
+      "loss": 0.9488,
+      "step": 1553
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8538092657953228e-05,
+      "loss": 0.9789,
+      "step": 1554
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8535930890373467e-05,
+      "loss": 0.7364,
+      "step": 1555
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8533767651875975e-05,
+      "loss": 0.6788,
+      "step": 1556
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8531602942833525e-05,
+      "loss": 0.9911,
+      "step": 1557
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8529436763619143e-05,
+      "loss": 0.9112,
+      "step": 1558
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8527269114606102e-05,
+      "loss": 0.6079,
+      "step": 1559
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8525099996167935e-05,
+      "loss": 1.421,
+      "step": 1560
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8522929408678426e-05,
+      "loss": 1.1257,
+      "step": 1561
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8520757352511616e-05,
+      "loss": 1.2997,
+      "step": 1562
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8518583828041787e-05,
+      "loss": 0.8801,
+      "step": 1563
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.851640883564349e-05,
+      "loss": 0.5895,
+      "step": 1564
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.851423237569152e-05,
+      "loss": 1.109,
+      "step": 1565
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.851205444856092e-05,
+      "loss": 1.7001,
+      "step": 1566
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8509875054627e-05,
+      "loss": 0.8888,
+      "step": 1567
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.850769419426531e-05,
+      "loss": 0.9414,
+      "step": 1568
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8505511867851656e-05,
+      "loss": 0.8884,
+      "step": 1569
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8503328075762102e-05,
+      "loss": 0.8514,
+      "step": 1570
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8501142818372964e-05,
+      "loss": 1.2059,
+      "step": 1571
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8498956096060797e-05,
+      "loss": 0.8272,
+      "step": 1572
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8496767909202423e-05,
+      "loss": 0.8478,
+      "step": 1573
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8494578258174918e-05,
+      "loss": 0.8492,
+      "step": 1574
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8492387143355594e-05,
+      "loss": 0.881,
+      "step": 1575
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8490194565122034e-05,
+      "loss": 0.9716,
+      "step": 1576
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.848800052385206e-05,
+      "loss": 0.5691,
+      "step": 1577
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8485805019923752e-05,
+      "loss": 0.7024,
+      "step": 1578
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.848360805371544e-05,
+      "loss": 0.8785,
+      "step": 1579
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8481409625605708e-05,
+      "loss": 0.7476,
+      "step": 1580
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.847920973597339e-05,
+      "loss": 1.0528,
+      "step": 1581
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8477008385197575e-05,
+      "loss": 0.9144,
+      "step": 1582
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.84748055736576e-05,
+      "loss": 1.346,
+      "step": 1583
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8472601301733047e-05,
+      "loss": 0.4696,
+      "step": 1584
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.847039556980377e-05,
+      "loss": 0.512,
+      "step": 1585
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8468188378249858e-05,
+      "loss": 1.0187,
+      "step": 1586
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8465979727451653e-05,
+      "loss": 0.5902,
+      "step": 1587
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8463769617789754e-05,
+      "loss": 0.8336,
+      "step": 1588
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.846155804964501e-05,
+      "loss": 0.9812,
+      "step": 1589
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8459345023398514e-05,
+      "loss": 1.1258,
+      "step": 1590
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8457130539431623e-05,
+      "loss": 0.8952,
+      "step": 1591
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8454914598125936e-05,
+      "loss": 1.0795,
+      "step": 1592
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8452697199863305e-05,
+      "loss": 1.1386,
+      "step": 1593
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8450478345025832e-05,
+      "loss": 0.8102,
+      "step": 1594
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8448258033995877e-05,
+      "loss": 1.2893,
+      "step": 1595
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.844603626715604e-05,
+      "loss": 0.9141,
+      "step": 1596
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8443813044889182e-05,
+      "loss": 0.9623,
+      "step": 1597
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8441588367578405e-05,
+      "loss": 0.7117,
+      "step": 1598
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8439362235607074e-05,
+      "loss": 0.8423,
+      "step": 1599
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8437134649358788e-05,
+      "loss": 0.8152,
+      "step": 1600
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8434905609217415e-05,
+      "loss": 1.373,
+      "step": 1601
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.843267511556706e-05,
+      "loss": 0.7881,
+      "step": 1602
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8430443168792087e-05,
+      "loss": 0.8423,
+      "step": 1603
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8428209769277107e-05,
+      "loss": 1.1305,
+      "step": 1604
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8425974917406974e-05,
+      "loss": 0.7358,
+      "step": 1605
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8423738613566806e-05,
+      "loss": 1.1168,
+      "step": 1606
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.842150085814196e-05,
+      "loss": 0.8614,
+      "step": 1607
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8419261651518053e-05,
+      "loss": 0.7585,
+      "step": 1608
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8417020994080945e-05,
+      "loss": 1.2857,
+      "step": 1609
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8414778886216744e-05,
+      "loss": 0.5264,
+      "step": 1610
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.7996,
+      "step": 1611
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8410290320752768e-05,
+      "loss": 1.1891,
+      "step": 1612
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8408043863926462e-05,
+      "loss": 0.9617,
+      "step": 1613
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8405795958220013e-05,
+      "loss": 0.9647,
+      "step": 1614
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.840354660402078e-05,
+      "loss": 0.8767,
+      "step": 1615
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8401295801716375e-05,
+      "loss": 0.8715,
+      "step": 1616
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8399043551694655e-05,
+      "loss": 0.7427,
+      "step": 1617
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8396789854343726e-05,
+      "loss": 0.9496,
+      "step": 1618
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8394534710051956e-05,
+      "loss": 0.8333,
+      "step": 1619
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8392278119207944e-05,
+      "loss": 0.6817,
+      "step": 1620
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8390020082200553e-05,
+      "loss": 1.1953,
+      "step": 1621
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.838776059941889e-05,
+      "loss": 1.0039,
+      "step": 1622
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8385499671252305e-05,
+      "loss": 1.034,
+      "step": 1623
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8383237298090408e-05,
+      "loss": 1.1376,
+      "step": 1624
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8380973480323052e-05,
+      "loss": 0.767,
+      "step": 1625
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8378708218340336e-05,
+      "loss": 0.892,
+      "step": 1626
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8376441512532617e-05,
+      "loss": 1.4894,
+      "step": 1627
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.837417336329049e-05,
+      "loss": 0.3494,
+      "step": 1628
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8371903771004812e-05,
+      "loss": 0.5374,
+      "step": 1629
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8369632736066673e-05,
+      "loss": 0.8769,
+      "step": 1630
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8367360258867418e-05,
+      "loss": 1.0445,
+      "step": 1631
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.836508633979865e-05,
+      "loss": 0.9009,
+      "step": 1632
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8362810979252206e-05,
+      "loss": 1.1726,
+      "step": 1633
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.836053417762018e-05,
+      "loss": 1.0203,
+      "step": 1634
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.835825593529492e-05,
+      "loss": 0.7018,
+      "step": 1635
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8355976252668996e-05,
+      "loss": 0.5793,
+      "step": 1636
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8353695130135262e-05,
+      "loss": 0.7308,
+      "step": 1637
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.835141256808679e-05,
+      "loss": 0.8243,
+      "step": 1638
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.834912856691692e-05,
+      "loss": 0.9348,
+      "step": 1639
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8346843127019234e-05,
+      "loss": 0.5903,
+      "step": 1640
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8344556248787556e-05,
+      "loss": 0.9843,
+      "step": 1641
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8342267932615962e-05,
+      "loss": 0.8209,
+      "step": 1642
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.833997817889878e-05,
+      "loss": 0.602,
+      "step": 1643
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8337686988030584e-05,
+      "loss": 1.183,
+      "step": 1644
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8335394360406184e-05,
+      "loss": 0.9021,
+      "step": 1645
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.833310029642066e-05,
+      "loss": 0.9465,
+      "step": 1646
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8330804796469314e-05,
+      "loss": 1.3885,
+      "step": 1647
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8328507860947718e-05,
+      "loss": 0.9377,
+      "step": 1648
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8326209490251678e-05,
+      "loss": 1.1174,
+      "step": 1649
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.832390968477725e-05,
+      "loss": 1.5199,
+      "step": 1650
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8321608444920738e-05,
+      "loss": 0.8447,
+      "step": 1651
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8319305771078696e-05,
+      "loss": 0.7095,
+      "step": 1652
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.831700166364792e-05,
+      "loss": 1.0705,
+      "step": 1653
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8314696123025456e-05,
+      "loss": 0.6815,
+      "step": 1654
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8312389149608595e-05,
+      "loss": 1.1792,
+      "step": 1655
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.831008074379488e-05,
+      "loss": 0.7358,
+      "step": 1656
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8307770905982092e-05,
+      "loss": 0.9142,
+      "step": 1657
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.830545963656827e-05,
+      "loss": 0.941,
+      "step": 1658
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.830314693595169e-05,
+      "loss": 0.7598,
+      "step": 1659
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8300832804530877e-05,
+      "loss": 0.6559,
+      "step": 1660
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8298517242704607e-05,
+      "loss": 1.1319,
+      "step": 1661
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8296200250871897e-05,
+      "loss": 1.279,
+      "step": 1662
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8293881829432015e-05,
+      "loss": 0.684,
+      "step": 1663
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.829156197878447e-05,
+      "loss": 0.6668,
+      "step": 1664
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.828924069932902e-05,
+      "loss": 0.9145,
+      "step": 1665
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8286917991465674e-05,
+      "loss": 1.0257,
+      "step": 1666
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.828459385559468e-05,
+      "loss": 1.0696,
+      "step": 1667
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8282268292116536e-05,
+      "loss": 1.0375,
+      "step": 1668
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.827994130143198e-05,
+      "loss": 0.973,
+      "step": 1669
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.827761288394201e-05,
+      "loss": 0.9831,
+      "step": 1670
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8275283040047848e-05,
+      "loss": 0.9126,
+      "step": 1671
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8272951770150987e-05,
+      "loss": 1.124,
+      "step": 1672
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8270619074653143e-05,
+      "loss": 0.8471,
+      "step": 1673
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8268284953956293e-05,
+      "loss": 0.8864,
+      "step": 1674
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8265949408462657e-05,
+      "loss": 1.3291,
+      "step": 1675
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.826361243857469e-05,
+      "loss": 1.1018,
+      "step": 1676
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8261274044695106e-05,
+      "loss": 1.1252,
+      "step": 1677
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8258934227226856e-05,
+      "loss": 0.6155,
+      "step": 1678
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.825659298657314e-05,
+      "loss": 0.5234,
+      "step": 1679
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8254250323137407e-05,
+      "loss": 0.6481,
+      "step": 1680
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8251906237323338e-05,
+      "loss": 0.9985,
+      "step": 1681
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.824956072953487e-05,
+      "loss": 0.9365,
+      "step": 1682
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8247213800176192e-05,
+      "loss": 0.8506,
+      "step": 1683
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8244865449651716e-05,
+      "loss": 0.7187,
+      "step": 1684
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8242515678366114e-05,
+      "loss": 0.927,
+      "step": 1685
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8240164486724305e-05,
+      "loss": 0.7319,
+      "step": 1686
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8237811875131446e-05,
+      "loss": 0.9628,
+      "step": 1687
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.823545784399294e-05,
+      "loss": 0.744,
+      "step": 1688
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8233102393714435e-05,
+      "loss": 0.9068,
+      "step": 1689
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.823074552470182e-05,
+      "loss": 0.7367,
+      "step": 1690
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8228387237361245e-05,
+      "loss": 0.7525,
+      "step": 1691
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.822602753209908e-05,
+      "loss": 0.5501,
+      "step": 1692
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8223666409321953e-05,
+      "loss": 1.019,
+      "step": 1693
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8221303869436738e-05,
+      "loss": 0.7139,
+      "step": 1694
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8218939912850547e-05,
+      "loss": 1.2624,
+      "step": 1695
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.821657453997074e-05,
+      "loss": 0.8207,
+      "step": 1696
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8214207751204917e-05,
+      "loss": 0.7859,
+      "step": 1697
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.821183954696093e-05,
+      "loss": 0.8942,
+      "step": 1698
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8209469927646863e-05,
+      "loss": 0.9378,
+      "step": 1699
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8207098893671054e-05,
+      "loss": 0.9628,
+      "step": 1700
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8204726445442084e-05,
+      "loss": 0.9162,
+      "step": 1701
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.820235258336877e-05,
+      "loss": 0.9792,
+      "step": 1702
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8199977307860178e-05,
+      "loss": 0.8397,
+      "step": 1703
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8197600619325622e-05,
+      "loss": 0.8388,
+      "step": 1704
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8195222518174654e-05,
+      "loss": 0.6026,
+      "step": 1705
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8192843004817063e-05,
+      "loss": 0.827,
+      "step": 1706
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8190462079662897e-05,
+      "loss": 0.9253,
+      "step": 1707
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.818807974312243e-05,
+      "loss": 1.0776,
+      "step": 1708
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8185695995606196e-05,
+      "loss": 0.8291,
+      "step": 1709
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8183310837524965e-05,
+      "loss": 0.7661,
+      "step": 1710
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8180924269289737e-05,
+      "loss": 1.2197,
+      "step": 1711
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.817853629131178e-05,
+      "loss": 1.1421,
+      "step": 1712
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8176146904002587e-05,
+      "loss": 0.6034,
+      "step": 1713
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.81737561077739e-05,
+      "loss": 1.015,
+      "step": 1714
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.81713639030377e-05,
+      "loss": 0.8046,
+      "step": 1715
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8168970290206213e-05,
+      "loss": 0.8654,
+      "step": 1716
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8166575269691914e-05,
+      "loss": 0.6145,
+      "step": 1717
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8164178841907508e-05,
+      "loss": 1.0808,
+      "step": 1718
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8161781007265958e-05,
+      "loss": 1.2165,
+      "step": 1719
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8159381766180453e-05,
+      "loss": 0.785,
+      "step": 1720
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.815698111906443e-05,
+      "loss": 0.6653,
+      "step": 1721
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8154579066331577e-05,
+      "loss": 0.6981,
+      "step": 1722
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8152175608395814e-05,
+      "loss": 1.4175,
+      "step": 1723
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8149770745671308e-05,
+      "loss": 0.6102,
+      "step": 1724
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8147364478572463e-05,
+      "loss": 0.7399,
+      "step": 1725
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8144956807513933e-05,
+      "loss": 0.7851,
+      "step": 1726
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8142547732910606e-05,
+      "loss": 0.6379,
+      "step": 1727
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.814013725517762e-05,
+      "loss": 1.0369,
+      "step": 1728
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.813772537473035e-05,
+      "loss": 0.9569,
+      "step": 1729
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.813531209198441e-05,
+      "loss": 1.0077,
+      "step": 1730
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8132897407355657e-05,
+      "loss": 0.9006,
+      "step": 1731
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8130481321260193e-05,
+      "loss": 0.8361,
+      "step": 1732
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.812806383411436e-05,
+      "loss": 0.8414,
+      "step": 1733
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.812564494633474e-05,
+      "loss": 1.1887,
+      "step": 1734
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8123224658338166e-05,
+      "loss": 0.7373,
+      "step": 1735
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8120802970541687e-05,
+      "loss": 0.4915,
+      "step": 1736
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8118379883362626e-05,
+      "loss": 0.6693,
+      "step": 1737
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8115955397218524e-05,
+      "loss": 0.5406,
+      "step": 1738
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.811352951252717e-05,
+      "loss": 0.6501,
+      "step": 1739
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8111102229706593e-05,
+      "loss": 0.6049,
+      "step": 1740
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.810867354917507e-05,
+      "loss": 0.5617,
+      "step": 1741
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8106243471351105e-05,
+      "loss": 0.5624,
+      "step": 1742
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8103811996653458e-05,
+      "loss": 0.7127,
+      "step": 1743
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8101379125501116e-05,
+      "loss": 0.4516,
+      "step": 1744
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.809894485831332e-05,
+      "loss": 0.5177,
+      "step": 1745
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8096509195509536e-05,
+      "loss": 0.968,
+      "step": 1746
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.809407213750949e-05,
+      "loss": 0.6109,
+      "step": 1747
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8091633684733132e-05,
+      "loss": 1.0129,
+      "step": 1748
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8089193837600653e-05,
+      "loss": 0.6806,
+      "step": 1749
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8086752596532497e-05,
+      "loss": 0.6258,
+      "step": 1750
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8084309961949338e-05,
+      "loss": 0.676,
+      "step": 1751
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.808186593427209e-05,
+      "loss": 0.6069,
+      "step": 1752
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8079420513921913e-05,
+      "loss": 0.6739,
+      "step": 1753
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.80769737013202e-05,
+      "loss": 0.5008,
+      "step": 1754
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.807452549688859e-05,
+      "loss": 0.7011,
+      "step": 1755
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.807207590104896e-05,
+      "loss": 0.7604,
+      "step": 1756
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8069624914223425e-05,
+      "loss": 0.5797,
+      "step": 1757
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8067172536834343e-05,
+      "loss": 0.5233,
+      "step": 1758
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8064718769304304e-05,
+      "loss": 0.6166,
+      "step": 1759
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8062263612056145e-05,
+      "loss": 0.8225,
+      "step": 1760
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8059807065512945e-05,
+      "loss": 0.8383,
+      "step": 1761
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8057349130098014e-05,
+      "loss": 0.9657,
+      "step": 1762
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8054889806234906e-05,
+      "loss": 0.8804,
+      "step": 1763
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.805242909434741e-05,
+      "loss": 0.6463,
+      "step": 1764
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8049966994859565e-05,
+      "loss": 0.7468,
+      "step": 1765
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8047503508195636e-05,
+      "loss": 0.8217,
+      "step": 1766
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8045038634780136e-05,
+      "loss": 0.4711,
+      "step": 1767
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.804257237503781e-05,
+      "loss": 1.0406,
+      "step": 1768
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8040104729393652e-05,
+      "loss": 0.6552,
+      "step": 1769
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8037635698272878e-05,
+      "loss": 0.6801,
+      "step": 1770
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8035165282100963e-05,
+      "loss": 0.7727,
+      "step": 1771
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8032693481303604e-05,
+      "loss": 0.5532,
+      "step": 1772
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8030220296306752e-05,
+      "loss": 0.5827,
+      "step": 1773
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8027745727536578e-05,
+      "loss": 0.7761,
+      "step": 1774
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.802526977541951e-05,
+      "loss": 0.5919,
+      "step": 1775
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.80227924403822e-05,
+      "loss": 0.8207,
+      "step": 1776
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8020313722851547e-05,
+      "loss": 0.5281,
+      "step": 1777
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8017833623254685e-05,
+      "loss": 0.4988,
+      "step": 1778
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8015352142018984e-05,
+      "loss": 0.4161,
+      "step": 1779
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.801286927957206e-05,
+      "loss": 0.6046,
+      "step": 1780
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.801038503634176e-05,
+      "loss": 0.5877,
+      "step": 1781
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.800789941275617e-05,
+      "loss": 0.5421,
+      "step": 1782
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8005412409243604e-05,
+      "loss": 0.5874,
+      "step": 1783
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8002924026232643e-05,
+      "loss": 0.6863,
+      "step": 1784
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8000434264152074e-05,
+      "loss": 0.677,
+      "step": 1785
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7997943123430936e-05,
+      "loss": 0.8075,
+      "step": 1786
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.799545060449851e-05,
+      "loss": 0.7307,
+      "step": 1787
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7992956707784307e-05,
+      "loss": 0.6866,
+      "step": 1788
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.799046143371807e-05,
+      "loss": 0.7648,
+      "step": 1789
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7987964782729793e-05,
+      "loss": 0.7349,
+      "step": 1790
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7985466755249703e-05,
+      "loss": 0.7658,
+      "step": 1791
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7982967351708253e-05,
+      "loss": 0.8237,
+      "step": 1792
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.798046657253615e-05,
+      "loss": 0.6271,
+      "step": 1793
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7977964418164328e-05,
+      "loss": 0.7649,
+      "step": 1794
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.797546088902396e-05,
+      "loss": 0.8457,
+      "step": 1795
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7972955985546452e-05,
+      "loss": 0.6207,
+      "step": 1796
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7970449708163455e-05,
+      "loss": 0.7756,
+      "step": 1797
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.796794205730685e-05,
+      "loss": 0.5519,
+      "step": 1798
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.796543303340876e-05,
+      "loss": 0.5001,
+      "step": 1799
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.796292263690154e-05,
+      "loss": 0.8946,
+      "step": 1800
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7960410868217786e-05,
+      "loss": 0.6339,
+      "step": 1801
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7957897727790327e-05,
+      "loss": 0.7323,
+      "step": 1802
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7955383216052224e-05,
+      "loss": 0.6311,
+      "step": 1803
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.795286733343678e-05,
+      "loss": 0.742,
+      "step": 1804
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7950350080377545e-05,
+      "loss": 0.3959,
+      "step": 1805
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7947831457308276e-05,
+      "loss": 0.5645,
+      "step": 1806
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7945311464663002e-05,
+      "loss": 0.634,
+      "step": 1807
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.794279010287596e-05,
+      "loss": 0.4558,
+      "step": 1808
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.794026737238163e-05,
+      "loss": 0.6115,
+      "step": 1809
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.793774327361474e-05,
+      "loss": 0.8031,
+      "step": 1810
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7935217807010238e-05,
+      "loss": 0.5608,
+      "step": 1811
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7932690973003316e-05,
+      "loss": 0.5464,
+      "step": 1812
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.79301627720294e-05,
+      "loss": 0.4197,
+      "step": 1813
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7927633204524153e-05,
+      "loss": 0.51,
+      "step": 1814
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7925102270923465e-05,
+      "loss": 0.7034,
+      "step": 1815
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.792256997166348e-05,
+      "loss": 0.5432,
+      "step": 1816
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7920036307180554e-05,
+      "loss": 0.7405,
+      "step": 1817
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7917501277911295e-05,
+      "loss": 0.4201,
+      "step": 1818
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7914964884292543e-05,
+      "loss": 0.9322,
+      "step": 1819
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.791242712676137e-05,
+      "loss": 0.4983,
+      "step": 1820
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7909888005755077e-05,
+      "loss": 0.7574,
+      "step": 1821
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.790734752171122e-05,
+      "loss": 0.5758,
+      "step": 1822
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7904805675067568e-05,
+      "loss": 0.4325,
+      "step": 1823
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7902262466262136e-05,
+      "loss": 0.5053,
+      "step": 1824
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7899717895733175e-05,
+      "loss": 0.5162,
+      "step": 1825
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.789717196391916e-05,
+      "loss": 0.9115,
+      "step": 1826
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7894624671258813e-05,
+      "loss": 0.7502,
+      "step": 1827
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7892076018191085e-05,
+      "loss": 0.7192,
+      "step": 1828
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.788952600515516e-05,
+      "loss": 0.5429,
+      "step": 1829
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.788697463259046e-05,
+      "loss": 0.8751,
+      "step": 1830
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.788442190093664e-05,
+      "loss": 0.4343,
+      "step": 1831
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7881867810633582e-05,
+      "loss": 0.5883,
+      "step": 1832
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7879312362121414e-05,
+      "loss": 0.3509,
+      "step": 1833
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7876755555840498e-05,
+      "loss": 0.6079,
+      "step": 1834
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7874197392231414e-05,
+      "loss": 0.5354,
+      "step": 1835
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7871637871734994e-05,
+      "loss": 0.7175,
+      "step": 1836
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.786907699479229e-05,
+      "loss": 0.6706,
+      "step": 1837
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.78665147618446e-05,
+      "loss": 0.4389,
+      "step": 1838
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.786395117333345e-05,
+      "loss": 0.5751,
+      "step": 1839
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7861386229700597e-05,
+      "loss": 0.5922,
+      "step": 1840
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.785881993138803e-05,
+      "loss": 0.4846,
+      "step": 1841
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7856252278837987e-05,
+      "loss": 0.6419,
+      "step": 1842
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7853683272492913e-05,
+      "loss": 0.4507,
+      "step": 1843
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7851112912795513e-05,
+      "loss": 0.5721,
+      "step": 1844
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7848541200188708e-05,
+      "loss": 0.751,
+      "step": 1845
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7845968135115657e-05,
+      "loss": 0.9964,
+      "step": 1846
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7843393718019753e-05,
+      "loss": 0.6348,
+      "step": 1847
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.784081794934462e-05,
+      "loss": 0.4113,
+      "step": 1848
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7838240829534125e-05,
+      "loss": 0.7642,
+      "step": 1849
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7835662359032345e-05,
+      "loss": 0.6739,
+      "step": 1850
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7833082538283615e-05,
+      "loss": 0.4774,
+      "step": 1851
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7830501367732484e-05,
+      "loss": 0.6662,
+      "step": 1852
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7827918847823746e-05,
+      "loss": 0.4888,
+      "step": 1853
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7825334979002425e-05,
+      "loss": 0.5941,
+      "step": 1854
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7822749761713767e-05,
+      "loss": 0.6743,
+      "step": 1855
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7820163196403266e-05,
+      "loss": 0.5123,
+      "step": 1856
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7817575283516637e-05,
+      "loss": 0.5098,
+      "step": 1857
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.781498602349983e-05,
+      "loss": 0.5561,
+      "step": 1858
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7812395416799034e-05,
+      "loss": 0.4527,
+      "step": 1859
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7809803463860658e-05,
+      "loss": 0.8124,
+      "step": 1860
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7807210165131352e-05,
+      "loss": 0.5585,
+      "step": 1861
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7804615521057993e-05,
+      "loss": 0.7108,
+      "step": 1862
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7802019532087692e-05,
+      "loss": 0.9543,
+      "step": 1863
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7799422198667798e-05,
+      "loss": 0.6701,
+      "step": 1864
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7796823521245876e-05,
+      "loss": 0.7742,
+      "step": 1865
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.779422350026974e-05,
+      "loss": 0.661,
+      "step": 1866
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7791622136187422e-05,
+      "loss": 0.7269,
+      "step": 1867
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7789019429447192e-05,
+      "loss": 0.8751,
+      "step": 1868
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.778641538049755e-05,
+      "loss": 0.7153,
+      "step": 1869
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7783809989787234e-05,
+      "loss": 0.5632,
+      "step": 1870
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7781203257765197e-05,
+      "loss": 0.6984,
+      "step": 1871
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.777859518488064e-05,
+      "loss": 0.5866,
+      "step": 1872
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7775985771582986e-05,
+      "loss": 0.6133,
+      "step": 1873
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.777337501832189e-05,
+      "loss": 0.7626,
+      "step": 1874
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7770762925547235e-05,
+      "loss": 0.6351,
+      "step": 1875
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.776814949370915e-05,
+      "loss": 0.555,
+      "step": 1876
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7765534723257972e-05,
+      "loss": 0.771,
+      "step": 1877
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7762918614644283e-05,
+      "loss": 0.64,
+      "step": 1878
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7760301168318897e-05,
+      "loss": 1.0157,
+      "step": 1879
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.775768238473285e-05,
+      "loss": 0.7011,
+      "step": 1880
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7755062264337412e-05,
+      "loss": 0.6854,
+      "step": 1881
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.775244080758409e-05,
+      "loss": 0.6449,
+      "step": 1882
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7749818014924612e-05,
+      "loss": 0.3599,
+      "step": 1883
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7747193886810935e-05,
+      "loss": 0.8309,
+      "step": 1884
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7744568423695257e-05,
+      "loss": 0.5504,
+      "step": 1885
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.774194162603e-05,
+      "loss": 0.7901,
+      "step": 1886
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7739313494267808e-05,
+      "loss": 0.5587,
+      "step": 1887
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.773668402886157e-05,
+      "loss": 1.0259,
+      "step": 1888
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7734053230264394e-05,
+      "loss": 0.5993,
+      "step": 1889
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7731421098929625e-05,
+      "loss": 0.6405,
+      "step": 1890
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7728787635310828e-05,
+      "loss": 0.5086,
+      "step": 1891
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7726152839861806e-05,
+      "loss": 0.6217,
+      "step": 1892
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7723516713036588e-05,
+      "loss": 0.7445,
+      "step": 1893
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7720879255289436e-05,
+      "loss": 0.8307,
+      "step": 1894
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7718240467074833e-05,
+      "loss": 0.7947,
+      "step": 1895
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7715600348847508e-05,
+      "loss": 0.5594,
+      "step": 1896
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7712958901062397e-05,
+      "loss": 0.6092,
+      "step": 1897
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7710316124174677e-05,
+      "loss": 0.5784,
+      "step": 1898
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.770767201863976e-05,
+      "loss": 0.408,
+      "step": 1899
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7705026584913274e-05,
+      "loss": 1.0004,
+      "step": 1900
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7702379823451087e-05,
+      "loss": 0.6774,
+      "step": 1901
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.769973173470928e-05,
+      "loss": 0.822,
+      "step": 1902
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.769708231914419e-05,
+      "loss": 0.559,
+      "step": 1903
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7694431577212356e-05,
+      "loss": 0.3774,
+      "step": 1904
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.769177950937056e-05,
+      "loss": 0.4788,
+      "step": 1905
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.76891261160758e-05,
+      "loss": 1.0332,
+      "step": 1906
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7686471397785322e-05,
+      "loss": 0.6953,
+      "step": 1907
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7683815354956583e-05,
+      "loss": 0.5973,
+      "step": 1908
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7681157988047272e-05,
+      "loss": 0.6944,
+      "step": 1909
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.767849929751531e-05,
+      "loss": 0.4635,
+      "step": 1910
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7675839283818848e-05,
+      "loss": 0.9001,
+      "step": 1911
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7673177947416258e-05,
+      "loss": 0.535,
+      "step": 1912
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7670515288766146e-05,
+      "loss": 0.7833,
+      "step": 1913
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7667851308327338e-05,
+      "loss": 0.6265,
+      "step": 1914
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.76651860065589e-05,
+      "loss": 0.4397,
+      "step": 1915
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7662519383920114e-05,
+      "loss": 0.6514,
+      "step": 1916
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7659851440870498e-05,
+      "loss": 0.6146,
+      "step": 1917
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.765718217786979e-05,
+      "loss": 0.7826,
+      "step": 1918
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7654511595377956e-05,
+      "loss": 0.4035,
+      "step": 1919
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7651839693855203e-05,
+      "loss": 0.5438,
+      "step": 1920
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7649166473761947e-05,
+      "loss": 1.1535,
+      "step": 1921
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7646491935558844e-05,
+      "loss": 0.4168,
+      "step": 1922
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.764381607970677e-05,
+      "loss": 0.5216,
+      "step": 1923
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7641138906666828e-05,
+      "loss": 0.6655,
+      "step": 1924
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7638460416900355e-05,
+      "loss": 0.657,
+      "step": 1925
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7635780610868906e-05,
+      "loss": 0.6581,
+      "step": 1926
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7633099489034272e-05,
+      "loss": 0.797,
+      "step": 1927
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.763041705185846e-05,
+      "loss": 0.2898,
+      "step": 1928
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7627733299803714e-05,
+      "loss": 0.5058,
+      "step": 1929
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7625048233332494e-05,
+      "loss": 0.6193,
+      "step": 1930
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7622361852907506e-05,
+      "loss": 0.7814,
+      "step": 1931
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7619674158991654e-05,
+      "loss": 0.8803,
+      "step": 1932
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7616985152048092e-05,
+      "loss": 0.56,
+      "step": 1933
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.761429483254019e-05,
+      "loss": 1.0577,
+      "step": 1934
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.761160320093154e-05,
+      "loss": 0.7272,
+      "step": 1935
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7608910257685973e-05,
+      "loss": 0.5159,
+      "step": 1936
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.760621600326754e-05,
+      "loss": 0.7017,
+      "step": 1937
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7603520438140507e-05,
+      "loss": 0.7425,
+      "step": 1938
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.760082356276939e-05,
+      "loss": 0.7313,
+      "step": 1939
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7598125377618904e-05,
+      "loss": 0.457,
+      "step": 1940
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.759542588315401e-05,
+      "loss": 0.4953,
+      "step": 1941
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.759272507983988e-05,
+      "loss": 0.6931,
+      "step": 1942
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7590022968141928e-05,
+      "loss": 0.4024,
+      "step": 1943
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7587319548525772e-05,
+      "loss": 0.6655,
+      "step": 1944
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7584614821457275e-05,
+      "loss": 0.5118,
+      "step": 1945
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7581908787402516e-05,
+      "loss": 0.8595,
+      "step": 1946
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.75792014468278e-05,
+      "loss": 0.4893,
+      "step": 1947
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7576492800199657e-05,
+      "loss": 0.5483,
+      "step": 1948
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7573782847984848e-05,
+      "loss": 0.7115,
+      "step": 1949
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7571071590650348e-05,
+      "loss": 0.4002,
+      "step": 1950
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7568359028663365e-05,
+      "loss": 0.8394,
+      "step": 1951
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.756564516249133e-05,
+      "loss": 0.4715,
+      "step": 1952
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7562929992601895e-05,
+      "loss": 0.608,
+      "step": 1953
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7560213519462945e-05,
+      "loss": 0.3881,
+      "step": 1954
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.8236,
+      "step": 1955
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7554776665309138e-05,
+      "loss": 0.4796,
+      "step": 1956
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7552056285231166e-05,
+      "loss": 0.6489,
+      "step": 1957
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7549334603777436e-05,
+      "loss": 0.5731,
+      "step": 1958
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.754661162141696e-05,
+      "loss": 0.6179,
+      "step": 1959
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7543887338618955e-05,
+      "loss": 0.745,
+      "step": 1960
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7541161755852884e-05,
+      "loss": 0.681,
+      "step": 1961
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.753843487358841e-05,
+      "loss": 0.7347,
+      "step": 1962
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7535706692295436e-05,
+      "loss": 0.5427,
+      "step": 1963
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7532977212444083e-05,
+      "loss": 0.7693,
+      "step": 1964
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.75302464345047e-05,
+      "loss": 0.7056,
+      "step": 1965
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7527514358947853e-05,
+      "loss": 0.6616,
+      "step": 1966
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7524780986244334e-05,
+      "loss": 0.8217,
+      "step": 1967
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7522046316865165e-05,
+      "loss": 0.5064,
+      "step": 1968
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.751931035128158e-05,
+      "loss": 0.8797,
+      "step": 1969
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7516573089965047e-05,
+      "loss": 0.6006,
+      "step": 1970
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7513834533387256e-05,
+      "loss": 0.7127,
+      "step": 1971
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7511094682020107e-05,
+      "loss": 0.6063,
+      "step": 1972
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.750835353633574e-05,
+      "loss": 0.6593,
+      "step": 1973
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7505611096806512e-05,
+      "loss": 0.9297,
+      "step": 1974
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7502867363905e-05,
+      "loss": 0.5097,
+      "step": 1975
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7500122338104005e-05,
+      "loss": 0.8518,
+      "step": 1976
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.749737601987655e-05,
+      "loss": 0.5821,
+      "step": 1977
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.749462840969589e-05,
+      "loss": 0.5809,
+      "step": 1978
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7491879508035488e-05,
+      "loss": 0.6211,
+      "step": 1979
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7489129315369043e-05,
+      "loss": 0.7245,
+      "step": 1980
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7486377832170463e-05,
+      "loss": 0.8808,
+      "step": 1981
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7483625058913887e-05,
+      "loss": 0.6284,
+      "step": 1982
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7480870996073682e-05,
+      "loss": 0.4367,
+      "step": 1983
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.747811564412442e-05,
+      "loss": 0.9272,
+      "step": 1984
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7475359003540916e-05,
+      "loss": 0.5227,
+      "step": 1985
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.747260107479819e-05,
+      "loss": 0.757,
+      "step": 1986
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.746984185837149e-05,
+      "loss": 0.4533,
+      "step": 1987
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7467081354736282e-05,
+      "loss": 0.6301,
+      "step": 1988
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7464319564368268e-05,
+      "loss": 0.5568,
+      "step": 1989
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7461556487743355e-05,
+      "loss": 0.5944,
+      "step": 1990
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.745879212533768e-05,
+      "loss": 0.7349,
+      "step": 1991
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.74560264776276e-05,
+      "loss": 0.7009,
+      "step": 1992
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7453259545089695e-05,
+      "loss": 0.6198,
+      "step": 1993
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7450491328200758e-05,
+      "loss": 0.8243,
+      "step": 1994
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.744772182743782e-05,
+      "loss": 0.4192,
+      "step": 1995
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7444951043278114e-05,
+      "loss": 0.5095,
+      "step": 1996
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7442178976199112e-05,
+      "loss": 0.5803,
+      "step": 1997
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7439405626678496e-05,
+      "loss": 0.7254,
+      "step": 1998
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.743663099519417e-05,
+      "loss": 0.6141,
+      "step": 1999
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7433855082224256e-05,
+      "loss": 0.8312,
+      "step": 2000
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7431077888247108e-05,
+      "loss": 0.5331,
+      "step": 2001
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7428299413741294e-05,
+      "loss": 0.7238,
+      "step": 2002
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7425519659185596e-05,
+      "loss": 0.5329,
+      "step": 2003
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7422738625059035e-05,
+      "loss": 0.572,
+      "step": 2004
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.741995631184083e-05,
+      "loss": 0.7405,
+      "step": 2005
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7417172720010436e-05,
+      "loss": 1.0834,
+      "step": 2006
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7414387850047522e-05,
+      "loss": 0.6482,
+      "step": 2007
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7411601702431976e-05,
+      "loss": 0.6333,
+      "step": 2008
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.740881427764392e-05,
+      "loss": 0.724,
+      "step": 2009
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7406025576163673e-05,
+      "loss": 0.8309,
+      "step": 2010
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.740323559847179e-05,
+      "loss": 0.5736,
+      "step": 2011
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7400444345049043e-05,
+      "loss": 0.5192,
+      "step": 2012
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7397651816376423e-05,
+      "loss": 0.5376,
+      "step": 2013
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7394858012935135e-05,
+      "loss": 0.5252,
+      "step": 2014
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.739206293520662e-05,
+      "loss": 0.7366,
+      "step": 2015
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.738926658367252e-05,
+      "loss": 0.5549,
+      "step": 2016
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7386468958814706e-05,
+      "loss": 0.7096,
+      "step": 2017
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7383670061115266e-05,
+      "loss": 0.6289,
+      "step": 2018
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.738086989105651e-05,
+      "loss": 0.6948,
+      "step": 2019
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7378068449120963e-05,
+      "loss": 0.6511,
+      "step": 2020
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7375265735791372e-05,
+      "loss": 0.7579,
+      "step": 2021
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7372461751550704e-05,
+      "loss": 0.579,
+      "step": 2022
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.736965649688214e-05,
+      "loss": 1.0343,
+      "step": 2023
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.736684997226909e-05,
+      "loss": 0.6928,
+      "step": 2024
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7364042178195176e-05,
+      "loss": 0.5496,
+      "step": 2025
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.736123311514423e-05,
+      "loss": 0.3867,
+      "step": 2026
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.735842278360032e-05,
+      "loss": 0.7713,
+      "step": 2027
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7355611184047717e-05,
+      "loss": 0.9361,
+      "step": 2028
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.735279831697093e-05,
+      "loss": 0.4062,
+      "step": 2029
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.734998418285466e-05,
+      "loss": 0.6575,
+      "step": 2030
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7347168782183852e-05,
+      "loss": 0.701,
+      "step": 2031
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7344352115443653e-05,
+      "loss": 0.7201,
+      "step": 2032
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.734153418311943e-05,
+      "loss": 0.4181,
+      "step": 2033
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.733871498569678e-05,
+      "loss": 0.7156,
+      "step": 2034
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.73358945236615e-05,
+      "loss": 0.8928,
+      "step": 2035
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7333072797499614e-05,
+      "loss": 1.0604,
+      "step": 2036
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7330249807697374e-05,
+      "loss": 0.8557,
+      "step": 2037
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7327425554741228e-05,
+      "loss": 0.7371,
+      "step": 2038
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7324600039117862e-05,
+      "loss": 0.5849,
+      "step": 2039
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7321773261314167e-05,
+      "loss": 0.4011,
+      "step": 2040
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7318945221817255e-05,
+      "loss": 0.6845,
+      "step": 2041
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.731611592111446e-05,
+      "loss": 0.9577,
+      "step": 2042
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7313285359693322e-05,
+      "loss": 0.5278,
+      "step": 2043
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7310453538041607e-05,
+      "loss": 0.5656,
+      "step": 2044
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.73076204566473e-05,
+      "loss": 0.5799,
+      "step": 2045
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.73047861159986e-05,
+      "loss": 0.761,
+      "step": 2046
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7301950516583922e-05,
+      "loss": 0.3893,
+      "step": 2047
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7299113658891894e-05,
+      "loss": 0.5464,
+      "step": 2048
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.729627554341137e-05,
+      "loss": 0.7193,
+      "step": 2049
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7293436170631414e-05,
+      "loss": 0.5443,
+      "step": 2050
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7290595541041312e-05,
+      "loss": 0.5415,
+      "step": 2051
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7287753655130557e-05,
+      "loss": 0.5914,
+      "step": 2052
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.728491051338887e-05,
+      "loss": 0.7633,
+      "step": 2053
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7282066116306184e-05,
+      "loss": 0.7815,
+      "step": 2054
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7279220464372644e-05,
+      "loss": 0.881,
+      "step": 2055
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7276373558078617e-05,
+      "loss": 0.751,
+      "step": 2056
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7273525397914677e-05,
+      "loss": 0.6074,
+      "step": 2057
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7270675984371634e-05,
+      "loss": 0.6564,
+      "step": 2058
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7267825317940494e-05,
+      "loss": 0.6246,
+      "step": 2059
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7264973399112483e-05,
+      "loss": 0.4798,
+      "step": 2060
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7262120228379053e-05,
+      "loss": 0.8119,
+      "step": 2061
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7259265806231855e-05,
+      "loss": 0.7011,
+      "step": 2062
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.725641013316277e-05,
+      "loss": 0.7151,
+      "step": 2063
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7253553209663895e-05,
+      "loss": 0.5506,
+      "step": 2064
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.725069503622753e-05,
+      "loss": 0.887,
+      "step": 2065
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7247835613346197e-05,
+      "loss": 0.7727,
+      "step": 2066
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.724497494151264e-05,
+      "loss": 0.6127,
+      "step": 2067
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7242113021219803e-05,
+      "loss": 0.7524,
+      "step": 2068
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7239249852960863e-05,
+      "loss": 0.5769,
+      "step": 2069
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.72363854372292e-05,
+      "loss": 0.6542,
+      "step": 2070
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.723351977451841e-05,
+      "loss": 0.8598,
+      "step": 2071
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.723065286532231e-05,
+      "loss": 0.9653,
+      "step": 2072
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7227784710134924e-05,
+      "loss": 0.6397,
+      "step": 2073
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7224915309450497e-05,
+      "loss": 0.751,
+      "step": 2074
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7222044663763484e-05,
+      "loss": 0.5674,
+      "step": 2075
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7219172773568556e-05,
+      "loss": 0.6632,
+      "step": 2076
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7216299639360605e-05,
+      "loss": 0.6207,
+      "step": 2077
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7213425261634727e-05,
+      "loss": 0.8495,
+      "step": 2078
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7210549640886237e-05,
+      "loss": 0.5419,
+      "step": 2079
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7207672777610667e-05,
+      "loss": 0.4567,
+      "step": 2080
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7204794672303757e-05,
+      "loss": 0.5998,
+      "step": 2081
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7201915325461463e-05,
+      "loss": 0.5006,
+      "step": 2082
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7199034737579962e-05,
+      "loss": 1.0038,
+      "step": 2083
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.719615290915563e-05,
+      "loss": 0.7045,
+      "step": 2084
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7193269840685074e-05,
+      "loss": 0.7068,
+      "step": 2085
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7190385532665105e-05,
+      "loss": 0.6353,
+      "step": 2086
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7187499985592743e-05,
+      "loss": 0.7026,
+      "step": 2087
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.718461319996523e-05,
+      "loss": 1.1769,
+      "step": 2088
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7181725176280028e-05,
+      "loss": 0.6697,
+      "step": 2089
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7178835915034793e-05,
+      "loss": 0.7797,
+      "step": 2090
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7175945416727405e-05,
+      "loss": 0.6555,
+      "step": 2091
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7173053681855964e-05,
+      "loss": 0.6432,
+      "step": 2092
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.717016071091877e-05,
+      "loss": 0.6336,
+      "step": 2093
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7167266504414342e-05,
+      "loss": 0.5817,
+      "step": 2094
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.716437106284141e-05,
+      "loss": 0.7654,
+      "step": 2095
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7161474386698927e-05,
+      "loss": 0.4947,
+      "step": 2096
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.715857647648604e-05,
+      "loss": 0.5137,
+      "step": 2097
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7155677332702127e-05,
+      "loss": 0.6804,
+      "step": 2098
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7152776955846768e-05,
+      "loss": 0.9247,
+      "step": 2099
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.714987534641975e-05,
+      "loss": 0.4838,
+      "step": 2100
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7146972504921097e-05,
+      "loss": 0.57,
+      "step": 2101
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7144068431851013e-05,
+      "loss": 0.8266,
+      "step": 2102
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.714116312770994e-05,
+      "loss": 0.8076,
+      "step": 2103
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7138256592998515e-05,
+      "loss": 0.5691,
+      "step": 2104
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.71353488282176e-05,
+      "loss": 0.3072,
+      "step": 2105
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.713243983386826e-05,
+      "loss": 0.6802,
+      "step": 2106
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7129529610451775e-05,
+      "loss": 0.7643,
+      "step": 2107
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7126618158469638e-05,
+      "loss": 0.6146,
+      "step": 2108
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7123705478423552e-05,
+      "loss": 0.7563,
+      "step": 2109
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.712079157081543e-05,
+      "loss": 0.4259,
+      "step": 2110
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7117876436147404e-05,
+      "loss": 0.5126,
+      "step": 2111
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7114960074921808e-05,
+      "loss": 0.8804,
+      "step": 2112
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.711204248764119e-05,
+      "loss": 0.8621,
+      "step": 2113
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7109123674808318e-05,
+      "loss": 0.5223,
+      "step": 2114
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7106203636926154e-05,
+      "loss": 0.5085,
+      "step": 2115
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7103282374497884e-05,
+      "loss": 0.6458,
+      "step": 2116
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.710035988802691e-05,
+      "loss": 0.5616,
+      "step": 2117
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7097436178016827e-05,
+      "loss": 0.7769,
+      "step": 2118
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7094511244971456e-05,
+      "loss": 0.4488,
+      "step": 2119
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7091585089394824e-05,
+      "loss": 0.7504,
+      "step": 2120
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.708865771179116e-05,
+      "loss": 0.595,
+      "step": 2121
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7085729112664924e-05,
+      "loss": 0.6385,
+      "step": 2122
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7082799292520767e-05,
+      "loss": 0.6192,
+      "step": 2123
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.707986825186356e-05,
+      "loss": 0.7415,
+      "step": 2124
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7076935991198383e-05,
+      "loss": 0.8139,
+      "step": 2125
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.707400251103052e-05,
+      "loss": 0.6232,
+      "step": 2126
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.6232,
+      "step": 2127
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.706813189420896e-05,
+      "loss": 1.1056,
+      "step": 2128
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.706519475856689e-05,
+      "loss": 0.4547,
+      "step": 2129
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7062256405445392e-05,
+      "loss": 0.6244,
+      "step": 2130
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7059316835350806e-05,
+      "loss": 0.6693,
+      "step": 2131
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7056376048789687e-05,
+      "loss": 0.7155,
+      "step": 2132
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7053434046268793e-05,
+      "loss": 0.5739,
+      "step": 2133
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7050490828295082e-05,
+      "loss": 0.9325,
+      "step": 2134
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.704754639537574e-05,
+      "loss": 0.5815,
+      "step": 2135
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7044600748018153e-05,
+      "loss": 0.7064,
+      "step": 2136
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.704165388672992e-05,
+      "loss": 0.7574,
+      "step": 2137
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7038705812018835e-05,
+      "loss": 0.6481,
+      "step": 2138
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7035756524392924e-05,
+      "loss": 0.6124,
+      "step": 2139
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7032806024360404e-05,
+      "loss": 0.5949,
+      "step": 2140
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7029854312429707e-05,
+      "loss": 0.6971,
+      "step": 2141
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7026901389109478e-05,
+      "loss": 0.5823,
+      "step": 2142
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7023947254908565e-05,
+      "loss": 0.6207,
+      "step": 2143
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.702099191033603e-05,
+      "loss": 0.5844,
+      "step": 2144
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.701803535590113e-05,
+      "loss": 0.6202,
+      "step": 2145
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.701507759211335e-05,
+      "loss": 0.8872,
+      "step": 2146
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7012118619482376e-05,
+      "loss": 0.7031,
+      "step": 2147
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7009158438518093e-05,
+      "loss": 0.7852,
+      "step": 2148
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.70061970497306e-05,
+      "loss": 0.7313,
+      "step": 2149
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7003234453630212e-05,
+      "loss": 1.0372,
+      "step": 2150
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7000270650727446e-05,
+      "loss": 0.9611,
+      "step": 2151
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.699730564153302e-05,
+      "loss": 0.697,
+      "step": 2152
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6994339426557872e-05,
+      "loss": 0.498,
+      "step": 2153
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.699137200631314e-05,
+      "loss": 0.4396,
+      "step": 2154
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6988403381310177e-05,
+      "loss": 0.5242,
+      "step": 2155
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6985433552060532e-05,
+      "loss": 0.7281,
+      "step": 2156
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6982462519075968e-05,
+      "loss": 0.5724,
+      "step": 2157
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6979490282868464e-05,
+      "loss": 0.5347,
+      "step": 2158
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6976516843950187e-05,
+      "loss": 0.6817,
+      "step": 2159
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.697354220283353e-05,
+      "loss": 0.5016,
+      "step": 2160
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6970566360031077e-05,
+      "loss": 0.853,
+      "step": 2161
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6967589316055636e-05,
+      "loss": 0.7576,
+      "step": 2162
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.696461107142021e-05,
+      "loss": 0.6552,
+      "step": 2163
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6961631626638012e-05,
+      "loss": 0.7486,
+      "step": 2164
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6958650982222466e-05,
+      "loss": 0.7096,
+      "step": 2165
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.695566913868719e-05,
+      "loss": 0.6062,
+      "step": 2166
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.695268609654602e-05,
+      "loss": 0.7917,
+      "step": 2167
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6949701856313007e-05,
+      "loss": 0.6374,
+      "step": 2168
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6946716418502386e-05,
+      "loss": 0.7917,
+      "step": 2169
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.694372978362861e-05,
+      "loss": 0.5021,
+      "step": 2170
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6940741952206342e-05,
+      "loss": 0.4303,
+      "step": 2171
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6937752924750444e-05,
+      "loss": 0.6978,
+      "step": 2172
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.693476270177599e-05,
+      "loss": 0.9133,
+      "step": 2173
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.693177128379826e-05,
+      "loss": 0.6158,
+      "step": 2174
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.692877867133273e-05,
+      "loss": 0.4397,
+      "step": 2175
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.69257848648951e-05,
+      "loss": 0.9392,
+      "step": 2176
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6922789865001254e-05,
+      "loss": 0.8788,
+      "step": 2177
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6919793672167293e-05,
+      "loss": 0.5508,
+      "step": 2178
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.691679628690953e-05,
+      "loss": 0.4891,
+      "step": 2179
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.691379770974447e-05,
+      "loss": 0.6017,
+      "step": 2180
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6910797941188838e-05,
+      "loss": 0.7297,
+      "step": 2181
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.690779698175955e-05,
+      "loss": 0.6538,
+      "step": 2182
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6904794831973733e-05,
+      "loss": 0.8483,
+      "step": 2183
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6901791492348722e-05,
+      "loss": 0.6765,
+      "step": 2184
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6898786963402057e-05,
+      "loss": 0.7949,
+      "step": 2185
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6895781245651474e-05,
+      "loss": 0.5615,
+      "step": 2186
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6892774339614927e-05,
+      "loss": 0.5523,
+      "step": 2187
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6889766245810563e-05,
+      "loss": 0.5902,
+      "step": 2188
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6886756964756743e-05,
+      "loss": 0.6658,
+      "step": 2189
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6883746496972026e-05,
+      "loss": 0.6014,
+      "step": 2190
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.688073484297518e-05,
+      "loss": 0.4539,
+      "step": 2191
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6877722003285173e-05,
+      "loss": 0.5974,
+      "step": 2192
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.687470797842118e-05,
+      "loss": 0.8232,
+      "step": 2193
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6871692768902582e-05,
+      "loss": 0.6389,
+      "step": 2194
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.686867637524896e-05,
+      "loss": 0.4123,
+      "step": 2195
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6865658797980106e-05,
+      "loss": 0.7864,
+      "step": 2196
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6862640037616e-05,
+      "loss": 0.6983,
+      "step": 2197
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.685962009467685e-05,
+      "loss": 0.6205,
+      "step": 2198
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6856598969683045e-05,
+      "loss": 0.5585,
+      "step": 2199
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6853576663155188e-05,
+      "loss": 0.977,
+      "step": 2200
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6850553175614093e-05,
+      "loss": 0.7092,
+      "step": 2201
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6847528507580764e-05,
+      "loss": 0.6427,
+      "step": 2202
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6844502659576414e-05,
+      "loss": 0.7645,
+      "step": 2203
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.684147563212246e-05,
+      "loss": 0.8006,
+      "step": 2204
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6838447425740524e-05,
+      "loss": 0.4545,
+      "step": 2205
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.683541804095242e-05,
+      "loss": 0.6422,
+      "step": 2206
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6832387478280185e-05,
+      "loss": 0.6833,
+      "step": 2207
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.682935573824604e-05,
+      "loss": 0.6024,
+      "step": 2208
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.682632282137242e-05,
+      "loss": 0.8108,
+      "step": 2209
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6823288728181962e-05,
+      "loss": 0.8306,
+      "step": 2210
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6820253459197493e-05,
+      "loss": 0.5308,
+      "step": 2211
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6817217014942063e-05,
+      "loss": 0.4724,
+      "step": 2212
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6814179395938915e-05,
+      "loss": 0.523,
+      "step": 2213
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6811140602711483e-05,
+      "loss": 0.7918,
+      "step": 2214
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.680810063578342e-05,
+      "loss": 0.5056,
+      "step": 2215
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6805059495678582e-05,
+      "loss": 0.6578,
+      "step": 2216
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.680201718292101e-05,
+      "loss": 0.5981,
+      "step": 2217
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6798973698034962e-05,
+      "loss": 0.5621,
+      "step": 2218
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.679592904154489e-05,
+      "loss": 0.6553,
+      "step": 2219
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.679288321397546e-05,
+      "loss": 0.6409,
+      "step": 2220
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6789836215851525e-05,
+      "loss": 0.518,
+      "step": 2221
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6786788047698147e-05,
+      "loss": 0.5971,
+      "step": 2222
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6783738710040588e-05,
+      "loss": 0.651,
+      "step": 2223
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.678068820340431e-05,
+      "loss": 0.6228,
+      "step": 2224
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.677763652831498e-05,
+      "loss": 0.5329,
+      "step": 2225
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6774583685298468e-05,
+      "loss": 0.8154,
+      "step": 2226
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.677152967488084e-05,
+      "loss": 0.8367,
+      "step": 2227
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6768474497588362e-05,
+      "loss": 0.5969,
+      "step": 2228
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.676541815394751e-05,
+      "loss": 0.6578,
+      "step": 2229
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.676236064448495e-05,
+      "loss": 0.5955,
+      "step": 2230
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.675930196972756e-05,
+      "loss": 0.5541,
+      "step": 2231
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6756242130202407e-05,
+      "loss": 0.5708,
+      "step": 2232
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.675318112643677e-05,
+      "loss": 0.4492,
+      "step": 2233
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.675011895895812e-05,
+      "loss": 0.3908,
+      "step": 2234
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6747055628294134e-05,
+      "loss": 0.7679,
+      "step": 2235
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6743991134972684e-05,
+      "loss": 0.8472,
+      "step": 2236
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6740925479521844e-05,
+      "loss": 1.106,
+      "step": 2237
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6737858662469902e-05,
+      "loss": 0.8499,
+      "step": 2238
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6734790684345322e-05,
+      "loss": 0.8129,
+      "step": 2239
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6731721545676784e-05,
+      "loss": 0.619,
+      "step": 2240
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6728651246993165e-05,
+      "loss": 0.5734,
+      "step": 2241
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.672557978882354e-05,
+      "loss": 0.4801,
+      "step": 2242
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6722507171697184e-05,
+      "loss": 0.7815,
+      "step": 2243
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.671943339614358e-05,
+      "loss": 0.6146,
+      "step": 2244
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.671635846269239e-05,
+      "loss": 0.7084,
+      "step": 2245
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6713282371873497e-05,
+      "loss": 0.5449,
+      "step": 2246
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6710205124216973e-05,
+      "loss": 0.5568,
+      "step": 2247
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6707126720253097e-05,
+      "loss": 0.9387,
+      "step": 2248
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6704047160512333e-05,
+      "loss": 0.5495,
+      "step": 2249
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.670096644552536e-05,
+      "loss": 0.5103,
+      "step": 2250
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.669788457582304e-05,
+      "loss": 0.6465,
+      "step": 2251
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6694801551936458e-05,
+      "loss": 0.8275,
+      "step": 2252
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.669171737439687e-05,
+      "loss": 0.7638,
+      "step": 2253
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6688632043735746e-05,
+      "loss": 0.8361,
+      "step": 2254
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.668554556048476e-05,
+      "loss": 0.7085,
+      "step": 2255
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6682457925175762e-05,
+      "loss": 0.8816,
+      "step": 2256
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6679369138340833e-05,
+      "loss": 0.5022,
+      "step": 2257
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6676279200512222e-05,
+      "loss": 0.547,
+      "step": 2258
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6673188112222394e-05,
+      "loss": 0.5578,
+      "step": 2259
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6670095874004012e-05,
+      "loss": 0.9191,
+      "step": 2260
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6667002486389924e-05,
+      "loss": 0.5388,
+      "step": 2261
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.666390794991319e-05,
+      "loss": 0.6102,
+      "step": 2262
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.666081226510706e-05,
+      "loss": 0.6249,
+      "step": 2263
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.665771543250499e-05,
+      "loss": 0.7351,
+      "step": 2264
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6654617452640622e-05,
+      "loss": 0.5141,
+      "step": 2265
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6651518326047807e-05,
+      "loss": 0.6868,
+      "step": 2266
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6648418053260585e-05,
+      "loss": 0.7258,
+      "step": 2267
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6645316634813195e-05,
+      "loss": 0.5754,
+      "step": 2268
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6642214071240085e-05,
+      "loss": 0.8516,
+      "step": 2269
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6639110363075884e-05,
+      "loss": 0.443,
+      "step": 2270
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6636005510855425e-05,
+      "loss": 0.8868,
+      "step": 2271
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6632899515113744e-05,
+      "loss": 0.706,
+      "step": 2272
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.662979237638606e-05,
+      "loss": 0.7365,
+      "step": 2273
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6626684095207805e-05,
+      "loss": 0.6485,
+      "step": 2274
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6623574672114596e-05,
+      "loss": 0.6224,
+      "step": 2275
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.662046410764225e-05,
+      "loss": 0.7295,
+      "step": 2276
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.661735240232679e-05,
+      "loss": 1.0264,
+      "step": 2277
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6614239556704413e-05,
+      "loss": 0.7344,
+      "step": 2278
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.661112557131154e-05,
+      "loss": 0.7325,
+      "step": 2279
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6608010446684764e-05,
+      "loss": 0.6053,
+      "step": 2280
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.66048941833609e-05,
+      "loss": 0.6082,
+      "step": 2281
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6601776781876932e-05,
+      "loss": 0.7254,
+      "step": 2282
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6598658242770054e-05,
+      "loss": 0.7853,
+      "step": 2283
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6595538566577663e-05,
+      "loss": 0.8934,
+      "step": 2284
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6592417753837338e-05,
+      "loss": 0.9431,
+      "step": 2285
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.658929580508686e-05,
+      "loss": 0.613,
+      "step": 2286
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6586172720864206e-05,
+      "loss": 0.757,
+      "step": 2287
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6583048501707547e-05,
+      "loss": 0.4491,
+      "step": 2288
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.657992314815525e-05,
+      "loss": 0.5638,
+      "step": 2289
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6576796660745885e-05,
+      "loss": 0.4346,
+      "step": 2290
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6573669040018202e-05,
+      "loss": 0.7123,
+      "step": 2291
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6570540286511157e-05,
+      "loss": 0.556,
+      "step": 2292
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6567410400763902e-05,
+      "loss": 0.5587,
+      "step": 2293
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.656427938331578e-05,
+      "loss": 0.5327,
+      "step": 2294
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6561147234706333e-05,
+      "loss": 0.5824,
+      "step": 2295
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6558013955475288e-05,
+      "loss": 0.8265,
+      "step": 2296
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6554879546162578e-05,
+      "loss": 0.4991,
+      "step": 2297
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6551744007308325e-05,
+      "loss": 0.9995,
+      "step": 2298
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.5838,
+      "step": 2299
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6545469543136667e-05,
+      "loss": 0.5585,
+      "step": 2300
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.654233061890048e-05,
+      "loss": 0.7344,
+      "step": 2301
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.653919056728519e-05,
+      "loss": 0.5959,
+      "step": 2302
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6536049388831897e-05,
+      "loss": 0.6035,
+      "step": 2303
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6532907084081882e-05,
+      "loss": 0.7194,
+      "step": 2304
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.652976365357664e-05,
+      "loss": 0.7143,
+      "step": 2305
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.652661909785785e-05,
+      "loss": 0.6835,
+      "step": 2306
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.652347341746737e-05,
+      "loss": 0.7664,
+      "step": 2307
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.652032661294728e-05,
+      "loss": 0.6772,
+      "step": 2308
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6517178684839834e-05,
+      "loss": 0.7381,
+      "step": 2309
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6514029633687487e-05,
+      "loss": 0.8095,
+      "step": 2310
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.651087946003288e-05,
+      "loss": 0.9214,
+      "step": 2311
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6507728164418864e-05,
+      "loss": 0.6552,
+      "step": 2312
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6504575747388462e-05,
+      "loss": 0.4508,
+      "step": 2313
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6501422209484907e-05,
+      "loss": 0.7331,
+      "step": 2314
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6498267551251618e-05,
+      "loss": 0.7456,
+      "step": 2315
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.64951117732322e-05,
+      "loss": 0.6104,
+      "step": 2316
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6491954875970474e-05,
+      "loss": 0.7811,
+      "step": 2317
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.648879686001043e-05,
+      "loss": 0.7039,
+      "step": 2318
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6485637725896256e-05,
+      "loss": 0.7316,
+      "step": 2319
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.648247747417234e-05,
+      "loss": 0.412,
+      "step": 2320
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6479316105383266e-05,
+      "loss": 0.7333,
+      "step": 2321
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.647615362007379e-05,
+      "loss": 0.8418,
+      "step": 2322
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6472990018788884e-05,
+      "loss": 0.5314,
+      "step": 2323
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6469825302073695e-05,
+      "loss": 0.5983,
+      "step": 2324
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.646665947047358e-05,
+      "loss": 0.3537,
+      "step": 2325
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.646349252453406e-05,
+      "loss": 0.8369,
+      "step": 2326
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6460324464800884e-05,
+      "loss": 0.5885,
+      "step": 2327
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6457155291819963e-05,
+      "loss": 0.6234,
+      "step": 2328
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6453985006137412e-05,
+      "loss": 0.8165,
+      "step": 2329
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6450813608299537e-05,
+      "loss": 0.6736,
+      "step": 2330
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.644764109885284e-05,
+      "loss": 0.5139,
+      "step": 2331
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.644446747834401e-05,
+      "loss": 0.7337,
+      "step": 2332
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6441292747319923e-05,
+      "loss": 0.5998,
+      "step": 2333
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6438116906327654e-05,
+      "loss": 0.7321,
+      "step": 2334
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6434939955914466e-05,
+      "loss": 0.787,
+      "step": 2335
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6431761896627808e-05,
+      "loss": 0.7298,
+      "step": 2336
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6428582729015333e-05,
+      "loss": 0.7614,
+      "step": 2337
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6425402453624872e-05,
+      "loss": 0.6888,
+      "step": 2338
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.642222107100446e-05,
+      "loss": 0.5836,
+      "step": 2339
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6419038581702306e-05,
+      "loss": 0.7457,
+      "step": 2340
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6415854986266824e-05,
+      "loss": 0.6189,
+      "step": 2341
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.641267028524661e-05,
+      "loss": 0.4652,
+      "step": 2342
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6409484479190457e-05,
+      "loss": 0.7194,
+      "step": 2343
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6406297568647347e-05,
+      "loss": 0.7008,
+      "step": 2344
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6403109554166447e-05,
+      "loss": 0.5387,
+      "step": 2345
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.639992043629712e-05,
+      "loss": 0.5238,
+      "step": 2346
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6396730215588913e-05,
+      "loss": 0.7724,
+      "step": 2347
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6393538892591575e-05,
+      "loss": 0.5296,
+      "step": 2348
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6390346467855027e-05,
+      "loss": 0.6356,
+      "step": 2349
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.63871529419294e-05,
+      "loss": 0.5891,
+      "step": 2350
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6383958315365e-05,
+      "loss": 0.4928,
+      "step": 2351
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6380762588712324e-05,
+      "loss": 0.4729,
+      "step": 2352
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6377565762522066e-05,
+      "loss": 0.5658,
+      "step": 2353
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6374367837345106e-05,
+      "loss": 0.7543,
+      "step": 2354
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6371168813732514e-05,
+      "loss": 0.6568,
+      "step": 2355
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.636796869223554e-05,
+      "loss": 0.7011,
+      "step": 2356
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.636476747340564e-05,
+      "loss": 1.1477,
+      "step": 2357
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.636156515779445e-05,
+      "loss": 0.8987,
+      "step": 2358
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6358361745953787e-05,
+      "loss": 0.7906,
+      "step": 2359
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.635515723843567e-05,
+      "loss": 0.43,
+      "step": 2360
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6351951635792306e-05,
+      "loss": 0.6018,
+      "step": 2361
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6348744938576082e-05,
+      "loss": 0.7459,
+      "step": 2362
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6345537147339578e-05,
+      "loss": 0.6335,
+      "step": 2363
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6342328262635567e-05,
+      "loss": 0.5683,
+      "step": 2364
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6339118285017005e-05,
+      "loss": 0.6146,
+      "step": 2365
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6335907215037033e-05,
+      "loss": 0.543,
+      "step": 2366
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.633269505324899e-05,
+      "loss": 0.5626,
+      "step": 2367
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6329481800206396e-05,
+      "loss": 0.6484,
+      "step": 2368
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6326267456462965e-05,
+      "loss": 1.033,
+      "step": 2369
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6323052022572586e-05,
+      "loss": 0.3871,
+      "step": 2370
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6319835499089358e-05,
+      "loss": 0.5044,
+      "step": 2371
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.631661788656754e-05,
+      "loss": 0.47,
+      "step": 2372
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.63133991855616e-05,
+      "loss": 0.4979,
+      "step": 2373
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6310179396626194e-05,
+      "loss": 0.6424,
+      "step": 2374
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6306958520316144e-05,
+      "loss": 0.6494,
+      "step": 2375
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.630373655718649e-05,
+      "loss": 0.5537,
+      "step": 2376
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6300513507792426e-05,
+      "loss": 1.0583,
+      "step": 2377
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.629728937268936e-05,
+      "loss": 0.4538,
+      "step": 2378
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6294064152432878e-05,
+      "loss": 0.3839,
+      "step": 2379
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.629083784757875e-05,
+      "loss": 0.5466,
+      "step": 2380
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.628761045868293e-05,
+      "loss": 0.6317,
+      "step": 2381
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6284381986301577e-05,
+      "loss": 0.8395,
+      "step": 2382
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6281152430991012e-05,
+      "loss": 0.522,
+      "step": 2383
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6277921793307758e-05,
+      "loss": 0.6637,
+      "step": 2384
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.627469007380852e-05,
+      "loss": 0.8951,
+      "step": 2385
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6271457273050195e-05,
+      "loss": 0.7303,
+      "step": 2386
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.626822339158985e-05,
+      "loss": 0.552,
+      "step": 2387
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6264988429984764e-05,
+      "loss": 0.5698,
+      "step": 2388
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.626175238879238e-05,
+      "loss": 0.655,
+      "step": 2389
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6258515268570337e-05,
+      "loss": 0.5667,
+      "step": 2390
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6255277069876454e-05,
+      "loss": 0.5728,
+      "step": 2391
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6252037793268748e-05,
+      "loss": 1.084,
+      "step": 2392
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6248797439305403e-05,
+      "loss": 0.6767,
+      "step": 2393
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6245556008544806e-05,
+      "loss": 0.8771,
+      "step": 2394
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6242313501545522e-05,
+      "loss": 0.8457,
+      "step": 2395
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6239069918866302e-05,
+      "loss": 0.8702,
+      "step": 2396
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6235825261066078e-05,
+      "loss": 0.4001,
+      "step": 2397
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.623257952870398e-05,
+      "loss": 0.6565,
+      "step": 2398
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6229332722339308e-05,
+      "loss": 0.6879,
+      "step": 2399
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6226084842531557e-05,
+      "loss": 0.6247,
+      "step": 2400
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6222835889840402e-05,
+      "loss": 0.48,
+      "step": 2401
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6219585864825706e-05,
+      "loss": 0.6482,
+      "step": 2402
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.621633476804752e-05,
+      "loss": 0.3619,
+      "step": 2403
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6213082600066065e-05,
+      "loss": 0.7831,
+      "step": 2404
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6209829361441767e-05,
+      "loss": 0.6362,
+      "step": 2405
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6206575052735224e-05,
+      "loss": 0.6476,
+      "step": 2406
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6203319674507216e-05,
+      "loss": 0.7106,
+      "step": 2407
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6200063227318716e-05,
+      "loss": 0.9175,
+      "step": 2408
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.619680571173088e-05,
+      "loss": 0.5077,
+      "step": 2409
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.619354712830504e-05,
+      "loss": 0.6618,
+      "step": 2410
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6190287477602716e-05,
+      "loss": 0.7651,
+      "step": 2411
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6187026760185623e-05,
+      "loss": 0.5286,
+      "step": 2412
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.618376497661564e-05,
+      "loss": 0.7089,
+      "step": 2413
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6180502127454847e-05,
+      "loss": 0.7136,
+      "step": 2414
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6177238213265497e-05,
+      "loss": 0.8844,
+      "step": 2415
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.617397323461003e-05,
+      "loss": 0.7566,
+      "step": 2416
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6170707192051067e-05,
+      "loss": 0.6016,
+      "step": 2417
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6167440086151423e-05,
+      "loss": 0.5405,
+      "step": 2418
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6164171917474078e-05,
+      "loss": 0.4907,
+      "step": 2419
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6160902686582212e-05,
+      "loss": 0.6295,
+      "step": 2420
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.615763239403918e-05,
+      "loss": 0.4696,
+      "step": 2421
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.615436104040852e-05,
+      "loss": 0.3805,
+      "step": 2422
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6151088626253954e-05,
+      "loss": 0.6665,
+      "step": 2423
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6147815152139385e-05,
+      "loss": 0.3516,
+      "step": 2424
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6144540618628907e-05,
+      "loss": 0.5178,
+      "step": 2425
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.614126502628678e-05,
+      "loss": 0.6293,
+      "step": 2426
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6137988375677466e-05,
+      "loss": 0.5355,
+      "step": 2427
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6134710667365598e-05,
+      "loss": 0.7715,
+      "step": 2428
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6131431901915987e-05,
+      "loss": 0.4127,
+      "step": 2429
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6128152079893636e-05,
+      "loss": 0.7208,
+      "step": 2430
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.612487120186373e-05,
+      "loss": 0.6531,
+      "step": 2431
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.612158926839162e-05,
+      "loss": 0.5806,
+      "step": 2432
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.611830628004287e-05,
+      "loss": 0.6042,
+      "step": 2433
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.611502223738319e-05,
+      "loss": 0.6274,
+      "step": 2434
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6111737140978495e-05,
+      "loss": 0.4077,
+      "step": 2435
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6108450991394876e-05,
+      "loss": 0.5208,
+      "step": 2436
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6105163789198607e-05,
+      "loss": 0.6424,
+      "step": 2437
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6101875534956134e-05,
+      "loss": 0.584,
+      "step": 2438
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6098586229234096e-05,
+      "loss": 0.8576,
+      "step": 2439
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6095295872599306e-05,
+      "loss": 0.4709,
+      "step": 2440
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.609200446561877e-05,
+      "loss": 0.6257,
+      "step": 2441
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.608871200885965e-05,
+      "loss": 0.6965,
+      "step": 2442
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6085418502889315e-05,
+      "loss": 0.7742,
+      "step": 2443
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6082123948275305e-05,
+      "loss": 0.4766,
+      "step": 2444
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6078828345585332e-05,
+      "loss": 0.5226,
+      "step": 2445
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6075531695387304e-05,
+      "loss": 0.6512,
+      "step": 2446
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6072233998249303e-05,
+      "loss": 0.5633,
+      "step": 2447
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6068935254739582e-05,
+      "loss": 0.5118,
+      "step": 2448
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.606563546542659e-05,
+      "loss": 0.771,
+      "step": 2449
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6062334630878945e-05,
+      "loss": 0.8098,
+      "step": 2450
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6059032751665454e-05,
+      "loss": 0.6232,
+      "step": 2451
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6055729828355097e-05,
+      "loss": 0.6519,
+      "step": 2452
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6052425861517035e-05,
+      "loss": 0.8099,
+      "step": 2453
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6049120851720607e-05,
+      "loss": 0.5284,
+      "step": 2454
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.604581479953534e-05,
+      "loss": 0.6621,
+      "step": 2455
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6042507705530933e-05,
+      "loss": 0.5358,
+      "step": 2456
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.603919957027727e-05,
+      "loss": 0.771,
+      "step": 2457
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6035890394344408e-05,
+      "loss": 0.5328,
+      "step": 2458
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6032580178302585e-05,
+      "loss": 0.4478,
+      "step": 2459
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.602926892272222e-05,
+      "loss": 0.7656,
+      "step": 2460
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6025956628173916e-05,
+      "loss": 0.6427,
+      "step": 2461
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6022643295228447e-05,
+      "loss": 0.7419,
+      "step": 2462
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6019328924456764e-05,
+      "loss": 0.9312,
+      "step": 2463
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.601601351643001e-05,
+      "loss": 0.678,
+      "step": 2464
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6012697071719498e-05,
+      "loss": 0.6424,
+      "step": 2465
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.600937959089671e-05,
+      "loss": 0.4571,
+      "step": 2466
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.600606107453333e-05,
+      "loss": 0.7341,
+      "step": 2467
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6002741523201197e-05,
+      "loss": 0.8035,
+      "step": 2468
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.599942093747234e-05,
+      "loss": 0.6476,
+      "step": 2469
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.599609931791897e-05,
+      "loss": 0.6036,
+      "step": 2470
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.599277666511347e-05,
+      "loss": 0.8358,
+      "step": 2471
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5989452979628398e-05,
+      "loss": 0.5985,
+      "step": 2472
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.59861282620365e-05,
+      "loss": 0.8092,
+      "step": 2473
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.598280251291068e-05,
+      "loss": 0.7387,
+      "step": 2474
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.597947573282405e-05,
+      "loss": 0.7112,
+      "step": 2475
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5976147922349873e-05,
+      "loss": 0.5919,
+      "step": 2476
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5972819082061605e-05,
+      "loss": 0.5409,
+      "step": 2477
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5969489212532865e-05,
+      "loss": 0.521,
+      "step": 2478
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5966158314337472e-05,
+      "loss": 0.5699,
+      "step": 2479
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5962826388049402e-05,
+      "loss": 0.5916,
+      "step": 2480
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.595949343424281e-05,
+      "loss": 0.889,
+      "step": 2481
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.595615945349204e-05,
+      "loss": 0.8469,
+      "step": 2482
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5952824446371608e-05,
+      "loss": 0.9762,
+      "step": 2483
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5949488413456196e-05,
+      "loss": 0.5461,
+      "step": 2484
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5946151355320675e-05,
+      "loss": 0.9382,
+      "step": 2485
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5942813272540093e-05,
+      "loss": 0.707,
+      "step": 2486
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.593947416568967e-05,
+      "loss": 0.7075,
+      "step": 2487
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5936134035344802e-05,
+      "loss": 0.4934,
+      "step": 2488
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5932792882081062e-05,
+      "loss": 0.6123,
+      "step": 2489
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.59294507064742e-05,
+      "loss": 0.8363,
+      "step": 2490
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.592610750910014e-05,
+      "loss": 0.6199,
+      "step": 2491
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.592276329053499e-05,
+      "loss": 0.5308,
+      "step": 2492
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.591941805135502e-05,
+      "loss": 0.6986,
+      "step": 2493
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.591607179213669e-05,
+      "loss": 0.7237,
+      "step": 2494
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.591272451345663e-05,
+      "loss": 0.8717,
+      "step": 2495
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.590937621589164e-05,
+      "loss": 0.7511,
+      "step": 2496
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5906026900018708e-05,
+      "loss": 0.5466,
+      "step": 2497
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.590267656641499e-05,
+      "loss": 0.5898,
+      "step": 2498
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.589932521565781e-05,
+      "loss": 0.707,
+      "step": 2499
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5895972848324684e-05,
+      "loss": 0.6847,
+      "step": 2500
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.589261946499329e-05,
+      "loss": 0.8096,
+      "step": 2501
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5889265066241488e-05,
+      "loss": 0.6889,
+      "step": 2502
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.588590965264731e-05,
+      "loss": 0.7168,
+      "step": 2503
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5882553224788955e-05,
+      "loss": 0.5597,
+      "step": 2504
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.587919578324482e-05,
+      "loss": 0.8983,
+      "step": 2505
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5875837328593448e-05,
+      "loss": 0.4669,
+      "step": 2506
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.587247786141358e-05,
+      "loss": 0.6649,
+      "step": 2507
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5869117382284113e-05,
+      "loss": 0.5132,
+      "step": 2508
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5865755891784137e-05,
+      "loss": 0.4568,
+      "step": 2509
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5862393390492895e-05,
+      "loss": 0.6038,
+      "step": 2510
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5859029878989828e-05,
+      "loss": 0.6563,
+      "step": 2511
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.585566535785453e-05,
+      "loss": 0.6789,
+      "step": 2512
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.585229982766678e-05,
+      "loss": 0.5024,
+      "step": 2513
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.584893328900653e-05,
+      "loss": 0.4842,
+      "step": 2514
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5845565742453906e-05,
+      "loss": 0.7176,
+      "step": 2515
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.58421971885892e-05,
+      "loss": 0.5866,
+      "step": 2516
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.583882762799289e-05,
+      "loss": 0.5214,
+      "step": 2517
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5835457061245615e-05,
+      "loss": 0.4783,
+      "step": 2518
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.58320854889282e-05,
+      "loss": 0.766,
+      "step": 2519
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.582871291162163e-05,
+      "loss": 0.6004,
+      "step": 2520
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5825339329907074e-05,
+      "loss": 0.9036,
+      "step": 2521
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5821964744365873e-05,
+      "loss": 0.4549,
+      "step": 2522
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.581858915557953e-05,
+      "loss": 0.486,
+      "step": 2523
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5815212564129737e-05,
+      "loss": 0.6953,
+      "step": 2524
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.581183497059834e-05,
+      "loss": 0.4985,
+      "step": 2525
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.580845637556738e-05,
+      "loss": 0.8742,
+      "step": 2526
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.580507677961906e-05,
+      "loss": 0.7459,
+      "step": 2527
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.580169618333574e-05,
+      "loss": 0.7204,
+      "step": 2528
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5798314587299976e-05,
+      "loss": 0.5985,
+      "step": 2529
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.579493199209449e-05,
+      "loss": 0.4362,
+      "step": 2530
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5791548398302167e-05,
+      "loss": 0.6231,
+      "step": 2531
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5788163806506074e-05,
+      "loss": 0.5233,
+      "step": 2532
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.578477821728944e-05,
+      "loss": 0.9747,
+      "step": 2533
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5781391631235688e-05,
+      "loss": 0.5433,
+      "step": 2534
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.577800404892838e-05,
+      "loss": 0.8144,
+      "step": 2535
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.577461547095127e-05,
+      "loss": 0.6711,
+      "step": 2536
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5771225897888288e-05,
+      "loss": 0.5194,
+      "step": 2537
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.576783533032352e-05,
+      "loss": 0.6209,
+      "step": 2538
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5764443768841234e-05,
+      "loss": 0.7116,
+      "step": 2539
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.576105121402587e-05,
+      "loss": 0.8289,
+      "step": 2540
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.575765766646203e-05,
+      "loss": 0.4586,
+      "step": 2541
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5754263126734494e-05,
+      "loss": 0.4626,
+      "step": 2542
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5750867595428205e-05,
+      "loss": 0.7116,
+      "step": 2543
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5747471073128297e-05,
+      "loss": 0.6779,
+      "step": 2544
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.574407356042005e-05,
+      "loss": 0.5168,
+      "step": 2545
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5740675057888934e-05,
+      "loss": 0.5099,
+      "step": 2546
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5737275566120577e-05,
+      "loss": 0.6586,
+      "step": 2547
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.573387508570078e-05,
+      "loss": 0.5096,
+      "step": 2548
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.573047361721552e-05,
+      "loss": 0.6055,
+      "step": 2549
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.572707116125094e-05,
+      "loss": 0.5284,
+      "step": 2550
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.572366771839335e-05,
+      "loss": 0.6703,
+      "step": 2551
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5720263289229237e-05,
+      "loss": 0.7642,
+      "step": 2552
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5716857874345258e-05,
+      "loss": 0.6023,
+      "step": 2553
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5713451474328226e-05,
+      "loss": 0.8944,
+      "step": 2554
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5710044089765144e-05,
+      "loss": 0.4423,
+      "step": 2555
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5706635721243173e-05,
+      "loss": 0.6176,
+      "step": 2556
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5703226369349642e-05,
+      "loss": 0.7126,
+      "step": 2557
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5699816034672056e-05,
+      "loss": 0.5671,
+      "step": 2558
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5696404717798087e-05,
+      "loss": 0.4155,
+      "step": 2559
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.569299241931557e-05,
+      "loss": 0.7838,
+      "step": 2560
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5689579139812524e-05,
+      "loss": 0.6766,
+      "step": 2561
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5686164879877118e-05,
+      "loss": 0.607,
+      "step": 2562
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5682749640097708e-05,
+      "loss": 0.4737,
+      "step": 2563
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5679333421062807e-05,
+      "loss": 0.8714,
+      "step": 2564
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5675916223361095e-05,
+      "loss": 0.5051,
+      "step": 2565
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5672498047581436e-05,
+      "loss": 0.5171,
+      "step": 2566
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5669078894312848e-05,
+      "loss": 0.8202,
+      "step": 2567
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.566565876414452e-05,
+      "loss": 0.6604,
+      "step": 2568
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.566223765766581e-05,
+      "loss": 0.4016,
+      "step": 2569
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5658815575466254e-05,
+      "loss": 0.6051,
+      "step": 2570
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.565539251813554e-05,
+      "loss": 1.0025,
+      "step": 2571
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.565196848626353e-05,
+      "loss": 0.5715,
+      "step": 2572
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5648543480440264e-05,
+      "loss": 0.9066,
+      "step": 2573
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5645117501255938e-05,
+      "loss": 0.6375,
+      "step": 2574
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5641690549300913e-05,
+      "loss": 0.9811,
+      "step": 2575
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5638262625165733e-05,
+      "loss": 0.5192,
+      "step": 2576
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5634833729441093e-05,
+      "loss": 1.0706,
+      "step": 2577
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.563140386271787e-05,
+      "loss": 0.6864,
+      "step": 2578
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5627973025587093e-05,
+      "loss": 0.7844,
+      "step": 2579
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.562454121863997e-05,
+      "loss": 0.5319,
+      "step": 2580
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5621108442467877e-05,
+      "loss": 0.6928,
+      "step": 2581
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5617674697662346e-05,
+      "loss": 0.6656,
+      "step": 2582
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5614239984815084e-05,
+      "loss": 0.7117,
+      "step": 2583
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5610804304517967e-05,
+      "loss": 0.3513,
+      "step": 2584
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5607367657363028e-05,
+      "loss": 0.4943,
+      "step": 2585
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5603930043942474e-05,
+      "loss": 0.644,
+      "step": 2586
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.560049146484868e-05,
+      "loss": 0.4207,
+      "step": 2587
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.559705192067418e-05,
+      "loss": 0.6896,
+      "step": 2588
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5593611412011685e-05,
+      "loss": 1.0006,
+      "step": 2589
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.559016993945406e-05,
+      "loss": 0.7985,
+      "step": 2590
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5586727503594347e-05,
+      "loss": 0.5953,
+      "step": 2591
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5583284105025743e-05,
+      "loss": 0.5347,
+      "step": 2592
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.557983974434162e-05,
+      "loss": 0.5405,
+      "step": 2593
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5576394422135518e-05,
+      "loss": 0.788,
+      "step": 2594
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5572948139001128e-05,
+      "loss": 0.8049,
+      "step": 2595
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.556950089553232e-05,
+      "loss": 0.6851,
+      "step": 2596
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5566052692323127e-05,
+      "loss": 0.6843,
+      "step": 2597
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5562603529967743e-05,
+      "loss": 0.6049,
+      "step": 2598
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5559153409060535e-05,
+      "loss": 0.878,
+      "step": 2599
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5555702330196024e-05,
+      "loss": 0.7566,
+      "step": 2600
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5552250293968903e-05,
+      "loss": 0.7161,
+      "step": 2601
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.554879730097404e-05,
+      "loss": 0.3921,
+      "step": 2602
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5545343351806443e-05,
+      "loss": 0.5582,
+      "step": 2603
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5541888447061304e-05,
+      "loss": 0.5861,
+      "step": 2604
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.553843258733398e-05,
+      "loss": 0.5767,
+      "step": 2605
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.553497577321998e-05,
+      "loss": 0.5303,
+      "step": 2606
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5531518005314987e-05,
+      "loss": 0.4315,
+      "step": 2607
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.552805928421485e-05,
+      "loss": 0.3801,
+      "step": 2608
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.552459961051557e-05,
+      "loss": 0.3882,
+      "step": 2609
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.552113898481333e-05,
+      "loss": 0.5724,
+      "step": 2610
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.551767740770446e-05,
+      "loss": 0.3597,
+      "step": 2611
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5514214879785465e-05,
+      "loss": 0.3472,
+      "step": 2612
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.551075140165301e-05,
+      "loss": 0.3906,
+      "step": 2613
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5507286973903925e-05,
+      "loss": 0.4689,
+      "step": 2614
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.55038215971352e-05,
+      "loss": 0.5281,
+      "step": 2615
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.550035527194399e-05,
+      "loss": 0.3839,
+      "step": 2616
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.549688799892762e-05,
+      "loss": 0.3878,
+      "step": 2617
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.549341977868357e-05,
+      "loss": 0.5017,
+      "step": 2618
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5489950611809484e-05,
+      "loss": 0.5633,
+      "step": 2619
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.548648049890318e-05,
+      "loss": 0.3262,
+      "step": 2620
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5483009440562617e-05,
+      "loss": 0.4887,
+      "step": 2621
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5479537437385938e-05,
+      "loss": 0.4145,
+      "step": 2622
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5476064489971444e-05,
+      "loss": 0.3751,
+      "step": 2623
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.547259059891759e-05,
+      "loss": 0.3802,
+      "step": 2624
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5469115764823e-05,
+      "loss": 0.4929,
+      "step": 2625
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5465639988286465e-05,
+      "loss": 0.4287,
+      "step": 2626
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5462163269906928e-05,
+      "loss": 0.4323,
+      "step": 2627
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5458685610283502e-05,
+      "loss": 0.5268,
+      "step": 2628
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5455207010015456e-05,
+      "loss": 0.3845,
+      "step": 2629
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5451727469702233e-05,
+      "loss": 0.4836,
+      "step": 2630
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.544824698994342e-05,
+      "loss": 0.495,
+      "step": 2631
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5444765571338782e-05,
+      "loss": 0.5223,
+      "step": 2632
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.544128321448824e-05,
+      "loss": 0.5695,
+      "step": 2633
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5437799919991873e-05,
+      "loss": 0.3893,
+      "step": 2634
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5434315688449924e-05,
+      "loss": 0.5814,
+      "step": 2635
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5430830520462806e-05,
+      "loss": 0.4975,
+      "step": 2636
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5427344416631076e-05,
+      "loss": 0.4027,
+      "step": 2637
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.542385737755547e-05,
+      "loss": 0.515,
+      "step": 2638
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5420369403836867e-05,
+      "loss": 0.3568,
+      "step": 2639
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5416880496076328e-05,
+      "loss": 0.325,
+      "step": 2640
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.541339065487506e-05,
+      "loss": 0.4917,
+      "step": 2641
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.540989988083443e-05,
+      "loss": 0.593,
+      "step": 2642
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.4399,
+      "step": 2643
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5402915536641392e-05,
+      "loss": 0.6167,
+      "step": 2644
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.539942196769253e-05,
+      "loss": 0.3516,
+      "step": 2645
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5395927468311404e-05,
+      "loss": 0.5673,
+      "step": 2646
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5392432039100192e-05,
+      "loss": 0.5118,
+      "step": 2647
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5388935680661223e-05,
+      "loss": 0.5078,
+      "step": 2648
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5385438393596994e-05,
+      "loss": 0.5407,
+      "step": 2649
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5381940178510163e-05,
+      "loss": 0.5307,
+      "step": 2650
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5378441036003543e-05,
+      "loss": 0.4356,
+      "step": 2651
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.537494096668011e-05,
+      "loss": 0.3207,
+      "step": 2652
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5371439971142993e-05,
+      "loss": 0.3945,
+      "step": 2653
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5367938049995493e-05,
+      "loss": 0.5191,
+      "step": 2654
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5364435203841058e-05,
+      "loss": 0.3615,
+      "step": 2655
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5360931433283308e-05,
+      "loss": 0.3515,
+      "step": 2656
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5357426738926008e-05,
+      "loss": 0.5633,
+      "step": 2657
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5353921121373092e-05,
+      "loss": 0.6994,
+      "step": 2658
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.535041458122865e-05,
+      "loss": 0.4186,
+      "step": 2659
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.534690711909694e-05,
+      "loss": 0.4431,
+      "step": 2660
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5343398735582352e-05,
+      "loss": 0.307,
+      "step": 2661
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.533988943128947e-05,
+      "loss": 0.3597,
+      "step": 2662
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5336379206823013e-05,
+      "loss": 0.4738,
+      "step": 2663
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5332868062787868e-05,
+      "loss": 0.4744,
+      "step": 2664
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5329355999789074e-05,
+      "loss": 0.5469,
+      "step": 2665
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5325843018431832e-05,
+      "loss": 0.6122,
+      "step": 2666
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5322329119321508e-05,
+      "loss": 0.5144,
+      "step": 2667
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5318814303063616e-05,
+      "loss": 0.7296,
+      "step": 2668
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5315298570263827e-05,
+      "loss": 0.5936,
+      "step": 2669
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5311781921527983e-05,
+      "loss": 0.3122,
+      "step": 2670
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5308264357462068e-05,
+      "loss": 0.4757,
+      "step": 2671
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5304745878672232e-05,
+      "loss": 0.4276,
+      "step": 2672
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.530122648576479e-05,
+      "loss": 0.5841,
+      "step": 2673
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5297706179346197e-05,
+      "loss": 0.3414,
+      "step": 2674
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.529418496002308e-05,
+      "loss": 0.5583,
+      "step": 2675
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.529066282840221e-05,
+      "loss": 0.5668,
+      "step": 2676
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5287139785090534e-05,
+      "loss": 0.4696,
+      "step": 2677
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5283615830695144e-05,
+      "loss": 0.631,
+      "step": 2678
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.528009096582328e-05,
+      "loss": 0.6091,
+      "step": 2679
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.527656519108236e-05,
+      "loss": 0.3751,
+      "step": 2680
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.527303850707994e-05,
+      "loss": 0.4608,
+      "step": 2681
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5269510914423753e-05,
+      "loss": 0.4319,
+      "step": 2682
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5265982413721662e-05,
+      "loss": 0.365,
+      "step": 2683
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5262453005581713e-05,
+      "loss": 0.4458,
+      "step": 2684
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5258922690612085e-05,
+      "loss": 0.5005,
+      "step": 2685
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.525539146942113e-05,
+      "loss": 0.4497,
+      "step": 2686
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5251859342617353e-05,
+      "loss": 0.4085,
+      "step": 2687
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5248326310809404e-05,
+      "loss": 0.46,
+      "step": 2688
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.524479237460611e-05,
+      "loss": 0.466,
+      "step": 2689
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5241257534616432e-05,
+      "loss": 0.5825,
+      "step": 2690
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5237721791449497e-05,
+      "loss": 0.4091,
+      "step": 2691
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5234185145714593e-05,
+      "loss": 0.333,
+      "step": 2692
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5230647598021153e-05,
+      "loss": 0.5372,
+      "step": 2693
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5227109148978765e-05,
+      "loss": 0.3374,
+      "step": 2694
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5223569799197185e-05,
+      "loss": 0.4579,
+      "step": 2695
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5220029549286315e-05,
+      "loss": 0.6216,
+      "step": 2696
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5216488399856207e-05,
+      "loss": 0.5634,
+      "step": 2697
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5212946351517081e-05,
+      "loss": 0.5918,
+      "step": 2698
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5209403404879305e-05,
+      "loss": 0.5757,
+      "step": 2699
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.52058595605534e-05,
+      "loss": 0.4258,
+      "step": 2700
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5202314819150044e-05,
+      "loss": 0.476,
+      "step": 2701
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5198769181280066e-05,
+      "loss": 0.4089,
+      "step": 2702
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5195222647554459e-05,
+      "loss": 0.2975,
+      "step": 2703
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5191675218584366e-05,
+      "loss": 0.4257,
+      "step": 2704
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5188126894981072e-05,
+      "loss": 0.3129,
+      "step": 2705
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5184577677356033e-05,
+      "loss": 0.4164,
+      "step": 2706
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5181027566320858e-05,
+      "loss": 0.3383,
+      "step": 2707
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5177476562487297e-05,
+      "loss": 0.4425,
+      "step": 2708
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5173924666467259e-05,
+      "loss": 0.8146,
+      "step": 2709
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5170371878872818e-05,
+      "loss": 0.4259,
+      "step": 2710
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5166818200316192e-05,
+      "loss": 0.5508,
+      "step": 2711
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5163263631409746e-05,
+      "loss": 0.4579,
+      "step": 2712
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.515970817276601e-05,
+      "loss": 0.4796,
+      "step": 2713
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.515615182499767e-05,
+      "loss": 0.5974,
+      "step": 2714
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5152594588717544e-05,
+      "loss": 0.3926,
+      "step": 2715
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5149036464538627e-05,
+      "loss": 0.3951,
+      "step": 2716
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5145477453074056e-05,
+      "loss": 0.352,
+      "step": 2717
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5141917554937125e-05,
+      "loss": 0.3808,
+      "step": 2718
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5138356770741274e-05,
+      "loss": 0.3512,
+      "step": 2719
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5134795101100102e-05,
+      "loss": 0.575,
+      "step": 2720
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5131232546627355e-05,
+      "loss": 0.3969,
+      "step": 2721
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.512766910793694e-05,
+      "loss": 0.6076,
+      "step": 2722
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5124104785642909e-05,
+      "loss": 0.6058,
+      "step": 2723
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5120539580359466e-05,
+      "loss": 0.4438,
+      "step": 2724
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5116973492700977e-05,
+      "loss": 0.5281,
+      "step": 2725
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5113406523281943e-05,
+      "loss": 0.5125,
+      "step": 2726
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5109838672717034e-05,
+      "loss": 0.4792,
+      "step": 2727
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5106269941621064e-05,
+      "loss": 0.3712,
+      "step": 2728
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5102700330609e-05,
+      "loss": 0.5764,
+      "step": 2729
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5099129840295952e-05,
+      "loss": 0.3317,
+      "step": 2730
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5095558471297196e-05,
+      "loss": 0.5664,
+      "step": 2731
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5091986224228156e-05,
+      "loss": 0.7296,
+      "step": 2732
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5088413099704396e-05,
+      "loss": 0.468,
+      "step": 2733
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5084839098341647e-05,
+      "loss": 0.5953,
+      "step": 2734
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5081264220755775e-05,
+      "loss": 0.4873,
+      "step": 2735
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5077688467562817e-05,
+      "loss": 0.3376,
+      "step": 2736
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5074111839378938e-05,
+      "loss": 0.5559,
+      "step": 2737
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5070534336820473e-05,
+      "loss": 0.3836,
+      "step": 2738
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5066955960503893e-05,
+      "loss": 0.467,
+      "step": 2739
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5063376711045834e-05,
+      "loss": 0.3647,
+      "step": 2740
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.505979658906307e-05,
+      "loss": 0.502,
+      "step": 2741
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5056215595172528e-05,
+      "loss": 0.5722,
+      "step": 2742
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5052633729991296e-05,
+      "loss": 0.5175,
+      "step": 2743
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5049050994136596e-05,
+      "loss": 0.5074,
+      "step": 2744
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5045467388225805e-05,
+      "loss": 0.4788,
+      "step": 2745
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5041882912876462e-05,
+      "loss": 0.4779,
+      "step": 2746
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5038297568706244e-05,
+      "loss": 0.4965,
+      "step": 2747
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5034711356332973e-05,
+      "loss": 0.352,
+      "step": 2748
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5031124276374633e-05,
+      "loss": 0.4941,
+      "step": 2749
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5027536329449355e-05,
+      "loss": 0.337,
+      "step": 2750
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.502394751617541e-05,
+      "loss": 0.7245,
+      "step": 2751
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5020357837171229e-05,
+      "loss": 0.486,
+      "step": 2752
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5016767293055388e-05,
+      "loss": 0.3777,
+      "step": 2753
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.501317588444661e-05,
+      "loss": 0.4419,
+      "step": 2754
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5009583611963772e-05,
+      "loss": 0.7296,
+      "step": 2755
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5005990476225893e-05,
+      "loss": 0.4393,
+      "step": 2756
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5002396477852146e-05,
+      "loss": 0.3912,
+      "step": 2757
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4998801617461857e-05,
+      "loss": 0.5286,
+      "step": 2758
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4995205895674489e-05,
+      "loss": 0.3465,
+      "step": 2759
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4991609313109658e-05,
+      "loss": 0.3303,
+      "step": 2760
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4988011870387138e-05,
+      "loss": 0.4433,
+      "step": 2761
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4984413568126831e-05,
+      "loss": 0.4148,
+      "step": 2762
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4980814406948806e-05,
+      "loss": 0.486,
+      "step": 2763
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4977214387473278e-05,
+      "loss": 0.4438,
+      "step": 2764
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4973613510320595e-05,
+      "loss": 0.4144,
+      "step": 2765
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4970011776111269e-05,
+      "loss": 0.5404,
+      "step": 2766
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4966409185465949e-05,
+      "loss": 0.4508,
+      "step": 2767
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4962805739005439e-05,
+      "loss": 0.4416,
+      "step": 2768
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4959201437350688e-05,
+      "loss": 0.3667,
+      "step": 2769
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.495559628112279e-05,
+      "loss": 0.5171,
+      "step": 2770
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4951990270942991e-05,
+      "loss": 0.493,
+      "step": 2771
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4948383407432678e-05,
+      "loss": 0.809,
+      "step": 2772
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.494477569121339e-05,
+      "loss": 0.6398,
+      "step": 2773
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4941167122906808e-05,
+      "loss": 0.4008,
+      "step": 2774
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4937557703134767e-05,
+      "loss": 0.3786,
+      "step": 2775
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4933947432519244e-05,
+      "loss": 0.3889,
+      "step": 2776
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4930336311682365e-05,
+      "loss": 0.5385,
+      "step": 2777
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4926724341246397e-05,
+      "loss": 0.5497,
+      "step": 2778
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.492311152183376e-05,
+      "loss": 0.497,
+      "step": 2779
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4919497854067016e-05,
+      "loss": 0.6585,
+      "step": 2780
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4915883338568875e-05,
+      "loss": 0.3968,
+      "step": 2781
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4912267975962196e-05,
+      "loss": 0.3994,
+      "step": 2782
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4908651766869975e-05,
+      "loss": 0.4556,
+      "step": 2783
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4905034711915366e-05,
+      "loss": 0.4174,
+      "step": 2784
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4901416811721658e-05,
+      "loss": 0.4658,
+      "step": 2785
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4897798066912294e-05,
+      "loss": 0.5497,
+      "step": 2786
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4894178478110856e-05,
+      "loss": 0.4695,
+      "step": 2787
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4890558045941074e-05,
+      "loss": 0.5701,
+      "step": 2788
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4886936771026826e-05,
+      "loss": 0.3691,
+      "step": 2789
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.488331465399213e-05,
+      "loss": 0.3921,
+      "step": 2790
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4879691695461151e-05,
+      "loss": 0.4363,
+      "step": 2791
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4876067896058202e-05,
+      "loss": 0.4848,
+      "step": 2792
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.487244325640774e-05,
+      "loss": 0.5194,
+      "step": 2793
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4868817777134363e-05,
+      "loss": 0.549,
+      "step": 2794
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4865191458862816e-05,
+      "loss": 0.3968,
+      "step": 2795
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.486156430221799e-05,
+      "loss": 0.324,
+      "step": 2796
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4857936307824923e-05,
+      "loss": 0.3139,
+      "step": 2797
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4854307476308783e-05,
+      "loss": 0.3302,
+      "step": 2798
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4850677808294902e-05,
+      "loss": 0.4004,
+      "step": 2799
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4847047304408744e-05,
+      "loss": 0.3996,
+      "step": 2800
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.484341596527592e-05,
+      "loss": 0.4915,
+      "step": 2801
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4839783791522187e-05,
+      "loss": 0.4096,
+      "step": 2802
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4836150783773442e-05,
+      "loss": 0.7132,
+      "step": 2803
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4832516942655726e-05,
+      "loss": 0.3916,
+      "step": 2804
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4828882268795228e-05,
+      "loss": 0.5386,
+      "step": 2805
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4825246762818278e-05,
+      "loss": 0.5293,
+      "step": 2806
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4821610425351343e-05,
+      "loss": 0.4402,
+      "step": 2807
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4817973257021047e-05,
+      "loss": 0.5019,
+      "step": 2808
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4814335258454144e-05,
+      "loss": 0.4952,
+      "step": 2809
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4810696430277537e-05,
+      "loss": 0.3438,
+      "step": 2810
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4807056773118276e-05,
+      "loss": 0.466,
+      "step": 2811
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4803416287603549e-05,
+      "loss": 0.7865,
+      "step": 2812
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4799774974360679e-05,
+      "loss": 0.4752,
+      "step": 2813
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4796132834017147e-05,
+      "loss": 0.3689,
+      "step": 2814
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.479248986720057e-05,
+      "loss": 0.4845,
+      "step": 2815
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.47888460745387e-05,
+      "loss": 0.4573,
+      "step": 2816
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4785201456659446e-05,
+      "loss": 0.5053,
+      "step": 2817
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4781556014190845e-05,
+      "loss": 0.6259,
+      "step": 2818
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4777909747761085e-05,
+      "loss": 0.4284,
+      "step": 2819
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4774262657998493e-05,
+      "loss": 0.4824,
+      "step": 2820
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4770614745531538e-05,
+      "loss": 0.464,
+      "step": 2821
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4766966010988828e-05,
+      "loss": 0.3801,
+      "step": 2822
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4763316454999121e-05,
+      "loss": 0.5945,
+      "step": 2823
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4759666078191308e-05,
+      "loss": 0.7367,
+      "step": 2824
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4756014881194422e-05,
+      "loss": 0.5577,
+      "step": 2825
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4752362864637648e-05,
+      "loss": 0.4223,
+      "step": 2826
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4748710029150296e-05,
+      "loss": 0.3509,
+      "step": 2827
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4745056375361826e-05,
+      "loss": 0.4772,
+      "step": 2828
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4741401903901843e-05,
+      "loss": 0.5041,
+      "step": 2829
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4737746615400087e-05,
+      "loss": 0.3653,
+      "step": 2830
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4734090510486435e-05,
+      "loss": 0.4162,
+      "step": 2831
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4730433589790916e-05,
+      "loss": 0.7153,
+      "step": 2832
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4726775853943688e-05,
+      "loss": 0.5692,
+      "step": 2833
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4723117303575056e-05,
+      "loss": 0.4063,
+      "step": 2834
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4719457939315468e-05,
+      "loss": 0.4826,
+      "step": 2835
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4715797761795506e-05,
+      "loss": 0.5411,
+      "step": 2836
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4712136771645891e-05,
+      "loss": 0.4302,
+      "step": 2837
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4708474969497491e-05,
+      "loss": 0.5036,
+      "step": 2838
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4704812355981308e-05,
+      "loss": 0.517,
+      "step": 2839
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4701148931728488e-05,
+      "loss": 0.5237,
+      "step": 2840
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4697484697370317e-05,
+      "loss": 0.3783,
+      "step": 2841
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4693819653538215e-05,
+      "loss": 0.4917,
+      "step": 2842
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4690153800863743e-05,
+      "loss": 0.4941,
+      "step": 2843
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4686487139978612e-05,
+      "loss": 0.4139,
+      "step": 2844
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4682819671514654e-05,
+      "loss": 0.4364,
+      "step": 2845
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4679151396103852e-05,
+      "loss": 0.5196,
+      "step": 2846
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4675482314378331e-05,
+      "loss": 0.3198,
+      "step": 2847
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4671812426970347e-05,
+      "loss": 0.53,
+      "step": 2848
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4668141734512294e-05,
+      "loss": 0.3032,
+      "step": 2849
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4664470237636712e-05,
+      "loss": 0.3896,
+      "step": 2850
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4660797936976278e-05,
+      "loss": 0.6393,
+      "step": 2851
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.46571248331638e-05,
+      "loss": 0.3185,
+      "step": 2852
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4653450926832236e-05,
+      "loss": 0.4795,
+      "step": 2853
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4649776218614671e-05,
+      "loss": 0.3548,
+      "step": 2854
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4646100709144335e-05,
+      "loss": 0.443,
+      "step": 2855
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4642424399054594e-05,
+      "loss": 0.518,
+      "step": 2856
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4638747288978954e-05,
+      "loss": 0.6018,
+      "step": 2857
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4635069379551054e-05,
+      "loss": 0.3637,
+      "step": 2858
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4631390671404682e-05,
+      "loss": 0.6501,
+      "step": 2859
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4627711165173746e-05,
+      "loss": 0.4869,
+      "step": 2860
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4624030861492304e-05,
+      "loss": 0.5844,
+      "step": 2861
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4620349760994553e-05,
+      "loss": 0.6758,
+      "step": 2862
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.461666786431482e-05,
+      "loss": 0.5023,
+      "step": 2863
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4612985172087567e-05,
+      "loss": 0.5369,
+      "step": 2864
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4609301684947403e-05,
+      "loss": 0.3501,
+      "step": 2865
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4605617403529076e-05,
+      "loss": 0.445,
+      "step": 2866
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.460193232846745e-05,
+      "loss": 0.437,
+      "step": 2867
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4598246460397548e-05,
+      "loss": 0.3592,
+      "step": 2868
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4594559799954523e-05,
+      "loss": 0.3906,
+      "step": 2869
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4590872347773657e-05,
+      "loss": 0.4301,
+      "step": 2870
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4587184104490377e-05,
+      "loss": 0.6342,
+      "step": 2871
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4583495070740248e-05,
+      "loss": 0.5961,
+      "step": 2872
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.457980524715896e-05,
+      "loss": 0.4334,
+      "step": 2873
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.457611463438235e-05,
+      "loss": 0.2637,
+      "step": 2874
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4572423233046386e-05,
+      "loss": 0.4458,
+      "step": 2875
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4568731043787171e-05,
+      "loss": 0.4387,
+      "step": 2876
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4565038067240956e-05,
+      "loss": 0.5618,
+      "step": 2877
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4561344304044102e-05,
+      "loss": 0.6085,
+      "step": 2878
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.455764975483313e-05,
+      "loss": 0.5215,
+      "step": 2879
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4553954420244688e-05,
+      "loss": 0.5699,
+      "step": 2880
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4550258300915551e-05,
+      "loss": 0.3023,
+      "step": 2881
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4546561397482646e-05,
+      "loss": 0.6082,
+      "step": 2882
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4542863710583022e-05,
+      "loss": 0.7194,
+      "step": 2883
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4539165240853865e-05,
+      "loss": 0.4877,
+      "step": 2884
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4535465988932502e-05,
+      "loss": 0.6326,
+      "step": 2885
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.453176595545639e-05,
+      "loss": 0.7624,
+      "step": 2886
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4528065141063119e-05,
+      "loss": 0.6171,
+      "step": 2887
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4524363546390417e-05,
+      "loss": 0.3057,
+      "step": 2888
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4520661172076147e-05,
+      "loss": 0.3948,
+      "step": 2889
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4516958018758302e-05,
+      "loss": 0.42,
+      "step": 2890
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4513254087075015e-05,
+      "loss": 0.4394,
+      "step": 2891
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4509549377664549e-05,
+      "loss": 0.4782,
+      "step": 2892
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4505843891165304e-05,
+      "loss": 0.5781,
+      "step": 2893
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4502137628215805e-05,
+      "loss": 0.5391,
+      "step": 2894
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4498430589454726e-05,
+      "loss": 0.4084,
+      "step": 2895
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4494722775520861e-05,
+      "loss": 0.4364,
+      "step": 2896
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.449101418705315e-05,
+      "loss": 0.4363,
+      "step": 2897
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4487304824690656e-05,
+      "loss": 0.4503,
+      "step": 2898
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4483594689072571e-05,
+      "loss": 0.5064,
+      "step": 2899
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4479883780838243e-05,
+      "loss": 0.484,
+      "step": 2900
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4476172100627127e-05,
+      "loss": 0.4711,
+      "step": 2901
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4472459649078827e-05,
+      "loss": 0.4027,
+      "step": 2902
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4468746426833075e-05,
+      "loss": 0.4698,
+      "step": 2903
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4465032434529737e-05,
+      "loss": 0.5644,
+      "step": 2904
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4461317672808809e-05,
+      "loss": 0.55,
+      "step": 2905
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4457602142310421e-05,
+      "loss": 0.7021,
+      "step": 2906
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4453885843674837e-05,
+      "loss": 0.4446,
+      "step": 2907
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4450168777542452e-05,
+      "loss": 0.5293,
+      "step": 2908
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4446450944553793e-05,
+      "loss": 0.6073,
+      "step": 2909
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4442732345349521e-05,
+      "loss": 0.4808,
+      "step": 2910
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4439012980570425e-05,
+      "loss": 0.5764,
+      "step": 2911
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4435292850857433e-05,
+      "loss": 0.574,
+      "step": 2912
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.44315719568516e-05,
+      "loss": 0.5718,
+      "step": 2913
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4427850299194108e-05,
+      "loss": 0.3932,
+      "step": 2914
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4424127878526278e-05,
+      "loss": 0.4777,
+      "step": 2915
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4420404695489565e-05,
+      "loss": 0.4551,
+      "step": 2916
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4416680750725542e-05,
+      "loss": 0.5016,
+      "step": 2917
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4412956044875928e-05,
+      "loss": 0.4361,
+      "step": 2918
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4409230578582566e-05,
+      "loss": 0.5889,
+      "step": 2919
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4405504352487429e-05,
+      "loss": 0.5206,
+      "step": 2920
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4401777367232625e-05,
+      "loss": 0.4531,
+      "step": 2921
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.439804962346039e-05,
+      "loss": 0.5,
+      "step": 2922
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4394321121813093e-05,
+      "loss": 0.3949,
+      "step": 2923
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4390591862933229e-05,
+      "loss": 0.4166,
+      "step": 2924
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4386861847463429e-05,
+      "loss": 0.4887,
+      "step": 2925
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.438313107604645e-05,
+      "loss": 0.4148,
+      "step": 2926
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4379399549325185e-05,
+      "loss": 0.4522,
+      "step": 2927
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.437566726794265e-05,
+      "loss": 0.5578,
+      "step": 2928
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4371934232541991e-05,
+      "loss": 0.5343,
+      "step": 2929
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4368200443766496e-05,
+      "loss": 0.4078,
+      "step": 2930
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.436446590225957e-05,
+      "loss": 0.448,
+      "step": 2931
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4360730608664749e-05,
+      "loss": 0.6021,
+      "step": 2932
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4356994563625703e-05,
+      "loss": 0.6376,
+      "step": 2933
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4353257767786235e-05,
+      "loss": 0.6289,
+      "step": 2934
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4349520221790263e-05,
+      "loss": 0.4613,
+      "step": 2935
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4345781926281852e-05,
+      "loss": 0.7518,
+      "step": 2936
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4342042881905184e-05,
+      "loss": 0.5626,
+      "step": 2937
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4338303089304577e-05,
+      "loss": 0.5555,
+      "step": 2938
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.433456254912447e-05,
+      "loss": 0.5685,
+      "step": 2939
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4330821262009438e-05,
+      "loss": 0.3783,
+      "step": 2940
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4327079228604177e-05,
+      "loss": 0.4144,
+      "step": 2941
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.432333644955353e-05,
+      "loss": 0.4964,
+      "step": 2942
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4319592925502443e-05,
+      "loss": 0.4017,
+      "step": 2943
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4315848657096006e-05,
+      "loss": 0.4864,
+      "step": 2944
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4312103644979441e-05,
+      "loss": 0.6003,
+      "step": 2945
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.430835788979808e-05,
+      "loss": 0.5582,
+      "step": 2946
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4304611392197399e-05,
+      "loss": 0.5,
+      "step": 2947
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4300864152822998e-05,
+      "loss": 0.5314,
+      "step": 2948
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4297116172320606e-05,
+      "loss": 0.3662,
+      "step": 2949
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4293367451336074e-05,
+      "loss": 0.443,
+      "step": 2950
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4289617990515383e-05,
+      "loss": 0.4677,
+      "step": 2951
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4285867790504648e-05,
+      "loss": 0.3556,
+      "step": 2952
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4282116851950103e-05,
+      "loss": 0.5855,
+      "step": 2953
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4278365175498113e-05,
+      "loss": 0.5176,
+      "step": 2954
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.427461276179517e-05,
+      "loss": 0.4042,
+      "step": 2955
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4270859611487891e-05,
+      "loss": 0.5353,
+      "step": 2956
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4267105725223023e-05,
+      "loss": 0.3908,
+      "step": 2957
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4263351103647438e-05,
+      "loss": 0.5033,
+      "step": 2958
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4259595747408133e-05,
+      "loss": 0.475,
+      "step": 2959
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.425583965715224e-05,
+      "loss": 0.6122,
+      "step": 2960
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4252082833527003e-05,
+      "loss": 0.3692,
+      "step": 2961
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4248325277179804e-05,
+      "loss": 0.475,
+      "step": 2962
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4244566988758152e-05,
+      "loss": 0.4377,
+      "step": 2963
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4240807968909674e-05,
+      "loss": 0.5494,
+      "step": 2964
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4237048218282124e-05,
+      "loss": 0.5948,
+      "step": 2965
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4233287737523389e-05,
+      "loss": 0.4454,
+      "step": 2966
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.422952652728148e-05,
+      "loss": 0.5433,
+      "step": 2967
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4225764588204526e-05,
+      "loss": 0.3904,
+      "step": 2968
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4222001920940787e-05,
+      "loss": 0.5771,
+      "step": 2969
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4218238526138653e-05,
+      "loss": 0.6617,
+      "step": 2970
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4214474404446633e-05,
+      "loss": 0.3623,
+      "step": 2971
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4210709556513363e-05,
+      "loss": 0.587,
+      "step": 2972
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4206943982987604e-05,
+      "loss": 0.5595,
+      "step": 2973
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4203177684518243e-05,
+      "loss": 0.5008,
+      "step": 2974
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.419941066175429e-05,
+      "loss": 0.4021,
+      "step": 2975
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4195642915344882e-05,
+      "loss": 0.4649,
+      "step": 2976
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4191874445939278e-05,
+      "loss": 0.4461,
+      "step": 2977
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4188105254186869e-05,
+      "loss": 0.2738,
+      "step": 2978
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4184335340737158e-05,
+      "loss": 0.4221,
+      "step": 2979
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.418056470623978e-05,
+      "loss": 0.4178,
+      "step": 2980
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.41767933513445e-05,
+      "loss": 0.3743,
+      "step": 2981
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4173021276701194e-05,
+      "loss": 0.4252,
+      "step": 2982
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4169248482959868e-05,
+      "loss": 0.4517,
+      "step": 2983
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4165474970770658e-05,
+      "loss": 0.6776,
+      "step": 2984
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4161700740783815e-05,
+      "loss": 0.5807,
+      "step": 2985
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4157925793649716e-05,
+      "loss": 0.4504,
+      "step": 2986
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.5982,
+      "step": 2987
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4150373750541887e-05,
+      "loss": 0.5961,
+      "step": 2988
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4146596655869529e-05,
+      "loss": 0.4708,
+      "step": 2989
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4142818846652664e-05,
+      "loss": 0.4906,
+      "step": 2990
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4139040323542284e-05,
+      "loss": 0.4679,
+      "step": 2991
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.413526108718951e-05,
+      "loss": 0.4802,
+      "step": 2992
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4131481138245582e-05,
+      "loss": 0.4447,
+      "step": 2993
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4127700477361864e-05,
+      "loss": 0.5527,
+      "step": 2994
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4123919105189836e-05,
+      "loss": 0.6672,
+      "step": 2995
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4120137022381118e-05,
+      "loss": 0.3921,
+      "step": 2996
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4116354229587432e-05,
+      "loss": 0.4215,
+      "step": 2997
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4112570727460632e-05,
+      "loss": 0.3966,
+      "step": 2998
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4108786516652704e-05,
+      "loss": 0.5744,
+      "step": 2999
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4105001597815733e-05,
+      "loss": 0.4829,
+      "step": 3000
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4101215971601945e-05,
+      "loss": 0.4342,
+      "step": 3001
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4097429638663682e-05,
+      "loss": 0.3647,
+      "step": 3002
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4093642599653406e-05,
+      "loss": 0.7257,
+      "step": 3003
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4089854855223703e-05,
+      "loss": 0.7125,
+      "step": 3004
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4086066406027284e-05,
+      "loss": 0.3522,
+      "step": 3005
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.408227725271697e-05,
+      "loss": 0.4049,
+      "step": 3006
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4078487395945712e-05,
+      "loss": 0.3745,
+      "step": 3007
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4074696836366587e-05,
+      "loss": 0.6683,
+      "step": 3008
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4070905574632779e-05,
+      "loss": 0.5804,
+      "step": 3009
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4067113611397606e-05,
+      "loss": 0.5542,
+      "step": 3010
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.40633209473145e-05,
+      "loss": 0.4916,
+      "step": 3011
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4059527583037017e-05,
+      "loss": 0.5177,
+      "step": 3012
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4055733519218829e-05,
+      "loss": 0.6195,
+      "step": 3013
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4051938756513737e-05,
+      "loss": 0.4766,
+      "step": 3014
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.404814329557565e-05,
+      "loss": 0.4815,
+      "step": 3015
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4044347137058612e-05,
+      "loss": 0.6907,
+      "step": 3016
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4040550281616777e-05,
+      "loss": 0.6541,
+      "step": 3017
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.403675272990442e-05,
+      "loss": 0.5416,
+      "step": 3018
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4032954482575938e-05,
+      "loss": 0.4246,
+      "step": 3019
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.402915554028585e-05,
+      "loss": 0.399,
+      "step": 3020
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4025355903688794e-05,
+      "loss": 0.7404,
+      "step": 3021
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.402155557343952e-05,
+      "loss": 0.4516,
+      "step": 3022
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.401775455019291e-05,
+      "loss": 0.3951,
+      "step": 3023
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4013952834603954e-05,
+      "loss": 0.6016,
+      "step": 3024
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.401015042732777e-05,
+      "loss": 0.4548,
+      "step": 3025
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.400634732901959e-05,
+      "loss": 0.4424,
+      "step": 3026
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4002543540334766e-05,
+      "loss": 0.5058,
+      "step": 3027
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.399873906192877e-05,
+      "loss": 0.4304,
+      "step": 3028
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3994933894457193e-05,
+      "loss": 0.3771,
+      "step": 3029
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3991128038575741e-05,
+      "loss": 0.3215,
+      "step": 3030
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.398732149494025e-05,
+      "loss": 0.4382,
+      "step": 3031
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3983514264206658e-05,
+      "loss": 0.3072,
+      "step": 3032
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3979706347031034e-05,
+      "loss": 0.4441,
+      "step": 3033
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.397589774406956e-05,
+      "loss": 0.5166,
+      "step": 3034
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3972088455978537e-05,
+      "loss": 0.4492,
+      "step": 3035
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3968278483414381e-05,
+      "loss": 0.4335,
+      "step": 3036
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3964467827033637e-05,
+      "loss": 0.3471,
+      "step": 3037
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3960656487492955e-05,
+      "loss": 0.5682,
+      "step": 3038
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3956844465449106e-05,
+      "loss": 0.5832,
+      "step": 3039
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3953031761558985e-05,
+      "loss": 0.4326,
+      "step": 3040
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3949218376479593e-05,
+      "loss": 0.5102,
+      "step": 3041
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3945404310868064e-05,
+      "loss": 0.4193,
+      "step": 3042
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3941589565381635e-05,
+      "loss": 0.4512,
+      "step": 3043
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3937774140677669e-05,
+      "loss": 0.5671,
+      "step": 3044
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3933958037413636e-05,
+      "loss": 0.3972,
+      "step": 3045
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3930141256247138e-05,
+      "loss": 0.5392,
+      "step": 3046
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3926323797835882e-05,
+      "loss": 0.4774,
+      "step": 3047
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.392250566283769e-05,
+      "loss": 0.5258,
+      "step": 3048
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3918686851910516e-05,
+      "loss": 0.5821,
+      "step": 3049
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3914867365712417e-05,
+      "loss": 0.4061,
+      "step": 3050
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.391104720490156e-05,
+      "loss": 0.557,
+      "step": 3051
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3907226370136251e-05,
+      "loss": 0.5049,
+      "step": 3052
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3903404862074891e-05,
+      "loss": 0.5786,
+      "step": 3053
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3899582681376009e-05,
+      "loss": 0.464,
+      "step": 3054
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3895759828698243e-05,
+      "loss": 0.6769,
+      "step": 3055
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3891936304700353e-05,
+      "loss": 0.4754,
+      "step": 3056
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3888112110041207e-05,
+      "loss": 0.3513,
+      "step": 3057
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3884287245379796e-05,
+      "loss": 0.5554,
+      "step": 3058
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3880461711375224e-05,
+      "loss": 0.5773,
+      "step": 3059
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3876635508686706e-05,
+      "loss": 0.4551,
+      "step": 3060
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3872808637973583e-05,
+      "loss": 0.391,
+      "step": 3061
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3868981099895295e-05,
+      "loss": 0.5486,
+      "step": 3062
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.386515289511141e-05,
+      "loss": 0.4946,
+      "step": 3063
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3861324024281615e-05,
+      "loss": 0.4246,
+      "step": 3064
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3857494488065693e-05,
+      "loss": 0.4822,
+      "step": 3065
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3853664287123551e-05,
+      "loss": 0.5262,
+      "step": 3066
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3849833422115221e-05,
+      "loss": 0.3454,
+      "step": 3067
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.384600189370084e-05,
+      "loss": 0.4984,
+      "step": 3068
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.384216970254065e-05,
+      "loss": 0.3227,
+      "step": 3069
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3838336849295026e-05,
+      "loss": 0.5199,
+      "step": 3070
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3834503334624443e-05,
+      "loss": 0.4226,
+      "step": 3071
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.38306691591895e-05,
+      "loss": 0.6119,
+      "step": 3072
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 0.5405,
+      "step": 3073
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3822998828669465e-05,
+      "loss": 0.4049,
+      "step": 3074
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3819162674906134e-05,
+      "loss": 0.5521,
+      "step": 3075
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3815325863021953e-05,
+      "loss": 0.5413,
+      "step": 3076
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3811488393678085e-05,
+      "loss": 0.5104,
+      "step": 3077
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3807650267535802e-05,
+      "loss": 0.4035,
+      "step": 3078
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.38038114852565e-05,
+      "loss": 0.3357,
+      "step": 3079
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3799972047501676e-05,
+      "loss": 0.3327,
+      "step": 3080
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.379613195493294e-05,
+      "loss": 0.5472,
+      "step": 3081
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.379229120821203e-05,
+      "loss": 0.457,
+      "step": 3082
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.378844980800078e-05,
+      "loss": 0.5035,
+      "step": 3083
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3784607754961141e-05,
+      "loss": 0.4892,
+      "step": 3084
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3780765049755184e-05,
+      "loss": 0.3543,
+      "step": 3085
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3776921693045084e-05,
+      "loss": 0.3247,
+      "step": 3086
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3773077685493128e-05,
+      "loss": 0.4779,
+      "step": 3087
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3769233027761724e-05,
+      "loss": 0.4295,
+      "step": 3088
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.376538772051338e-05,
+      "loss": 0.4446,
+      "step": 3089
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3761541764410729e-05,
+      "loss": 0.5331,
+      "step": 3090
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3757695160116502e-05,
+      "loss": 0.5113,
+      "step": 3091
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3753847908293554e-05,
+      "loss": 0.4311,
+      "step": 3092
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3750000009604843e-05,
+      "loss": 0.4545,
+      "step": 3093
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.374615146471344e-05,
+      "loss": 0.5248,
+      "step": 3094
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3742302274282532e-05,
+      "loss": 0.4447,
+      "step": 3095
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3738452438975416e-05,
+      "loss": 0.424,
+      "step": 3096
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3734601959455495e-05,
+      "loss": 0.6081,
+      "step": 3097
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3730750836386291e-05,
+      "loss": 0.6276,
+      "step": 3098
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3726899070431423e-05,
+      "loss": 0.4369,
+      "step": 3099
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3723046662254637e-05,
+      "loss": 0.5127,
+      "step": 3100
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3719193612519789e-05,
+      "loss": 0.5229,
+      "step": 3101
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3715339921890823e-05,
+      "loss": 0.4214,
+      "step": 3102
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3711485591031821e-05,
+      "loss": 0.6228,
+      "step": 3103
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3707630620606962e-05,
+      "loss": 0.5486,
+      "step": 3104
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3703775011280538e-05,
+      "loss": 0.4036,
+      "step": 3105
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.369991876371695e-05,
+      "loss": 0.3059,
+      "step": 3106
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3696061878580707e-05,
+      "loss": 0.473,
+      "step": 3107
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3692204356536434e-05,
+      "loss": 0.3602,
+      "step": 3108
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3688346198248859e-05,
+      "loss": 0.4257,
+      "step": 3109
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3684487404382823e-05,
+      "loss": 0.5017,
+      "step": 3110
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3680627975603275e-05,
+      "loss": 0.5451,
+      "step": 3111
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.367676791257528e-05,
+      "loss": 0.7489,
+      "step": 3112
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3672907215964e-05,
+      "loss": 0.403,
+      "step": 3113
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3669045886434715e-05,
+      "loss": 0.6601,
+      "step": 3114
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3665183924652817e-05,
+      "loss": 0.5643,
+      "step": 3115
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3661321331283796e-05,
+      "loss": 0.4498,
+      "step": 3116
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3657458106993257e-05,
+      "loss": 0.4932,
+      "step": 3117
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3653594252446917e-05,
+      "loss": 0.502,
+      "step": 3118
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3649729768310598e-05,
+      "loss": 0.3834,
+      "step": 3119
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3645864655250227e-05,
+      "loss": 0.5372,
+      "step": 3120
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3641998913931849e-05,
+      "loss": 0.5547,
+      "step": 3121
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3638132545021605e-05,
+      "loss": 0.3728,
+      "step": 3122
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3634265549185755e-05,
+      "loss": 0.6014,
+      "step": 3123
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3630397927090657e-05,
+      "loss": 0.5167,
+      "step": 3124
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.362652967940279e-05,
+      "loss": 0.5089,
+      "step": 3125
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3622660806788728e-05,
+      "loss": 0.5301,
+      "step": 3126
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.361879130991516e-05,
+      "loss": 0.4119,
+      "step": 3127
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.361492118944888e-05,
+      "loss": 0.3345,
+      "step": 3128
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3611050446056787e-05,
+      "loss": 0.5549,
+      "step": 3129
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3607179080405895e-05,
+      "loss": 0.4106,
+      "step": 3130
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3603307093163319e-05,
+      "loss": 0.5394,
+      "step": 3131
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3599434484996277e-05,
+      "loss": 0.4847,
+      "step": 3132
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3595561256572111e-05,
+      "loss": 0.278,
+      "step": 3133
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3591687408558248e-05,
+      "loss": 0.4935,
+      "step": 3134
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3587812941622238e-05,
+      "loss": 0.4182,
+      "step": 3135
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.358393785643173e-05,
+      "loss": 0.4117,
+      "step": 3136
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3580062153654484e-05,
+      "loss": 0.3162,
+      "step": 3137
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.357618583395836e-05,
+      "loss": 0.5996,
+      "step": 3138
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3572308898011328e-05,
+      "loss": 0.4589,
+      "step": 3139
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.356843134648147e-05,
+      "loss": 0.4914,
+      "step": 3140
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3564553180036962e-05,
+      "loss": 0.5413,
+      "step": 3141
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3560674399346096e-05,
+      "loss": 0.6338,
+      "step": 3142
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3556795005077266e-05,
+      "loss": 0.627,
+      "step": 3143
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.355291499789897e-05,
+      "loss": 0.4386,
+      "step": 3144
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3549034378479815e-05,
+      "loss": 0.8542,
+      "step": 3145
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3545153147488513e-05,
+      "loss": 0.6686,
+      "step": 3146
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3541271305593878e-05,
+      "loss": 0.3662,
+      "step": 3147
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3537388853464837e-05,
+      "loss": 0.3678,
+      "step": 3148
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3533505791770409e-05,
+      "loss": 0.4555,
+      "step": 3149
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3529622121179734e-05,
+      "loss": 0.6019,
+      "step": 3150
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3525737842362043e-05,
+      "loss": 0.4994,
+      "step": 3151
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3521852955986679e-05,
+      "loss": 0.3258,
+      "step": 3152
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3517967462723092e-05,
+      "loss": 0.483,
+      "step": 3153
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.351408136324083e-05,
+      "loss": 0.4999,
+      "step": 3154
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3510194658209547e-05,
+      "loss": 0.3485,
+      "step": 3155
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3506307348299007e-05,
+      "loss": 0.6839,
+      "step": 3156
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3502419434179068e-05,
+      "loss": 0.3981,
+      "step": 3157
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3498530916519705e-05,
+      "loss": 0.3707,
+      "step": 3158
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3494641795990986e-05,
+      "loss": 0.5243,
+      "step": 3159
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3490752073263086e-05,
+      "loss": 0.6222,
+      "step": 3160
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3486861749006286e-05,
+      "loss": 0.4703,
+      "step": 3161
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3482970823890975e-05,
+      "loss": 0.3257,
+      "step": 3162
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3479079298587634e-05,
+      "loss": 0.5109,
+      "step": 3163
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3475187173766851e-05,
+      "loss": 0.6087,
+      "step": 3164
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3471294450099327e-05,
+      "loss": 0.5186,
+      "step": 3165
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3467401128255855e-05,
+      "loss": 0.3854,
+      "step": 3166
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3463507208907332e-05,
+      "loss": 0.3977,
+      "step": 3167
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3459612692724768e-05,
+      "loss": 0.4853,
+      "step": 3168
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3455717580379266e-05,
+      "loss": 0.5361,
+      "step": 3169
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3451821872542028e-05,
+      "loss": 0.6876,
+      "step": 3170
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3447925569884374e-05,
+      "loss": 0.3443,
+      "step": 3171
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3444028673077717e-05,
+      "loss": 0.5794,
+      "step": 3172
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3440131182793567e-05,
+      "loss": 0.5267,
+      "step": 3173
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3436233099703549e-05,
+      "loss": 0.3955,
+      "step": 3174
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3432334424479377e-05,
+      "loss": 0.3732,
+      "step": 3175
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.342843515779288e-05,
+      "loss": 0.6641,
+      "step": 3176
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3424535300315978e-05,
+      "loss": 0.4593,
+      "step": 3177
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3420634852720697e-05,
+      "loss": 0.459,
+      "step": 3178
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3416733815679166e-05,
+      "loss": 0.5255,
+      "step": 3179
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3412832189863618e-05,
+      "loss": 0.7054,
+      "step": 3180
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3408929975946379e-05,
+      "loss": 0.4311,
+      "step": 3181
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3405027174599882e-05,
+      "loss": 0.4784,
+      "step": 3182
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3401123786496664e-05,
+      "loss": 0.5636,
+      "step": 3183
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3397219812309356e-05,
+      "loss": 0.5776,
+      "step": 3184
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3393315252710695e-05,
+      "loss": 0.509,
+      "step": 3185
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.338941010837352e-05,
+      "loss": 0.5683,
+      "step": 3186
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3385504379970764e-05,
+      "loss": 0.4302,
+      "step": 3187
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3381598068175465e-05,
+      "loss": 0.491,
+      "step": 3188
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3377691173660764e-05,
+      "loss": 0.7267,
+      "step": 3189
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.33737836970999e-05,
+      "loss": 0.5465,
+      "step": 3190
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3369875639166212e-05,
+      "loss": 0.3852,
+      "step": 3191
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3365967000533138e-05,
+      "loss": 0.335,
+      "step": 3192
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3362057781874215e-05,
+      "loss": 0.5639,
+      "step": 3193
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3358147983863089e-05,
+      "loss": 0.3904,
+      "step": 3194
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3354237607173494e-05,
+      "loss": 0.4697,
+      "step": 3195
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.335032665247927e-05,
+      "loss": 0.6255,
+      "step": 3196
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3346415120454354e-05,
+      "loss": 0.5112,
+      "step": 3197
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3342503011772792e-05,
+      "loss": 0.5452,
+      "step": 3198
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3338590327108709e-05,
+      "loss": 0.431,
+      "step": 3199
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.333467706713635e-05,
+      "loss": 0.5152,
+      "step": 3200
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3330763232530048e-05,
+      "loss": 0.6609,
+      "step": 3201
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3326848823964243e-05,
+      "loss": 0.5678,
+      "step": 3202
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3322933842113457e-05,
+      "loss": 0.3464,
+      "step": 3203
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3319018287652335e-05,
+      "loss": 0.3303,
+      "step": 3204
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3315102161255603e-05,
+      "loss": 0.5312,
+      "step": 3205
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3311185463598086e-05,
+      "loss": 0.4836,
+      "step": 3206
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3307268195354718e-05,
+      "loss": 0.665,
+      "step": 3207
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3303350357200526e-05,
+      "loss": 0.4941,
+      "step": 3208
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.329943194981063e-05,
+      "loss": 0.5923,
+      "step": 3209
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3295512973860259e-05,
+      "loss": 0.688,
+      "step": 3210
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3291593430024727e-05,
+      "loss": 0.4988,
+      "step": 3211
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3287673318979458e-05,
+      "loss": 0.6059,
+      "step": 3212
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3283752641399965e-05,
+      "loss": 0.506,
+      "step": 3213
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3279831397961865e-05,
+      "loss": 0.4837,
+      "step": 3214
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3275909589340865e-05,
+      "loss": 0.3972,
+      "step": 3215
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.327198721621278e-05,
+      "loss": 0.3322,
+      "step": 3216
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3268064279253512e-05,
+      "loss": 0.4237,
+      "step": 3217
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.326414077913906e-05,
+      "loss": 0.5616,
+      "step": 3218
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3260216716545534e-05,
+      "loss": 0.5,
+      "step": 3219
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3256292092149128e-05,
+      "loss": 0.4461,
+      "step": 3220
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.325236690662613e-05,
+      "loss": 0.4112,
+      "step": 3221
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3248441160652938e-05,
+      "loss": 0.5254,
+      "step": 3222
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3244514854906039e-05,
+      "loss": 0.3933,
+      "step": 3223
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3240587990062012e-05,
+      "loss": 0.4362,
+      "step": 3224
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3236660566797542e-05,
+      "loss": 0.4233,
+      "step": 3225
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3232732585789401e-05,
+      "loss": 0.4337,
+      "step": 3226
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3228804047714462e-05,
+      "loss": 0.3988,
+      "step": 3227
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.32248749532497e-05,
+      "loss": 0.476,
+      "step": 3228
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3220945303072172e-05,
+      "loss": 0.421,
+      "step": 3229
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3217015097859038e-05,
+      "loss": 0.4867,
+      "step": 3230
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3213084338287563e-05,
+      "loss": 0.3278,
+      "step": 3231
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3209153025035089e-05,
+      "loss": 0.4824,
+      "step": 3232
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3205221158779063e-05,
+      "loss": 0.5384,
+      "step": 3233
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3201288740197033e-05,
+      "loss": 0.6315,
+      "step": 3234
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.319735576996663e-05,
+      "loss": 0.5202,
+      "step": 3235
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.319342224876559e-05,
+      "loss": 0.3255,
+      "step": 3236
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3189488177271741e-05,
+      "loss": 0.4692,
+      "step": 3237
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3185553556163e-05,
+      "loss": 0.5153,
+      "step": 3238
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.318161838611739e-05,
+      "loss": 0.5423,
+      "step": 3239
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3177682667813019e-05,
+      "loss": 0.4748,
+      "step": 3240
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3173746401928092e-05,
+      "loss": 0.3874,
+      "step": 3241
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3169809589140909e-05,
+      "loss": 0.3432,
+      "step": 3242
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3165872230129869e-05,
+      "loss": 0.4251,
+      "step": 3243
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3161934325573454e-05,
+      "loss": 0.3494,
+      "step": 3244
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3157995876150252e-05,
+      "loss": 0.4689,
+      "step": 3245
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3154056882538935e-05,
+      "loss": 0.3974,
+      "step": 3246
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3150117345418277e-05,
+      "loss": 0.2981,
+      "step": 3247
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3146177265467137e-05,
+      "loss": 0.5307,
+      "step": 3248
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3142236643364481e-05,
+      "loss": 0.5299,
+      "step": 3249
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3138295479789352e-05,
+      "loss": 0.4538,
+      "step": 3250
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3134353775420895e-05,
+      "loss": 0.4233,
+      "step": 3251
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.313041153093835e-05,
+      "loss": 0.3335,
+      "step": 3252
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3126468747021044e-05,
+      "loss": 0.5014,
+      "step": 3253
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3122525424348404e-05,
+      "loss": 0.5408,
+      "step": 3254
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3118581563599946e-05,
+      "loss": 0.455,
+      "step": 3255
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3114637165455277e-05,
+      "loss": 0.6565,
+      "step": 3256
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.31106922305941e-05,
+      "loss": 0.6161,
+      "step": 3257
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3106746759696209e-05,
+      "loss": 0.4857,
+      "step": 3258
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3102800753441488e-05,
+      "loss": 0.4081,
+      "step": 3259
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3098854212509918e-05,
+      "loss": 0.445,
+      "step": 3260
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3094907137581567e-05,
+      "loss": 0.4537,
+      "step": 3261
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3090959529336601e-05,
+      "loss": 0.4975,
+      "step": 3262
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3087011388455275e-05,
+      "loss": 0.4056,
+      "step": 3263
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3083062715617933e-05,
+      "loss": 0.3622,
+      "step": 3264
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.307911351150501e-05,
+      "loss": 0.6248,
+      "step": 3265
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3075163776797047e-05,
+      "loss": 0.6368,
+      "step": 3266
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3071213512174655e-05,
+      "loss": 0.4909,
+      "step": 3267
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3067262718318548e-05,
+      "loss": 0.4009,
+      "step": 3268
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3063311395909531e-05,
+      "loss": 0.5205,
+      "step": 3269
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3059359545628501e-05,
+      "loss": 0.5129,
+      "step": 3270
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3055407168156438e-05,
+      "loss": 0.3313,
+      "step": 3271
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3051454264174424e-05,
+      "loss": 0.5913,
+      "step": 3272
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3047500834363623e-05,
+      "loss": 0.4672,
+      "step": 3273
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3043546879405293e-05,
+      "loss": 0.5096,
+      "step": 3274
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3039592399980785e-05,
+      "loss": 0.3765,
+      "step": 3275
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3035637396771534e-05,
+      "loss": 0.338,
+      "step": 3276
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3031681870459074e-05,
+      "loss": 0.4224,
+      "step": 3277
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.302772582172502e-05,
+      "loss": 0.4236,
+      "step": 3278
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3023769251251081e-05,
+      "loss": 0.3396,
+      "step": 3279
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3019812159719056e-05,
+      "loss": 0.4298,
+      "step": 3280
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.0,
+      "learning_rate": 1.301585454781084e-05,
+      "loss": 0.7381,
+      "step": 3281
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3011896416208404e-05,
+      "loss": 0.3901,
+      "step": 3282
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3007937765593818e-05,
+      "loss": 0.5679,
+      "step": 3283
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.300397859664924e-05,
+      "loss": 0.6097,
+      "step": 3284
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3000018910056922e-05,
+      "loss": 0.4819,
+      "step": 3285
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2996058706499191e-05,
+      "loss": 0.4868,
+      "step": 3286
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2992097986658476e-05,
+      "loss": 0.3989,
+      "step": 3287
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2988136751217292e-05,
+      "loss": 0.4221,
+      "step": 3288
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2984175000858241e-05,
+      "loss": 0.4584,
+      "step": 3289
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2980212736264015e-05,
+      "loss": 0.5606,
+      "step": 3290
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2976249958117395e-05,
+      "loss": 0.5336,
+      "step": 3291
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2972286667101248e-05,
+      "loss": 0.317,
+      "step": 3292
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2968322863898533e-05,
+      "loss": 0.642,
+      "step": 3293
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2964358549192294e-05,
+      "loss": 0.3921,
+      "step": 3294
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2960393723665663e-05,
+      "loss": 0.3628,
+      "step": 3295
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.295642838800187e-05,
+      "loss": 0.5007,
+      "step": 3296
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2952462542884212e-05,
+      "loss": 0.5039,
+      "step": 3297
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2948496188996094e-05,
+      "loss": 0.4303,
+      "step": 3298
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2944529327021002e-05,
+      "loss": 0.4876,
+      "step": 3299
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2940561957642503e-05,
+      "loss": 0.3188,
+      "step": 3300
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2936594081544258e-05,
+      "loss": 0.5112,
+      "step": 3301
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.293262569941002e-05,
+      "loss": 0.5774,
+      "step": 3302
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2928656811923621e-05,
+      "loss": 0.6066,
+      "step": 3303
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2924687419768976e-05,
+      "loss": 0.4366,
+      "step": 3304
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.29207175236301e-05,
+      "loss": 0.4655,
+      "step": 3305
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.291674712419109e-05,
+      "loss": 0.4802,
+      "step": 3306
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.0,
+      "learning_rate": 1.291277622213612e-05,
+      "loss": 0.5102,
+      "step": 3307
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2908804818149468e-05,
+      "loss": 0.7958,
+      "step": 3308
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2904832912915481e-05,
+      "loss": 0.4422,
+      "step": 3309
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2900860507118606e-05,
+      "loss": 0.4265,
+      "step": 3310
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.289688760144337e-05,
+      "loss": 0.2877,
+      "step": 3311
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2892914196574387e-05,
+      "loss": 0.5285,
+      "step": 3312
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2888940293196352e-05,
+      "loss": 0.5326,
+      "step": 3313
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2884965891994061e-05,
+      "loss": 0.4234,
+      "step": 3314
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2880990993652379e-05,
+      "loss": 0.5113,
+      "step": 3315
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.287701559885626e-05,
+      "loss": 0.5361,
+      "step": 3316
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2873039708290753e-05,
+      "loss": 0.3466,
+      "step": 3317
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2869063322640984e-05,
+      "loss": 0.4718,
+      "step": 3318
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2865086442592165e-05,
+      "loss": 0.5163,
+      "step": 3319
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2861109068829598e-05,
+      "loss": 0.4837,
+      "step": 3320
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2857131202038668e-05,
+      "loss": 0.6164,
+      "step": 3321
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2853152842904839e-05,
+      "loss": 0.4602,
+      "step": 3322
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2849173992113669e-05,
+      "loss": 0.55,
+      "step": 3323
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2845194650350795e-05,
+      "loss": 0.5363,
+      "step": 3324
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2841214818301935e-05,
+      "loss": 0.5854,
+      "step": 3325
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2837234496652901e-05,
+      "loss": 0.4057,
+      "step": 3326
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2833253686089588e-05,
+      "loss": 0.5955,
+      "step": 3327
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.282927238729797e-05,
+      "loss": 0.521,
+      "step": 3328
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2825290600964104e-05,
+      "loss": 0.4226,
+      "step": 3329
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2821308327774138e-05,
+      "loss": 0.509,
+      "step": 3330
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.6553,
+      "step": 3331
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2813342323570899e-05,
+      "loss": 0.4073,
+      "step": 3332
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2809358593930332e-05,
+      "loss": 0.6066,
+      "step": 3333
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2805374380179076e-05,
+      "loss": 0.3673,
+      "step": 3334
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2801389683003701e-05,
+      "loss": 0.4718,
+      "step": 3335
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2797404503090846e-05,
+      "loss": 0.5711,
+      "step": 3336
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2793418841127242e-05,
+      "loss": 0.3421,
+      "step": 3337
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.27894326977997e-05,
+      "loss": 0.4992,
+      "step": 3338
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2785446073795118e-05,
+      "loss": 0.6623,
+      "step": 3339
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.278145896980047e-05,
+      "loss": 0.3459,
+      "step": 3340
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.277747138650282e-05,
+      "loss": 0.5237,
+      "step": 3341
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2773483324589311e-05,
+      "loss": 0.802,
+      "step": 3342
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2769494784747164e-05,
+      "loss": 0.408,
+      "step": 3343
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2765505767663695e-05,
+      "loss": 0.5354,
+      "step": 3344
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2761516274026289e-05,
+      "loss": 0.5825,
+      "step": 3345
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2757526304522418e-05,
+      "loss": 0.5381,
+      "step": 3346
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2753535859839638e-05,
+      "loss": 0.3926,
+      "step": 3347
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2749544940665586e-05,
+      "loss": 0.5067,
+      "step": 3348
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2745553547687977e-05,
+      "loss": 0.5219,
+      "step": 3349
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2741561681594616e-05,
+      "loss": 0.4346,
+      "step": 3350
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2737569343073379e-05,
+      "loss": 0.3767,
+      "step": 3351
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2733576532812229e-05,
+      "loss": 0.494,
+      "step": 3352
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2729583251499213e-05,
+      "loss": 0.4717,
+      "step": 3353
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2725589499822457e-05,
+      "loss": 0.3617,
+      "step": 3354
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.272159527847016e-05,
+      "loss": 0.5471,
+      "step": 3355
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2717600588130617e-05,
+      "loss": 0.266,
+      "step": 3356
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2713605429492193e-05,
+      "loss": 0.3744,
+      "step": 3357
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2709609803243335e-05,
+      "loss": 0.3864,
+      "step": 3358
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2705613710072575e-05,
+      "loss": 0.4277,
+      "step": 3359
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2701617150668522e-05,
+      "loss": 0.535,
+      "step": 3360
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2697620125719862e-05,
+      "loss": 0.5273,
+      "step": 3361
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2693622635915372e-05,
+      "loss": 0.385,
+      "step": 3362
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2689624681943897e-05,
+      "loss": 0.4975,
+      "step": 3363
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2685626264494368e-05,
+      "loss": 0.5363,
+      "step": 3364
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2681627384255796e-05,
+      "loss": 0.4204,
+      "step": 3365
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2677628041917274e-05,
+      "loss": 0.3935,
+      "step": 3366
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2673628238167965e-05,
+      "loss": 0.5082,
+      "step": 3367
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2669627973697125e-05,
+      "loss": 0.525,
+      "step": 3368
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2665627249194078e-05,
+      "loss": 0.4706,
+      "step": 3369
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2661626065348231e-05,
+      "loss": 0.5878,
+      "step": 3370
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2657624422849077e-05,
+      "loss": 0.6504,
+      "step": 3371
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2653622322386175e-05,
+      "loss": 0.5442,
+      "step": 3372
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.264961976464917e-05,
+      "loss": 0.4967,
+      "step": 3373
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2645616750327792e-05,
+      "loss": 0.3593,
+      "step": 3374
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2641613280111837e-05,
+      "loss": 0.3492,
+      "step": 3375
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2637609354691186e-05,
+      "loss": 0.2991,
+      "step": 3376
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2633604974755807e-05,
+      "loss": 0.3882,
+      "step": 3377
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2629600140995726e-05,
+      "loss": 0.7615,
+      "step": 3378
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2625594854101066e-05,
+      "loss": 0.3983,
+      "step": 3379
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2621589114762018e-05,
+      "loss": 0.5159,
+      "step": 3380
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2617582923668855e-05,
+      "loss": 0.4698,
+      "step": 3381
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2613576281511929e-05,
+      "loss": 0.456,
+      "step": 3382
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.260956918898166e-05,
+      "loss": 0.4447,
+      "step": 3383
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2605561646768558e-05,
+      "loss": 0.3255,
+      "step": 3384
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2601553655563207e-05,
+      "loss": 0.6547,
+      "step": 3385
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2597545216056269e-05,
+      "loss": 0.3963,
+      "step": 3386
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2593536328938471e-05,
+      "loss": 0.5888,
+      "step": 3387
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.258952699490064e-05,
+      "loss": 0.4494,
+      "step": 3388
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.258551721463366e-05,
+      "loss": 0.5197,
+      "step": 3389
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2581506988828498e-05,
+      "loss": 0.4769,
+      "step": 3390
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.25774963181762e-05,
+      "loss": 0.6599,
+      "step": 3391
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2573485203367894e-05,
+      "loss": 0.6682,
+      "step": 3392
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2569473645094772e-05,
+      "loss": 0.4547,
+      "step": 3393
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2565461644048107e-05,
+      "loss": 0.5422,
+      "step": 3394
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2561449200919253e-05,
+      "loss": 0.5127,
+      "step": 3395
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2557436316399638e-05,
+      "loss": 0.5351,
+      "step": 3396
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2553422991180766e-05,
+      "loss": 0.6959,
+      "step": 3397
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2549409225954213e-05,
+      "loss": 0.5698,
+      "step": 3398
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.254539502141163e-05,
+      "loss": 0.4668,
+      "step": 3399
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2541380378244758e-05,
+      "loss": 0.3538,
+      "step": 3400
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2537365297145396e-05,
+      "loss": 0.5522,
+      "step": 3401
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2533349778805426e-05,
+      "loss": 0.5245,
+      "step": 3402
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2529333823916807e-05,
+      "loss": 0.5311,
+      "step": 3403
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2525317433171575e-05,
+      "loss": 0.5682,
+      "step": 3404
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2521300607261826e-05,
+      "loss": 0.4502,
+      "step": 3405
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2517283346879751e-05,
+      "loss": 0.3502,
+      "step": 3406
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2513265652717613e-05,
+      "loss": 0.7456,
+      "step": 3407
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2509247525467725e-05,
+      "loss": 0.4005,
+      "step": 3408
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2505228965822511e-05,
+      "loss": 0.4023,
+      "step": 3409
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2501209974474444e-05,
+      "loss": 0.4091,
+      "step": 3410
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2497190552116082e-05,
+      "loss": 0.4296,
+      "step": 3411
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2493170699440055e-05,
+      "loss": 0.4151,
+      "step": 3412
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2489150417139067e-05,
+      "loss": 0.5359,
+      "step": 3413
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2485129705905892e-05,
+      "loss": 0.6034,
+      "step": 3414
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2481108566433388e-05,
+      "loss": 0.5828,
+      "step": 3415
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2477086999414476e-05,
+      "loss": 0.4753,
+      "step": 3416
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2473065005542155e-05,
+      "loss": 0.5537,
+      "step": 3417
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2469042585509502e-05,
+      "loss": 0.5308,
+      "step": 3418
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2465019740009662e-05,
+      "loss": 0.6997,
+      "step": 3419
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.246099646973585e-05,
+      "loss": 0.4297,
+      "step": 3420
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2456972775381364e-05,
+      "loss": 0.4622,
+      "step": 3421
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2452948657639572e-05,
+      "loss": 0.5564,
+      "step": 3422
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2448924117203904e-05,
+      "loss": 0.2683,
+      "step": 3423
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.244489915476788e-05,
+      "loss": 0.4266,
+      "step": 3424
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2440873771025079e-05,
+      "loss": 0.4904,
+      "step": 3425
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2436847966669161e-05,
+      "loss": 0.4095,
+      "step": 3426
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2432821742393854e-05,
+      "loss": 0.4649,
+      "step": 3427
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2428795098892963e-05,
+      "loss": 0.4208,
+      "step": 3428
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2424768036860359e-05,
+      "loss": 0.472,
+      "step": 3429
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.242074055698999e-05,
+      "loss": 0.6499,
+      "step": 3430
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2416712659975869e-05,
+      "loss": 0.4049,
+      "step": 3431
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2412684346512092e-05,
+      "loss": 0.5285,
+      "step": 3432
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2408655617292822e-05,
+      "loss": 0.4747,
+      "step": 3433
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2404626473012287e-05,
+      "loss": 0.3733,
+      "step": 3434
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2400596914364792e-05,
+      "loss": 0.5675,
+      "step": 3435
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2396566942044724e-05,
+      "loss": 0.5954,
+      "step": 3436
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2392536556746517e-05,
+      "loss": 0.4116,
+      "step": 3437
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2388505759164696e-05,
+      "loss": 0.5533,
+      "step": 3438
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.238447454999385e-05,
+      "loss": 0.4757,
+      "step": 3439
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2380442929928644e-05,
+      "loss": 0.7623,
+      "step": 3440
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2376410899663802e-05,
+      "loss": 0.4585,
+      "step": 3441
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2372378459894132e-05,
+      "loss": 0.3128,
+      "step": 3442
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2368345611314508e-05,
+      "loss": 0.7388,
+      "step": 3443
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2364312354619868e-05,
+      "loss": 0.6566,
+      "step": 3444
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.236027869050523e-05,
+      "loss": 0.3542,
+      "step": 3445
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2356244619665674e-05,
+      "loss": 0.3651,
+      "step": 3446
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2352210142796359e-05,
+      "loss": 0.5636,
+      "step": 3447
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2348175260592503e-05,
+      "loss": 0.526,
+      "step": 3448
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2344139973749406e-05,
+      "loss": 0.526,
+      "step": 3449
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2340104282962427e-05,
+      "loss": 0.411,
+      "step": 3450
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2336068188927002e-05,
+      "loss": 0.5031,
+      "step": 3451
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.233203169233863e-05,
+      "loss": 0.5061,
+      "step": 3452
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2327994793892885e-05,
+      "loss": 0.5539,
+      "step": 3453
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2323957494285414e-05,
+      "loss": 0.4404,
+      "step": 3454
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.231991979421192e-05,
+      "loss": 0.4693,
+      "step": 3455
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2315881694368179e-05,
+      "loss": 0.599,
+      "step": 3456
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2311843195450046e-05,
+      "loss": 0.5128,
+      "step": 3457
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2307804298153441e-05,
+      "loss": 0.4963,
+      "step": 3458
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2303765003174342e-05,
+      "loss": 0.5928,
+      "step": 3459
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2299725311208807e-05,
+      "loss": 0.3204,
+      "step": 3460
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2295685222952958e-05,
+      "loss": 0.459,
+      "step": 3461
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2291644739102986e-05,
+      "loss": 0.4148,
+      "step": 3462
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.228760386035515e-05,
+      "loss": 0.5344,
+      "step": 3463
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2283562587405777e-05,
+      "loss": 0.4785,
+      "step": 3464
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2279520920951262e-05,
+      "loss": 0.3531,
+      "step": 3465
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.227547886168807e-05,
+      "loss": 0.6574,
+      "step": 3466
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2271436410312727e-05,
+      "loss": 0.4422,
+      "step": 3467
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2267393567521832e-05,
+      "loss": 0.4538,
+      "step": 3468
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2263350334012059e-05,
+      "loss": 0.4213,
+      "step": 3469
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.225930671048013e-05,
+      "loss": 0.4239,
+      "step": 3470
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2255262697622848e-05,
+      "loss": 0.4712,
+      "step": 3471
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2251218296137084e-05,
+      "loss": 0.3725,
+      "step": 3472
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2247173506719768e-05,
+      "loss": 0.3626,
+      "step": 3473
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2243128330067904e-05,
+      "loss": 0.3043,
+      "step": 3474
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2239082766878557e-05,
+      "loss": 0.3645,
+      "step": 3475
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2235036817848865e-05,
+      "loss": 0.4014,
+      "step": 3476
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2230990483676024e-05,
+      "loss": 0.4215,
+      "step": 3477
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2226943765057304e-05,
+      "loss": 0.277,
+      "step": 3478
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.222289666269004e-05,
+      "loss": 0.3308,
+      "step": 3479
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2218849177271627e-05,
+      "loss": 0.3223,
+      "step": 3480
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2214801309499538e-05,
+      "loss": 0.4327,
+      "step": 3481
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2210753060071297e-05,
+      "loss": 0.2972,
+      "step": 3482
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2206704429684504e-05,
+      "loss": 0.4833,
+      "step": 3483
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2202655419036822e-05,
+      "loss": 0.4452,
+      "step": 3484
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2198606028825978e-05,
+      "loss": 0.3277,
+      "step": 3485
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2194556259749768e-05,
+      "loss": 0.335,
+      "step": 3486
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2190506112506054e-05,
+      "loss": 0.3909,
+      "step": 3487
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2186455587792753e-05,
+      "loss": 0.4561,
+      "step": 3488
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2182404686307857e-05,
+      "loss": 0.3795,
+      "step": 3489
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2178353408749427e-05,
+      "loss": 0.3561,
+      "step": 3490
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2174301755815572e-05,
+      "loss": 0.3698,
+      "step": 3491
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2170249728204479e-05,
+      "loss": 0.3333,
+      "step": 3492
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2166197326614399e-05,
+      "loss": 0.3966,
+      "step": 3493
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2162144551743644e-05,
+      "loss": 0.3454,
+      "step": 3494
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2158091404290588e-05,
+      "loss": 0.3425,
+      "step": 3495
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2154037884953676e-05,
+      "loss": 0.3548,
+      "step": 3496
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.214998399443141e-05,
+      "loss": 0.2866,
+      "step": 3497
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2145929733422364e-05,
+      "loss": 0.3178,
+      "step": 3498
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2141875102625166e-05,
+      "loss": 0.4735,
+      "step": 3499
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2137820102738516e-05,
+      "loss": 0.3313,
+      "step": 3500
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2133764734461173e-05,
+      "loss": 0.3182,
+      "step": 3501
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2129708998491961e-05,
+      "loss": 0.4941,
+      "step": 3502
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2125652895529766e-05,
+      "loss": 0.3629,
+      "step": 3503
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2121596426273544e-05,
+      "loss": 0.3762,
+      "step": 3504
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2117539591422308e-05,
+      "loss": 0.4275,
+      "step": 3505
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2113482391675124e-05,
+      "loss": 0.4041,
+      "step": 3506
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2109424827731144e-05,
+      "loss": 0.3838,
+      "step": 3507
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2105366900289566e-05,
+      "loss": 0.5046,
+      "step": 3508
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2101308610049653e-05,
+      "loss": 0.3706,
+      "step": 3509
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2097249957710737e-05,
+      "loss": 0.3363,
+      "step": 3510
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2093190943972204e-05,
+      "loss": 0.404,
+      "step": 3511
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2089131569533506e-05,
+      "loss": 0.2632,
+      "step": 3512
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.208507183509416e-05,
+      "loss": 0.3658,
+      "step": 3513
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2081011741353739e-05,
+      "loss": 0.4035,
+      "step": 3514
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2076951289011884e-05,
+      "loss": 0.359,
+      "step": 3515
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2072890478768296e-05,
+      "loss": 0.3476,
+      "step": 3516
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2068829311322731e-05,
+      "loss": 0.4675,
+      "step": 3517
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2064767787375017e-05,
+      "loss": 0.3085,
+      "step": 3518
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2060705907625042e-05,
+      "loss": 0.4766,
+      "step": 3519
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2056643672772746e-05,
+      "loss": 0.3663,
+      "step": 3520
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2052581083518135e-05,
+      "loss": 0.3164,
+      "step": 3521
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2048518140561285e-05,
+      "loss": 0.3867,
+      "step": 3522
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.204445484460232e-05,
+      "loss": 0.3247,
+      "step": 3523
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2040391196341427e-05,
+      "loss": 0.3059,
+      "step": 3524
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2036327196478866e-05,
+      "loss": 0.3114,
+      "step": 3525
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.203226284571494e-05,
+      "loss": 0.4147,
+      "step": 3526
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2028198144750023e-05,
+      "loss": 0.3473,
+      "step": 3527
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.202413309428455e-05,
+      "loss": 0.3461,
+      "step": 3528
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.202006769501901e-05,
+      "loss": 0.2994,
+      "step": 3529
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2016001947653957e-05,
+      "loss": 0.3082,
+      "step": 3530
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2011935852890004e-05,
+      "loss": 0.3385,
+      "step": 3531
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2007869411427824e-05,
+      "loss": 0.4588,
+      "step": 3532
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2003802623968144e-05,
+      "loss": 0.3187,
+      "step": 3533
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1999735491211764e-05,
+      "loss": 0.2887,
+      "step": 3534
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.199566801385953e-05,
+      "loss": 0.2431,
+      "step": 3535
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.199160019261235e-05,
+      "loss": 0.5337,
+      "step": 3536
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.19875320281712e-05,
+      "loss": 0.3808,
+      "step": 3537
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1983463521237106e-05,
+      "loss": 0.4567,
+      "step": 3538
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1979394672511156e-05,
+      "loss": 0.3324,
+      "step": 3539
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1975325482694497e-05,
+      "loss": 0.3393,
+      "step": 3540
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.0,
+      "learning_rate": 1.197125595248834e-05,
+      "loss": 0.3966,
+      "step": 3541
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1967186082593935e-05,
+      "loss": 0.3425,
+      "step": 3542
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.196311587371262e-05,
+      "loss": 0.3052,
+      "step": 3543
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.195904532654577e-05,
+      "loss": 0.3886,
+      "step": 3544
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1954974441794827e-05,
+      "loss": 0.2927,
+      "step": 3545
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1950903220161286e-05,
+      "loss": 0.2549,
+      "step": 3546
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.19468316623467e-05,
+      "loss": 0.3371,
+      "step": 3547
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.194275976905269e-05,
+      "loss": 0.4775,
+      "step": 3548
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1938687540980925e-05,
+      "loss": 0.4609,
+      "step": 3549
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.0,
+      "learning_rate": 1.193461497883313e-05,
+      "loss": 0.2557,
+      "step": 3550
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1930542083311096e-05,
+      "loss": 0.3027,
+      "step": 3551
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1926468855116671e-05,
+      "loss": 0.4204,
+      "step": 3552
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1922395294951747e-05,
+      "loss": 0.388,
+      "step": 3553
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.191832140351829e-05,
+      "loss": 0.3279,
+      "step": 3554
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1914247181518312e-05,
+      "loss": 0.284,
+      "step": 3555
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1910172629653891e-05,
+      "loss": 0.4064,
+      "step": 3556
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.190609774862715e-05,
+      "loss": 0.3145,
+      "step": 3557
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1902022539140278e-05,
+      "loss": 0.3793,
+      "step": 3558
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1897947001895524e-05,
+      "loss": 0.3352,
+      "step": 3559
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1893871137595178e-05,
+      "loss": 0.4911,
+      "step": 3560
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1889794946941603e-05,
+      "loss": 0.3046,
+      "step": 3561
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1885718430637205e-05,
+      "loss": 0.3048,
+      "step": 3562
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1881641589384456e-05,
+      "loss": 0.333,
+      "step": 3563
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1877564423885881e-05,
+      "loss": 0.3094,
+      "step": 3564
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.187348693484406e-05,
+      "loss": 0.2993,
+      "step": 3565
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1869409122961626e-05,
+      "loss": 0.3895,
+      "step": 3566
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1865330988941271e-05,
+      "loss": 0.3804,
+      "step": 3567
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1861252533485743e-05,
+      "loss": 0.339,
+      "step": 3568
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1857173757297843e-05,
+      "loss": 0.4273,
+      "step": 3569
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1853094661080435e-05,
+      "loss": 0.4394,
+      "step": 3570
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1849015245536424e-05,
+      "loss": 0.4986,
+      "step": 3571
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.184493551136878e-05,
+      "loss": 0.3829,
+      "step": 3572
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1840855459280526e-05,
+      "loss": 0.5138,
+      "step": 3573
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1836775089974743e-05,
+      "loss": 0.3414,
+      "step": 3574
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1832694404154557e-05,
+      "loss": 0.3868,
+      "step": 3575
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1828613402523156e-05,
+      "loss": 0.3766,
+      "step": 3576
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1824532085783788e-05,
+      "loss": 0.4167,
+      "step": 3577
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.182045045463974e-05,
+      "loss": 0.4689,
+      "step": 3578
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1816368509794365e-05,
+      "loss": 0.2707,
+      "step": 3579
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1812286251951069e-05,
+      "loss": 0.3528,
+      "step": 3580
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1808203681813304e-05,
+      "loss": 0.3863,
+      "step": 3581
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1804120800084587e-05,
+      "loss": 0.2887,
+      "step": 3582
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1800037607468477e-05,
+      "loss": 0.3259,
+      "step": 3583
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.17959541046686e-05,
+      "loss": 0.428,
+      "step": 3584
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1791870292388624e-05,
+      "loss": 0.3606,
+      "step": 3585
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1787786171332272e-05,
+      "loss": 0.3951,
+      "step": 3586
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1783701742203326e-05,
+      "loss": 0.3021,
+      "step": 3587
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1779617005705622e-05,
+      "loss": 0.3587,
+      "step": 3588
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1775531962543036e-05,
+      "loss": 0.3212,
+      "step": 3589
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1771446613419508e-05,
+      "loss": 0.3685,
+      "step": 3590
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1767360959039032e-05,
+      "loss": 0.2865,
+      "step": 3591
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.176327500010565e-05,
+      "loss": 0.4395,
+      "step": 3592
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1759188737323454e-05,
+      "loss": 0.4096,
+      "step": 3593
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1755102171396596e-05,
+      "loss": 0.4129,
+      "step": 3594
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1751015303029272e-05,
+      "loss": 0.3025,
+      "step": 3595
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1746928132925737e-05,
+      "loss": 0.454,
+      "step": 3596
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1742840661790294e-05,
+      "loss": 0.5186,
+      "step": 3597
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1738752890327299e-05,
+      "loss": 0.2969,
+      "step": 3598
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1734664819241159e-05,
+      "loss": 0.3839,
+      "step": 3599
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1730576449236334e-05,
+      "loss": 0.3271,
+      "step": 3600
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1726487781017337e-05,
+      "loss": 0.2812,
+      "step": 3601
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1722398815288727e-05,
+      "loss": 0.3731,
+      "step": 3602
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1718309552755118e-05,
+      "loss": 0.2719,
+      "step": 3603
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1714219994121175e-05,
+      "loss": 0.3053,
+      "step": 3604
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1710130140091611e-05,
+      "loss": 0.3521,
+      "step": 3605
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1706039991371203e-05,
+      "loss": 0.3256,
+      "step": 3606
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1701949548664753e-05,
+      "loss": 0.3863,
+      "step": 3607
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.169785881267714e-05,
+      "loss": 0.2455,
+      "step": 3608
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1693767784113281e-05,
+      "loss": 0.3908,
+      "step": 3609
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1689676463678143e-05,
+      "loss": 0.4181,
+      "step": 3610
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1685584852076746e-05,
+      "loss": 0.3687,
+      "step": 3611
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1681492950014157e-05,
+      "loss": 0.381,
+      "step": 3612
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1677400758195501e-05,
+      "loss": 0.5357,
+      "step": 3613
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1673308277325942e-05,
+      "loss": 0.3917,
+      "step": 3614
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.16692155081107e-05,
+      "loss": 0.2356,
+      "step": 3615
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1665122451255047e-05,
+      "loss": 0.3529,
+      "step": 3616
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1661029107464298e-05,
+      "loss": 0.3552,
+      "step": 3617
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1656935477443825e-05,
+      "loss": 0.4187,
+      "step": 3618
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1652841561899042e-05,
+      "loss": 0.4553,
+      "step": 3619
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1648747361535415e-05,
+      "loss": 0.5761,
+      "step": 3620
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1644652877058464e-05,
+      "loss": 0.4767,
+      "step": 3621
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1640558109173748e-05,
+      "loss": 0.2961,
+      "step": 3622
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1636463058586882e-05,
+      "loss": 0.3599,
+      "step": 3623
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1632367726003534e-05,
+      "loss": 0.387,
+      "step": 3624
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1628272112129402e-05,
+      "loss": 0.5327,
+      "step": 3625
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1624176217670255e-05,
+      "loss": 0.4639,
+      "step": 3626
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1620080043331901e-05,
+      "loss": 0.3377,
+      "step": 3627
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.0,
+      "learning_rate": 1.161598358982019e-05,
+      "loss": 0.3626,
+      "step": 3628
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.161188685784103e-05,
+      "loss": 0.4183,
+      "step": 3629
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1607789848100372e-05,
+      "loss": 0.346,
+      "step": 3630
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1603692561304214e-05,
+      "loss": 0.419,
+      "step": 3631
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1599594998158602e-05,
+      "loss": 0.3911,
+      "step": 3632
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1595497159369638e-05,
+      "loss": 0.5136,
+      "step": 3633
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1591399045643454e-05,
+      "loss": 0.3194,
+      "step": 3634
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1587300657686254e-05,
+      "loss": 0.3708,
+      "step": 3635
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.158320199620426e-05,
+      "loss": 0.4439,
+      "step": 3636
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1579103061903765e-05,
+      "loss": 0.3206,
+      "step": 3637
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1575003855491098e-05,
+      "loss": 0.2613,
+      "step": 3638
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1570904377672642e-05,
+      "loss": 0.3332,
+      "step": 3639
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1566804629154813e-05,
+      "loss": 0.3431,
+      "step": 3640
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.156270461064409e-05,
+      "loss": 0.4315,
+      "step": 3641
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1558604322846986e-05,
+      "loss": 0.2706,
+      "step": 3642
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1554503766470069e-05,
+      "loss": 0.3758,
+      "step": 3643
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.155040294221995e-05,
+      "loss": 0.3254,
+      "step": 3644
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1546301850803283e-05,
+      "loss": 0.4647,
+      "step": 3645
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1542200492926774e-05,
+      "loss": 0.4576,
+      "step": 3646
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.153809886929717e-05,
+      "loss": 0.3868,
+      "step": 3647
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1533996980621267e-05,
+      "loss": 0.4053,
+      "step": 3648
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1529894827605902e-05,
+      "loss": 0.4475,
+      "step": 3649
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1525792410957967e-05,
+      "loss": 0.4221,
+      "step": 3650
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1521689731384391e-05,
+      "loss": 0.3071,
+      "step": 3651
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1517586789592146e-05,
+      "loss": 0.4014,
+      "step": 3652
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1513483586288262e-05,
+      "loss": 0.4649,
+      "step": 3653
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1509380122179799e-05,
+      "loss": 0.4514,
+      "step": 3654
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.150527639797387e-05,
+      "loss": 0.4028,
+      "step": 3655
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1501172414377635e-05,
+      "loss": 0.4132,
+      "step": 3656
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1497068172098295e-05,
+      "loss": 0.369,
+      "step": 3657
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1492963671843092e-05,
+      "loss": 0.3508,
+      "step": 3658
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1488858914319321e-05,
+      "loss": 0.4668,
+      "step": 3659
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1484753900234316e-05,
+      "loss": 0.3043,
+      "step": 3660
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.148064863029545e-05,
+      "loss": 0.3422,
+      "step": 3661
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1476543105210154e-05,
+      "loss": 0.379,
+      "step": 3662
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.147243732568589e-05,
+      "loss": 0.2737,
+      "step": 3663
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1468331292430173e-05,
+      "loss": 0.3314,
+      "step": 3664
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1464225006150553e-05,
+      "loss": 0.5128,
+      "step": 3665
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1460118467554633e-05,
+      "loss": 0.4384,
+      "step": 3666
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1456011677350052e-05,
+      "loss": 0.3189,
+      "step": 3667
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1451904636244493e-05,
+      "loss": 0.4026,
+      "step": 3668
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1447797344945686e-05,
+      "loss": 0.2822,
+      "step": 3669
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1443689804161405e-05,
+      "loss": 0.3275,
+      "step": 3670
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1439582014599462e-05,
+      "loss": 0.3454,
+      "step": 3671
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1435473976967712e-05,
+      "loss": 0.3414,
+      "step": 3672
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1431365691974059e-05,
+      "loss": 0.3471,
+      "step": 3673
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1427257160326444e-05,
+      "loss": 0.3513,
+      "step": 3674
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.261,
+      "step": 3675
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1419039359901311e-05,
+      "loss": 0.3972,
+      "step": 3676
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1414930092539891e-05,
+      "loss": 0.3577,
+      "step": 3677
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1410820581356705e-05,
+      "loss": 0.4416,
+      "step": 3678
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.14067108270599e-05,
+      "loss": 0.3744,
+      "step": 3679
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1402600830357679e-05,
+      "loss": 0.5528,
+      "step": 3680
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.139849059195828e-05,
+      "loss": 0.3891,
+      "step": 3681
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1394380112569978e-05,
+      "loss": 0.3905,
+      "step": 3682
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1390269392901096e-05,
+      "loss": 0.4457,
+      "step": 3683
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1386158433659994e-05,
+      "loss": 0.3226,
+      "step": 3684
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1382047235555078e-05,
+      "loss": 0.3035,
+      "step": 3685
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1377935799294794e-05,
+      "loss": 0.355,
+      "step": 3686
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1373824125587625e-05,
+      "loss": 0.3682,
+      "step": 3687
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1369712215142095e-05,
+      "loss": 0.4245,
+      "step": 3688
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1365600068666781e-05,
+      "loss": 0.3426,
+      "step": 3689
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1361487686870278e-05,
+      "loss": 0.3457,
+      "step": 3690
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1357375070461241e-05,
+      "loss": 0.3087,
+      "step": 3691
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.135326222014836e-05,
+      "loss": 0.2343,
+      "step": 3692
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1349149136640367e-05,
+      "loss": 0.3737,
+      "step": 3693
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1345035820646022e-05,
+      "loss": 0.4569,
+      "step": 3694
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1340922272874143e-05,
+      "loss": 0.2959,
+      "step": 3695
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1336808494033577e-05,
+      "loss": 0.4473,
+      "step": 3696
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1332694484833213e-05,
+      "loss": 0.379,
+      "step": 3697
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.132858024598198e-05,
+      "loss": 0.4096,
+      "step": 3698
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1324465778188846e-05,
+      "loss": 0.4122,
+      "step": 3699
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1320351082162821e-05,
+      "loss": 0.3389,
+      "step": 3700
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1316236158612952e-05,
+      "loss": 0.3514,
+      "step": 3701
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1312121008248323e-05,
+      "loss": 0.304,
+      "step": 3702
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1308005631778061e-05,
+      "loss": 0.342,
+      "step": 3703
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1303890029911333e-05,
+      "loss": 0.4497,
+      "step": 3704
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.129977420335734e-05,
+      "loss": 0.301,
+      "step": 3705
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1295658152825323e-05,
+      "loss": 0.3898,
+      "step": 3706
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1291541879024568e-05,
+      "loss": 0.4751,
+      "step": 3707
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1287425382664388e-05,
+      "loss": 0.3366,
+      "step": 3708
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.128330866445414e-05,
+      "loss": 0.3978,
+      "step": 3709
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1279191725103226e-05,
+      "loss": 0.2839,
+      "step": 3710
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.127507456532108e-05,
+      "loss": 0.3253,
+      "step": 3711
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1270957185817168e-05,
+      "loss": 0.3592,
+      "step": 3712
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1266839587301004e-05,
+      "loss": 0.3813,
+      "step": 3713
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1262721770482135e-05,
+      "loss": 0.3959,
+      "step": 3714
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1258603736070145e-05,
+      "loss": 0.2466,
+      "step": 3715
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1254485484774655e-05,
+      "loss": 0.4141,
+      "step": 3716
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1250367017305328e-05,
+      "loss": 0.3173,
+      "step": 3717
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.124624833437186e-05,
+      "loss": 0.5803,
+      "step": 3718
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1242129436683988e-05,
+      "loss": 0.6174,
+      "step": 3719
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.123801032495148e-05,
+      "loss": 0.3875,
+      "step": 3720
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1233890999884143e-05,
+      "loss": 0.3538,
+      "step": 3721
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.122977146219183e-05,
+      "loss": 0.4668,
+      "step": 3722
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1225651712584413e-05,
+      "loss": 0.4087,
+      "step": 3723
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1221531751771813e-05,
+      "loss": 0.2933,
+      "step": 3724
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.121741158046399e-05,
+      "loss": 0.4321,
+      "step": 3725
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1213291199370927e-05,
+      "loss": 0.3948,
+      "step": 3726
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1209170609202657e-05,
+      "loss": 0.3143,
+      "step": 3727
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1205049810669241e-05,
+      "loss": 0.3685,
+      "step": 3728
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.120092880448078e-05,
+      "loss": 0.3212,
+      "step": 3729
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1196807591347406e-05,
+      "loss": 0.3801,
+      "step": 3730
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1192686171979288e-05,
+      "loss": 0.4119,
+      "step": 3731
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1188564547086634e-05,
+      "loss": 0.3529,
+      "step": 3732
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1184442717379687e-05,
+      "loss": 0.4828,
+      "step": 3733
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.118032068356872e-05,
+      "loss": 0.242,
+      "step": 3734
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1176198446364048e-05,
+      "loss": 0.338,
+      "step": 3735
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1172076006476016e-05,
+      "loss": 0.477,
+      "step": 3736
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1167953364615007e-05,
+      "loss": 0.4181,
+      "step": 3737
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1163830521491435e-05,
+      "loss": 0.3969,
+      "step": 3738
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1159707477815756e-05,
+      "loss": 0.2897,
+      "step": 3739
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1155584234298456e-05,
+      "loss": 0.423,
+      "step": 3740
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1151460791650046e-05,
+      "loss": 0.45,
+      "step": 3741
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.114733715058109e-05,
+      "loss": 0.3516,
+      "step": 3742
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1143213311802177e-05,
+      "loss": 0.2915,
+      "step": 3743
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.113908927602392e-05,
+      "loss": 0.3889,
+      "step": 3744
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1134965043956987e-05,
+      "loss": 0.3786,
+      "step": 3745
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1130840616312065e-05,
+      "loss": 0.5469,
+      "step": 3746
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1126715993799875e-05,
+      "loss": 0.3878,
+      "step": 3747
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1122591177131177e-05,
+      "loss": 0.371,
+      "step": 3748
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1118466167016763e-05,
+      "loss": 0.4156,
+      "step": 3749
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.111434096416746e-05,
+      "loss": 0.3816,
+      "step": 3750
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1110215569294122e-05,
+      "loss": 0.3347,
+      "step": 3751
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1106089983107642e-05,
+      "loss": 0.3909,
+      "step": 3752
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1101964206318938e-05,
+      "loss": 0.3007,
+      "step": 3753
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.109783823963898e-05,
+      "loss": 0.3314,
+      "step": 3754
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1093712083778748e-05,
+      "loss": 0.3679,
+      "step": 3755
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.108958573944926e-05,
+      "loss": 0.6113,
+      "step": 3756
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1085459207361579e-05,
+      "loss": 0.3491,
+      "step": 3757
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1081332488226795e-05,
+      "loss": 0.4241,
+      "step": 3758
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1077205582756015e-05,
+      "loss": 0.3544,
+      "step": 3759
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1073078491660398e-05,
+      "loss": 0.424,
+      "step": 3760
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1068951215651132e-05,
+      "loss": 0.3225,
+      "step": 3761
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1064823755439418e-05,
+      "loss": 0.3112,
+      "step": 3762
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1060696111736515e-05,
+      "loss": 0.3153,
+      "step": 3763
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.10565682852537e-05,
+      "loss": 0.4044,
+      "step": 3764
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1052440276702276e-05,
+      "loss": 0.3926,
+      "step": 3765
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1048312086793593e-05,
+      "loss": 0.3683,
+      "step": 3766
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1044183716239019e-05,
+      "loss": 0.4002,
+      "step": 3767
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1040055165749956e-05,
+      "loss": 0.5419,
+      "step": 3768
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1035926436037845e-05,
+      "loss": 0.4037,
+      "step": 3769
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1031797527814148e-05,
+      "loss": 0.3574,
+      "step": 3770
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1027668441790358e-05,
+      "loss": 0.3733,
+      "step": 3771
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.102353917867801e-05,
+      "loss": 0.4144,
+      "step": 3772
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1019409739188653e-05,
+      "loss": 0.4241,
+      "step": 3773
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.101528012403388e-05,
+      "loss": 0.4019,
+      "step": 3774
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.101115033392531e-05,
+      "loss": 0.3116,
+      "step": 3775
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1007020369574591e-05,
+      "loss": 0.3576,
+      "step": 3776
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1002890231693395e-05,
+      "loss": 0.3198,
+      "step": 3777
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0998759920993438e-05,
+      "loss": 0.4901,
+      "step": 3778
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.099462943818646e-05,
+      "loss": 0.3442,
+      "step": 3779
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0990498783984216e-05,
+      "loss": 0.4304,
+      "step": 3780
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0986367959098517e-05,
+      "loss": 0.4505,
+      "step": 3781
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0982236964241184e-05,
+      "loss": 0.4146,
+      "step": 3782
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0978105800124072e-05,
+      "loss": 0.4858,
+      "step": 3783
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0973974467459068e-05,
+      "loss": 0.3181,
+      "step": 3784
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0969842966958087e-05,
+      "loss": 0.4917,
+      "step": 3785
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0965711299333066e-05,
+      "loss": 0.3648,
+      "step": 3786
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0961579465295987e-05,
+      "loss": 0.4518,
+      "step": 3787
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0957447465558844e-05,
+      "loss": 0.2821,
+      "step": 3788
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0953315300833665e-05,
+      "loss": 0.4618,
+      "step": 3789
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0949182971832514e-05,
+      "loss": 0.4657,
+      "step": 3790
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0945050479267472e-05,
+      "loss": 0.347,
+      "step": 3791
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.094091782385065e-05,
+      "loss": 0.3244,
+      "step": 3792
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0936785006294197e-05,
+      "loss": 0.3148,
+      "step": 3793
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0932652027310282e-05,
+      "loss": 0.4684,
+      "step": 3794
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0928518887611099e-05,
+      "loss": 0.3699,
+      "step": 3795
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0924385587908876e-05,
+      "loss": 0.413,
+      "step": 3796
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.092025212891587e-05,
+      "loss": 0.3497,
+      "step": 3797
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.091611851134435e-05,
+      "loss": 0.3912,
+      "step": 3798
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0911984735906635e-05,
+      "loss": 0.2838,
+      "step": 3799
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0907850803315059e-05,
+      "loss": 0.3861,
+      "step": 3800
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.090371671428198e-05,
+      "loss": 0.2954,
+      "step": 3801
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0899582469519787e-05,
+      "loss": 0.3744,
+      "step": 3802
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0895448069740902e-05,
+      "loss": 0.3273,
+      "step": 3803
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.089131351565776e-05,
+      "loss": 0.4212,
+      "step": 3804
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0887178807982837e-05,
+      "loss": 0.4301,
+      "step": 3805
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0883043947428627e-05,
+      "loss": 0.3124,
+      "step": 3806
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0878908934707646e-05,
+      "loss": 0.5382,
+      "step": 3807
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0874773770532453e-05,
+      "loss": 0.3593,
+      "step": 3808
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0870638455615613e-05,
+      "loss": 0.4458,
+      "step": 3809
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0866502990669729e-05,
+      "loss": 0.3922,
+      "step": 3810
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0862367376407433e-05,
+      "loss": 0.4856,
+      "step": 3811
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0858231613541372e-05,
+      "loss": 0.4327,
+      "step": 3812
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0854095702784222e-05,
+      "loss": 0.5009,
+      "step": 3813
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0849959644848689e-05,
+      "loss": 0.3633,
+      "step": 3814
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.08458234404475e-05,
+      "loss": 0.6004,
+      "step": 3815
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0841687090293412e-05,
+      "loss": 0.374,
+      "step": 3816
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.08375505950992e-05,
+      "loss": 0.5373,
+      "step": 3817
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0833413955577667e-05,
+      "loss": 0.3175,
+      "step": 3818
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0829277172441648e-05,
+      "loss": 0.3214,
+      "step": 3819
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0825140246403989e-05,
+      "loss": 0.4051,
+      "step": 3820
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0821003178177572e-05,
+      "loss": 0.3064,
+      "step": 3821
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0816865968475299e-05,
+      "loss": 0.3821,
+      "step": 3822
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0812728618010098e-05,
+      "loss": 0.341,
+      "step": 3823
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0808591127494919e-05,
+      "loss": 0.3565,
+      "step": 3824
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0804453497642734e-05,
+      "loss": 0.3334,
+      "step": 3825
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.080031572916655e-05,
+      "loss": 0.5331,
+      "step": 3826
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0796177822779384e-05,
+      "loss": 0.41,
+      "step": 3827
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0792039779194283e-05,
+      "loss": 0.4489,
+      "step": 3828
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0787901599124322e-05,
+      "loss": 0.3561,
+      "step": 3829
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0783763283282594e-05,
+      "loss": 0.5295,
+      "step": 3830
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0779624832382211e-05,
+      "loss": 0.4321,
+      "step": 3831
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0775486247136322e-05,
+      "loss": 0.4444,
+      "step": 3832
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0771347528258084e-05,
+      "loss": 0.3689,
+      "step": 3833
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0767208676460689e-05,
+      "loss": 0.3649,
+      "step": 3834
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0763069692457346e-05,
+      "loss": 0.3486,
+      "step": 3835
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0758930576961286e-05,
+      "loss": 0.4765,
+      "step": 3836
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0754791330685763e-05,
+      "loss": 0.5279,
+      "step": 3837
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.075065195434406e-05,
+      "loss": 0.5594,
+      "step": 3838
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0746512448649473e-05,
+      "loss": 0.4096,
+      "step": 3839
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0742372814315323e-05,
+      "loss": 0.3731,
+      "step": 3840
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.073823305205496e-05,
+      "loss": 0.3972,
+      "step": 3841
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0734093162581747e-05,
+      "loss": 0.2965,
+      "step": 3842
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0729953146609076e-05,
+      "loss": 0.3492,
+      "step": 3843
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0725813004850357e-05,
+      "loss": 0.5758,
+      "step": 3844
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0721672738019015e-05,
+      "loss": 0.3782,
+      "step": 3845
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0717532346828514e-05,
+      "loss": 0.3661,
+      "step": 3846
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0713391831992324e-05,
+      "loss": 0.3829,
+      "step": 3847
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0709251194223942e-05,
+      "loss": 0.4912,
+      "step": 3848
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0705110434236888e-05,
+      "loss": 0.478,
+      "step": 3849
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0700969552744695e-05,
+      "loss": 0.3586,
+      "step": 3850
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0696828550460928e-05,
+      "loss": 0.3653,
+      "step": 3851
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0692687428099164e-05,
+      "loss": 0.2871,
+      "step": 3852
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0688546186373009e-05,
+      "loss": 0.3747,
+      "step": 3853
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.068440482599608e-05,
+      "loss": 0.6275,
+      "step": 3854
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0680263347682021e-05,
+      "loss": 0.3945,
+      "step": 3855
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0676121752144495e-05,
+      "loss": 0.3429,
+      "step": 3856
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.067198004009718e-05,
+      "loss": 0.345,
+      "step": 3857
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0667838212253787e-05,
+      "loss": 0.4147,
+      "step": 3858
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0663696269328034e-05,
+      "loss": 0.5598,
+      "step": 3859
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0659554212033663e-05,
+      "loss": 0.4445,
+      "step": 3860
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0655412041084438e-05,
+      "loss": 0.4222,
+      "step": 3861
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0651269757194142e-05,
+      "loss": 0.3552,
+      "step": 3862
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0647127361076573e-05,
+      "loss": 0.3539,
+      "step": 3863
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0642984853445553e-05,
+      "loss": 0.2805,
+      "step": 3864
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0638842235014923e-05,
+      "loss": 0.386,
+      "step": 3865
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.063469950649854e-05,
+      "loss": 0.356,
+      "step": 3866
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0630556668610286e-05,
+      "loss": 0.3817,
+      "step": 3867
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.062641372206405e-05,
+      "loss": 0.4517,
+      "step": 3868
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0622270667573757e-05,
+      "loss": 0.4699,
+      "step": 3869
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0618127505853336e-05,
+      "loss": 0.3747,
+      "step": 3870
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.061398423761674e-05,
+      "loss": 0.4935,
+      "step": 3871
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0609840863577936e-05,
+      "loss": 0.4121,
+      "step": 3872
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0605697384450922e-05,
+      "loss": 0.4234,
+      "step": 3873
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.06015538009497e-05,
+      "loss": 0.345,
+      "step": 3874
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.059741011378829e-05,
+      "loss": 0.352,
+      "step": 3875
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0593266323680748e-05,
+      "loss": 0.3399,
+      "step": 3876
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0589122431341129e-05,
+      "loss": 0.3337,
+      "step": 3877
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0584978437483504e-05,
+      "loss": 0.3978,
+      "step": 3878
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0580834342821978e-05,
+      "loss": 0.2915,
+      "step": 3879
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0576690148070666e-05,
+      "loss": 0.3345,
+      "step": 3880
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0572545853943688e-05,
+      "loss": 0.3161,
+      "step": 3881
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.05684014611552e-05,
+      "loss": 0.5279,
+      "step": 3882
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0564256970419367e-05,
+      "loss": 0.2585,
+      "step": 3883
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0560112382450365e-05,
+      "loss": 0.3691,
+      "step": 3884
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0555967697962399e-05,
+      "loss": 0.4072,
+      "step": 3885
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0551822917669679e-05,
+      "loss": 0.2988,
+      "step": 3886
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0547678042286435e-05,
+      "loss": 0.5151,
+      "step": 3887
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0543533072526925e-05,
+      "loss": 0.4352,
+      "step": 3888
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0539388009105404e-05,
+      "loss": 0.4246,
+      "step": 3889
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0535242852736152e-05,
+      "loss": 0.4331,
+      "step": 3890
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0531097604133473e-05,
+      "loss": 0.3866,
+      "step": 3891
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0526952264011672e-05,
+      "loss": 0.4322,
+      "step": 3892
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0522806833085076e-05,
+      "loss": 0.3574,
+      "step": 3893
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0518661312068032e-05,
+      "loss": 0.4832,
+      "step": 3894
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0514515701674904e-05,
+      "loss": 0.4308,
+      "step": 3895
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0510370002620056e-05,
+      "loss": 0.3325,
+      "step": 3896
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0506224215617887e-05,
+      "loss": 0.3953,
+      "step": 3897
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0502078341382798e-05,
+      "loss": 0.357,
+      "step": 3898
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0497932380629207e-05,
+      "loss": 0.4436,
+      "step": 3899
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0493786334071554e-05,
+      "loss": 0.414,
+      "step": 3900
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0489640202424284e-05,
+      "loss": 0.3581,
+      "step": 3901
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0485493986401865e-05,
+      "loss": 0.4547,
+      "step": 3902
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0481347686718775e-05,
+      "loss": 0.3818,
+      "step": 3903
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0477201304089503e-05,
+      "loss": 0.425,
+      "step": 3904
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0473054839228562e-05,
+      "loss": 0.3033,
+      "step": 3905
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0468908292850475e-05,
+      "loss": 0.3447,
+      "step": 3906
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0464761665669771e-05,
+      "loss": 0.3722,
+      "step": 3907
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0460614958401003e-05,
+      "loss": 0.3188,
+      "step": 3908
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.045646817175874e-05,
+      "loss": 0.3303,
+      "step": 3909
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0452321306457551e-05,
+      "loss": 0.3537,
+      "step": 3910
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0448174363212029e-05,
+      "loss": 0.3845,
+      "step": 3911
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0444027342736782e-05,
+      "loss": 0.4106,
+      "step": 3912
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0439880245746428e-05,
+      "loss": 0.3495,
+      "step": 3913
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0435733072955587e-05,
+      "loss": 0.3413,
+      "step": 3914
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0431585825078916e-05,
+      "loss": 0.3592,
+      "step": 3915
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0427438502831065e-05,
+      "loss": 0.315,
+      "step": 3916
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0423291106926704e-05,
+      "loss": 0.3218,
+      "step": 3917
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0419143638080518e-05,
+      "loss": 0.3837,
+      "step": 3918
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0414996097007199e-05,
+      "loss": 0.3944,
+      "step": 3919
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0410848484421454e-05,
+      "loss": 0.4059,
+      "step": 3920
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0406700801038006e-05,
+      "loss": 0.3554,
+      "step": 3921
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0402553047571584e-05,
+      "loss": 0.4601,
+      "step": 3922
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0398405224736927e-05,
+      "loss": 0.4291,
+      "step": 3923
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0394257333248805e-05,
+      "loss": 0.4405,
+      "step": 3924
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.039010937382197e-05,
+      "loss": 0.3786,
+      "step": 3925
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0385961347171209e-05,
+      "loss": 0.356,
+      "step": 3926
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0381813254011317e-05,
+      "loss": 0.2982,
+      "step": 3927
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0377665095057088e-05,
+      "loss": 0.298,
+      "step": 3928
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0373516871023336e-05,
+      "loss": 0.3886,
+      "step": 3929
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0369368582624893e-05,
+      "loss": 0.2582,
+      "step": 3930
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0365220230576592e-05,
+      "loss": 0.4328,
+      "step": 3931
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0361071815593275e-05,
+      "loss": 0.4177,
+      "step": 3932
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0356923338389807e-05,
+      "loss": 0.4319,
+      "step": 3933
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0352774799681052e-05,
+      "loss": 0.3647,
+      "step": 3934
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.034862620018189e-05,
+      "loss": 0.4753,
+      "step": 3935
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0344477540607214e-05,
+      "loss": 0.26,
+      "step": 3936
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0340328821671918e-05,
+      "loss": 0.4465,
+      "step": 3937
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0336180044090919e-05,
+      "loss": 0.432,
+      "step": 3938
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0332031208579133e-05,
+      "loss": 0.3006,
+      "step": 3939
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.032788231585149e-05,
+      "loss": 0.2848,
+      "step": 3940
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.032373336662293e-05,
+      "loss": 0.4072,
+      "step": 3941
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0319584361608407e-05,
+      "loss": 0.376,
+      "step": 3942
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0315435301522877e-05,
+      "loss": 0.3622,
+      "step": 3943
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0311286187081308e-05,
+      "loss": 0.3545,
+      "step": 3944
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0307137018998683e-05,
+      "loss": 0.3893,
+      "step": 3945
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0302987797989986e-05,
+      "loss": 0.4186,
+      "step": 3946
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0298838524770212e-05,
+      "loss": 0.3838,
+      "step": 3947
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.029468920005437e-05,
+      "loss": 0.3697,
+      "step": 3948
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0290539824557474e-05,
+      "loss": 0.4142,
+      "step": 3949
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0286390398994551e-05,
+      "loss": 0.4325,
+      "step": 3950
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0282240924080625e-05,
+      "loss": 0.4162,
+      "step": 3951
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0278091400530743e-05,
+      "loss": 0.3773,
+      "step": 3952
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.027394182905995e-05,
+      "loss": 0.3573,
+      "step": 3953
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0269792210383305e-05,
+      "loss": 0.3032,
+      "step": 3954
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0265642545215872e-05,
+      "loss": 0.441,
+      "step": 3955
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0261492834272728e-05,
+      "loss": 0.49,
+      "step": 3956
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0257343078268948e-05,
+      "loss": 0.3543,
+      "step": 3957
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0253193277919626e-05,
+      "loss": 0.3128,
+      "step": 3958
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0249043433939857e-05,
+      "loss": 0.2758,
+      "step": 3959
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0244893547044745e-05,
+      "loss": 0.3266,
+      "step": 3960
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.02407436179494e-05,
+      "loss": 0.3341,
+      "step": 3961
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0236593647368944e-05,
+      "loss": 0.2991,
+      "step": 3962
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0232443636018502e-05,
+      "loss": 0.4248,
+      "step": 3963
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0228293584613203e-05,
+      "loss": 0.3169,
+      "step": 3964
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0224143493868191e-05,
+      "loss": 0.365,
+      "step": 3965
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0219993364498612e-05,
+      "loss": 0.3377,
+      "step": 3966
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.021584319721962e-05,
+      "loss": 0.335,
+      "step": 3967
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.021169299274637e-05,
+      "loss": 0.4737,
+      "step": 3968
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0207542751794035e-05,
+      "loss": 0.3955,
+      "step": 3969
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0203392475077783e-05,
+      "loss": 0.5066,
+      "step": 3970
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0199242163312794e-05,
+      "loss": 0.3275,
+      "step": 3971
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0195091817214254e-05,
+      "loss": 0.3886,
+      "step": 3972
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.019094143749735e-05,
+      "loss": 0.2841,
+      "step": 3973
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0186791024877282e-05,
+      "loss": 0.4147,
+      "step": 3974
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0182640580069249e-05,
+      "loss": 0.2721,
+      "step": 3975
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0178490103788462e-05,
+      "loss": 0.3314,
+      "step": 3976
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0174339596750131e-05,
+      "loss": 0.3554,
+      "step": 3977
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.017018905966948e-05,
+      "loss": 0.3799,
+      "step": 3978
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0166038493261723e-05,
+      "loss": 0.4455,
+      "step": 3979
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0161887898242096e-05,
+      "loss": 0.356,
+      "step": 3980
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0157737275325829e-05,
+      "loss": 0.4378,
+      "step": 3981
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0153586625228162e-05,
+      "loss": 0.4533,
+      "step": 3982
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0149435948664338e-05,
+      "loss": 0.4903,
+      "step": 3983
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0145285246349605e-05,
+      "loss": 0.5149,
+      "step": 3984
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0141134518999216e-05,
+      "loss": 0.3199,
+      "step": 3985
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0136983767328424e-05,
+      "loss": 0.3843,
+      "step": 3986
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.013283299205249e-05,
+      "loss": 0.4153,
+      "step": 3987
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.012868219388668e-05,
+      "loss": 0.3755,
+      "step": 3988
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0124531373546262e-05,
+      "loss": 0.3509,
+      "step": 3989
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0120380531746507e-05,
+      "loss": 0.4563,
+      "step": 3990
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0116229669202692e-05,
+      "loss": 0.4134,
+      "step": 3991
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0112078786630104e-05,
+      "loss": 0.3089,
+      "step": 3992
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0107927884744014e-05,
+      "loss": 0.314,
+      "step": 3993
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0103776964259711e-05,
+      "loss": 0.5964,
+      "step": 3994
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0099626025892491e-05,
+      "loss": 0.4518,
+      "step": 3995
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0095475070357643e-05,
+      "loss": 0.3944,
+      "step": 3996
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.009132409837046e-05,
+      "loss": 0.3878,
+      "step": 3997
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0087173110646245e-05,
+      "loss": 0.4116,
+      "step": 3998
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0083022107900295e-05,
+      "loss": 0.3281,
+      "step": 3999
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0078871090847917e-05,
+      "loss": 0.4349,
+      "step": 4000
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0074720060204417e-05,
+      "loss": 0.414,
+      "step": 4001
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.00705690166851e-05,
+      "loss": 0.3983,
+      "step": 4002
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0066417961005283e-05,
+      "loss": 0.3525,
+      "step": 4003
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0062266893880275e-05,
+      "loss": 0.3481,
+      "step": 4004
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0058115816025389e-05,
+      "loss": 0.363,
+      "step": 4005
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0053964728155943e-05,
+      "loss": 0.3872,
+      "step": 4006
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0049813630987262e-05,
+      "loss": 0.382,
+      "step": 4007
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0045662525234657e-05,
+      "loss": 0.3283,
+      "step": 4008
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0041511411613449e-05,
+      "loss": 0.5004,
+      "step": 4009
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0037360290838971e-05,
+      "loss": 0.3221,
+      "step": 4010
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0033209163626539e-05,
+      "loss": 0.3599,
+      "step": 4011
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0029058030691477e-05,
+      "loss": 0.4032,
+      "step": 4012
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0024906892749115e-05,
+      "loss": 0.5077,
+      "step": 4013
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0020755750514786e-05,
+      "loss": 0.3228,
+      "step": 4014
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0016604604703803e-05,
+      "loss": 0.3895,
+      "step": 4015
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0012453456031506e-05,
+      "loss": 0.377,
+      "step": 4016
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.000830230521322e-05,
+      "loss": 0.387,
+      "step": 4017
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0004151152964276e-05,
+      "loss": 0.3869,
+      "step": 4018
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 0.2994,
+      "step": 4019
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.0,
+      "learning_rate": 9.99584884703573e-06,
+      "loss": 0.3991,
+      "step": 4020
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.0,
+      "learning_rate": 9.99169769478678e-06,
+      "loss": 0.4099,
+      "step": 4021
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.0,
+      "learning_rate": 9.987546543968497e-06,
+      "loss": 0.3602,
+      "step": 4022
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.0,
+      "learning_rate": 9.9833953952962e-06,
+      "loss": 0.4119,
+      "step": 4023
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.0,
+      "learning_rate": 9.979244249485218e-06,
+      "loss": 0.3656,
+      "step": 4024
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.0,
+      "learning_rate": 9.975093107250888e-06,
+      "loss": 0.419,
+      "step": 4025
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.0,
+      "learning_rate": 9.970941969308525e-06,
+      "loss": 0.3434,
+      "step": 4026
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.0,
+      "learning_rate": 9.966790836373465e-06,
+      "loss": 0.3488,
+      "step": 4027
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.0,
+      "learning_rate": 9.962639709161035e-06,
+      "loss": 0.3667,
+      "step": 4028
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.0,
+      "learning_rate": 9.958488588386551e-06,
+      "loss": 0.3841,
+      "step": 4029
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.0,
+      "learning_rate": 9.954337474765348e-06,
+      "loss": 0.4676,
+      "step": 4030
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.0,
+      "learning_rate": 9.950186369012744e-06,
+      "loss": 0.3458,
+      "step": 4031
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.0,
+      "learning_rate": 9.946035271844057e-06,
+      "loss": 0.4483,
+      "step": 4032
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.0,
+      "learning_rate": 9.941884183974614e-06,
+      "loss": 0.4992,
+      "step": 4033
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.0,
+      "learning_rate": 9.937733106119732e-06,
+      "loss": 0.3883,
+      "step": 4034
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.0,
+      "learning_rate": 9.933582038994719e-06,
+      "loss": 0.3927,
+      "step": 4035
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.0,
+      "learning_rate": 9.929430983314901e-06,
+      "loss": 0.3064,
+      "step": 4036
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.0,
+      "learning_rate": 9.925279939795583e-06,
+      "loss": 0.4323,
+      "step": 4037
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.0,
+      "learning_rate": 9.921128909152086e-06,
+      "loss": 0.4302,
+      "step": 4038
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.0,
+      "learning_rate": 9.916977892099707e-06,
+      "loss": 0.4789,
+      "step": 4039
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.0,
+      "learning_rate": 9.912826889353756e-06,
+      "loss": 0.3734,
+      "step": 4040
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.0,
+      "learning_rate": 9.908675901629542e-06,
+      "loss": 0.377,
+      "step": 4041
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.0,
+      "learning_rate": 9.904524929642362e-06,
+      "loss": 0.3439,
+      "step": 4042
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.0,
+      "learning_rate": 9.90037397410751e-06,
+      "loss": 0.3648,
+      "step": 4043
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.0,
+      "learning_rate": 9.896223035740292e-06,
+      "loss": 0.2527,
+      "step": 4044
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.0,
+      "learning_rate": 9.892072115255991e-06,
+      "loss": 0.3756,
+      "step": 4045
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.0,
+      "learning_rate": 9.8879212133699e-06,
+      "loss": 0.3415,
+      "step": 4046
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.0,
+      "learning_rate": 9.88377033079731e-06,
+      "loss": 0.4444,
+      "step": 4047
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.0,
+      "learning_rate": 9.879619468253496e-06,
+      "loss": 0.3328,
+      "step": 4048
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.0,
+      "learning_rate": 9.875468626453741e-06,
+      "loss": 0.4093,
+      "step": 4049
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.0,
+      "learning_rate": 9.871317806113322e-06,
+      "loss": 0.4448,
+      "step": 4050
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.0,
+      "learning_rate": 9.867167007947511e-06,
+      "loss": 0.2985,
+      "step": 4051
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.0,
+      "learning_rate": 9.863016232671581e-06,
+      "loss": 0.3587,
+      "step": 4052
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.0,
+      "learning_rate": 9.858865481000787e-06,
+      "loss": 0.3969,
+      "step": 4053
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.0,
+      "learning_rate": 9.854714753650395e-06,
+      "loss": 0.4861,
+      "step": 4054
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.0,
+      "learning_rate": 9.850564051335663e-06,
+      "loss": 0.2618,
+      "step": 4055
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.0,
+      "learning_rate": 9.84641337477184e-06,
+      "loss": 0.4768,
+      "step": 4056
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.0,
+      "learning_rate": 9.842262724674173e-06,
+      "loss": 0.5203,
+      "step": 4057
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.0,
+      "learning_rate": 9.838112101757908e-06,
+      "loss": 0.3587,
+      "step": 4058
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.0,
+      "learning_rate": 9.833961506738282e-06,
+      "loss": 0.428,
+      "step": 4059
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.0,
+      "learning_rate": 9.829810940330524e-06,
+      "loss": 0.3704,
+      "step": 4060
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.0,
+      "learning_rate": 9.82566040324987e-06,
+      "loss": 0.3099,
+      "step": 4061
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.0,
+      "learning_rate": 9.82150989621154e-06,
+      "loss": 0.3248,
+      "step": 4062
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.0,
+      "learning_rate": 9.817359419930753e-06,
+      "loss": 0.4564,
+      "step": 4063
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.0,
+      "learning_rate": 9.813208975122723e-06,
+      "loss": 0.3339,
+      "step": 4064
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.0,
+      "learning_rate": 9.809058562502653e-06,
+      "loss": 0.5775,
+      "step": 4065
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.0,
+      "learning_rate": 9.80490818278575e-06,
+      "loss": 0.3189,
+      "step": 4066
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.0,
+      "learning_rate": 9.80075783668721e-06,
+      "loss": 0.3153,
+      "step": 4067
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.0,
+      "learning_rate": 9.796607524922219e-06,
+      "loss": 0.2893,
+      "step": 4068
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.0,
+      "learning_rate": 9.792457248205967e-06,
+      "loss": 0.2996,
+      "step": 4069
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.0,
+      "learning_rate": 9.788307007253634e-06,
+      "loss": 0.4587,
+      "step": 4070
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.0,
+      "learning_rate": 9.784156802780385e-06,
+      "loss": 0.3579,
+      "step": 4071
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.0,
+      "learning_rate": 9.780006635501391e-06,
+      "loss": 0.3203,
+      "step": 4072
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.0,
+      "learning_rate": 9.77585650613181e-06,
+      "loss": 0.4412,
+      "step": 4073
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.0,
+      "learning_rate": 9.7717064153868e-06,
+      "loss": 0.416,
+      "step": 4074
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.0,
+      "learning_rate": 9.767556363981503e-06,
+      "loss": 0.3908,
+      "step": 4075
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.0,
+      "learning_rate": 9.763406352631058e-06,
+      "loss": 0.4803,
+      "step": 4076
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.0,
+      "learning_rate": 9.759256382050602e-06,
+      "loss": 0.3677,
+      "step": 4077
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.0,
+      "learning_rate": 9.75510645295526e-06,
+      "loss": 0.297,
+      "step": 4078
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.0,
+      "learning_rate": 9.750956566060145e-06,
+      "loss": 0.6437,
+      "step": 4079
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.0,
+      "learning_rate": 9.746806722080376e-06,
+      "loss": 0.4129,
+      "step": 4080
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.0,
+      "learning_rate": 9.742656921731057e-06,
+      "loss": 0.3667,
+      "step": 4081
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.0,
+      "learning_rate": 9.738507165727276e-06,
+      "loss": 0.3665,
+      "step": 4082
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.0,
+      "learning_rate": 9.734357454784131e-06,
+      "loss": 0.4047,
+      "step": 4083
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.0,
+      "learning_rate": 9.7302077896167e-06,
+      "loss": 0.3636,
+      "step": 4084
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.0,
+      "learning_rate": 9.726058170940053e-06,
+      "loss": 0.4373,
+      "step": 4085
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.0,
+      "learning_rate": 9.72190859946926e-06,
+      "loss": 0.3328,
+      "step": 4086
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.0,
+      "learning_rate": 9.717759075919375e-06,
+      "loss": 0.5227,
+      "step": 4087
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.0,
+      "learning_rate": 9.713609601005452e-06,
+      "loss": 0.4457,
+      "step": 4088
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.0,
+      "learning_rate": 9.709460175442527e-06,
+      "loss": 0.276,
+      "step": 4089
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.0,
+      "learning_rate": 9.70531079994563e-06,
+      "loss": 0.4899,
+      "step": 4090
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.0,
+      "learning_rate": 9.701161475229791e-06,
+      "loss": 0.6037,
+      "step": 4091
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.0,
+      "learning_rate": 9.69701220201002e-06,
+      "loss": 0.468,
+      "step": 4092
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.0,
+      "learning_rate": 9.692862981001319e-06,
+      "loss": 0.2827,
+      "step": 4093
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.0,
+      "learning_rate": 9.688713812918695e-06,
+      "loss": 0.2984,
+      "step": 4094
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.0,
+      "learning_rate": 9.684564698477128e-06,
+      "loss": 0.4173,
+      "step": 4095
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.0,
+      "learning_rate": 9.680415638391594e-06,
+      "loss": 0.3049,
+      "step": 4096
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.0,
+      "learning_rate": 9.676266633377073e-06,
+      "loss": 0.3626,
+      "step": 4097
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.0,
+      "learning_rate": 9.672117684148513e-06,
+      "loss": 0.3987,
+      "step": 4098
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.0,
+      "learning_rate": 9.66796879142087e-06,
+      "loss": 0.4474,
+      "step": 4099
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.0,
+      "learning_rate": 9.663819955909086e-06,
+      "loss": 0.3576,
+      "step": 4100
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.0,
+      "learning_rate": 9.659671178328083e-06,
+      "loss": 0.3832,
+      "step": 4101
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.0,
+      "learning_rate": 9.655522459392789e-06,
+      "loss": 0.5484,
+      "step": 4102
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.0,
+      "learning_rate": 9.651373799818114e-06,
+      "loss": 0.3599,
+      "step": 4103
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.0,
+      "learning_rate": 9.647225200318952e-06,
+      "loss": 0.3369,
+      "step": 4104
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.0,
+      "learning_rate": 9.643076661610197e-06,
+      "loss": 0.3815,
+      "step": 4105
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.0,
+      "learning_rate": 9.63892818440673e-06,
+      "loss": 0.3615,
+      "step": 4106
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.0,
+      "learning_rate": 9.634779769423412e-06,
+      "loss": 0.38,
+      "step": 4107
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.0,
+      "learning_rate": 9.63063141737511e-06,
+      "loss": 0.3794,
+      "step": 4108
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.0,
+      "learning_rate": 9.626483128976665e-06,
+      "loss": 0.4,
+      "step": 4109
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.0,
+      "learning_rate": 9.622334904942917e-06,
+      "loss": 0.3786,
+      "step": 4110
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.0,
+      "learning_rate": 9.618186745988688e-06,
+      "loss": 0.4908,
+      "step": 4111
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.0,
+      "learning_rate": 9.614038652828791e-06,
+      "loss": 0.2802,
+      "step": 4112
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.0,
+      "learning_rate": 9.60989062617803e-06,
+      "loss": 0.4401,
+      "step": 4113
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.6057426667512e-06,
+      "loss": 0.4509,
+      "step": 4114
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.601594775263073e-06,
+      "loss": 0.5321,
+      "step": 4115
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.59744695242842e-06,
+      "loss": 0.4011,
+      "step": 4116
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.593299198961999e-06,
+      "loss": 0.3594,
+      "step": 4117
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.589151515578547e-06,
+      "loss": 0.3338,
+      "step": 4118
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.585003902992804e-06,
+      "loss": 0.3722,
+      "step": 4119
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.580856361919484e-06,
+      "loss": 0.4026,
+      "step": 4120
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.576708893073297e-06,
+      "loss": 0.4539,
+      "step": 4121
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.572561497168937e-06,
+      "loss": 0.3655,
+      "step": 4122
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.568414174921085e-06,
+      "loss": 0.4443,
+      "step": 4123
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.564266927044414e-06,
+      "loss": 0.3733,
+      "step": 4124
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.560119754253578e-06,
+      "loss": 0.3377,
+      "step": 4125
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.555972657263218e-06,
+      "loss": 0.3476,
+      "step": 4126
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.551825636787974e-06,
+      "loss": 0.3514,
+      "step": 4127
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.547678693542454e-06,
+      "loss": 0.2863,
+      "step": 4128
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.543531828241263e-06,
+      "loss": 0.2585,
+      "step": 4129
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.539385041599e-06,
+      "loss": 0.4336,
+      "step": 4130
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.535238334330234e-06,
+      "loss": 0.3362,
+      "step": 4131
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.531091707149528e-06,
+      "loss": 0.4187,
+      "step": 4132
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.526945160771441e-06,
+      "loss": 0.4057,
+      "step": 4133
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.522798695910498e-06,
+      "loss": 0.4807,
+      "step": 4134
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.518652313281229e-06,
+      "loss": 0.3355,
+      "step": 4135
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.51450601359814e-06,
+      "loss": 0.3671,
+      "step": 4136
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.510359797575716e-06,
+      "loss": 0.5178,
+      "step": 4137
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.50621366592845e-06,
+      "loss": 0.408,
+      "step": 4138
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.502067619370794e-06,
+      "loss": 0.332,
+      "step": 4139
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.497921658617202e-06,
+      "loss": 0.3955,
+      "step": 4140
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.493775784382115e-06,
+      "loss": 0.4523,
+      "step": 4141
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.489629997379946e-06,
+      "loss": 0.3774,
+      "step": 4142
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.485484298325098e-06,
+      "loss": 0.4351,
+      "step": 4143
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.48133868793197e-06,
+      "loss": 0.3593,
+      "step": 4144
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.477193166914926e-06,
+      "loss": 0.4114,
+      "step": 4145
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.473047735988331e-06,
+      "loss": 0.4413,
+      "step": 4146
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.468902395866532e-06,
+      "loss": 0.3622,
+      "step": 4147
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.464757147263849e-06,
+      "loss": 0.2669,
+      "step": 4148
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.460611990894599e-06,
+      "loss": 0.3796,
+      "step": 4149
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.45646692747308e-06,
+      "loss": 0.334,
+      "step": 4150
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.452321957713563e-06,
+      "loss": 0.3069,
+      "step": 4151
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.448177082330324e-06,
+      "loss": 0.3967,
+      "step": 4152
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.444032302037608e-06,
+      "loss": 0.3415,
+      "step": 4153
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.439887617549637e-06,
+      "loss": 0.5422,
+      "step": 4154
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.435743029580638e-06,
+      "loss": 0.3687,
+      "step": 4155
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.431598538844801e-06,
+      "loss": 0.4134,
+      "step": 4156
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.427454146056315e-06,
+      "loss": 0.3751,
+      "step": 4157
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.42330985192934e-06,
+      "loss": 0.4294,
+      "step": 4158
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.419165657178022e-06,
+      "loss": 0.4349,
+      "step": 4159
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.4150215625165e-06,
+      "loss": 0.5214,
+      "step": 4160
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.410877568658876e-06,
+      "loss": 0.4176,
+      "step": 4161
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.406733676319253e-06,
+      "loss": 0.2483,
+      "step": 4162
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.402589886211711e-06,
+      "loss": 0.4357,
+      "step": 4163
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.398446199050305e-06,
+      "loss": 0.4223,
+      "step": 4164
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.394302615549081e-06,
+      "loss": 0.4456,
+      "step": 4165
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.390159136422068e-06,
+      "loss": 0.4933,
+      "step": 4166
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.386015762383262e-06,
+      "loss": 0.4486,
+      "step": 4167
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.381872494146667e-06,
+      "loss": 0.3568,
+      "step": 4168
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.377729332426247e-06,
+      "loss": 0.3849,
+      "step": 4169
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.373586277935948e-06,
+      "loss": 0.3967,
+      "step": 4170
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.369443331389718e-06,
+      "loss": 0.3034,
+      "step": 4171
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.365300493501463e-06,
+      "loss": 0.3551,
+      "step": 4172
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.361157764985079e-06,
+      "loss": 0.307,
+      "step": 4173
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.35701514655445e-06,
+      "loss": 0.3524,
+      "step": 4174
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.0,
+      "learning_rate": 9.35287263892343e-06,
+      "loss": 0.5345,
+      "step": 4175
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.0,
+      "learning_rate": 9.34873024280586e-06,
+      "loss": 0.4327,
+      "step": 4176
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.0,
+      "learning_rate": 9.344587958915566e-06,
+      "loss": 0.4353,
+      "step": 4177
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.0,
+      "learning_rate": 9.34044578796634e-06,
+      "loss": 0.4052,
+      "step": 4178
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.0,
+      "learning_rate": 9.336303730671968e-06,
+      "loss": 0.294,
+      "step": 4179
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.0,
+      "learning_rate": 9.332161787746216e-06,
+      "loss": 0.4488,
+      "step": 4180
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.0,
+      "learning_rate": 9.328019959902821e-06,
+      "loss": 0.4315,
+      "step": 4181
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.0,
+      "learning_rate": 9.323878247855509e-06,
+      "loss": 0.3861,
+      "step": 4182
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.0,
+      "learning_rate": 9.319736652317984e-06,
+      "loss": 0.3332,
+      "step": 4183
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.0,
+      "learning_rate": 9.315595174003922e-06,
+      "loss": 0.3611,
+      "step": 4184
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.0,
+      "learning_rate": 9.311453813626993e-06,
+      "loss": 0.3358,
+      "step": 4185
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.0,
+      "learning_rate": 9.30731257190084e-06,
+      "loss": 0.3381,
+      "step": 4186
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.0,
+      "learning_rate": 9.303171449539074e-06,
+      "loss": 0.3399,
+      "step": 4187
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.0,
+      "learning_rate": 9.299030447255308e-06,
+      "loss": 0.2962,
+      "step": 4188
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.0,
+      "learning_rate": 9.294889565763119e-06,
+      "loss": 0.3825,
+      "step": 4189
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.0,
+      "learning_rate": 9.29074880577606e-06,
+      "loss": 0.3843,
+      "step": 4190
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.0,
+      "learning_rate": 9.286608168007678e-06,
+      "loss": 0.3079,
+      "step": 4191
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.0,
+      "learning_rate": 9.282467653171486e-06,
+      "loss": 0.6341,
+      "step": 4192
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.0,
+      "learning_rate": 9.278327261980986e-06,
+      "loss": 0.4269,
+      "step": 4193
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.0,
+      "learning_rate": 9.274186995149648e-06,
+      "loss": 0.3036,
+      "step": 4194
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.0,
+      "learning_rate": 9.270046853390924e-06,
+      "loss": 0.4765,
+      "step": 4195
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.0,
+      "learning_rate": 9.265906837418255e-06,
+      "loss": 0.3076,
+      "step": 4196
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.0,
+      "learning_rate": 9.261766947945043e-06,
+      "loss": 0.4964,
+      "step": 4197
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.0,
+      "learning_rate": 9.257627185684679e-06,
+      "loss": 0.3558,
+      "step": 4198
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.0,
+      "learning_rate": 9.25348755135053e-06,
+      "loss": 0.3026,
+      "step": 4199
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.0,
+      "learning_rate": 9.249348045655946e-06,
+      "loss": 0.4326,
+      "step": 4200
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.0,
+      "learning_rate": 9.245208669314238e-06,
+      "loss": 0.5111,
+      "step": 4201
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.0,
+      "learning_rate": 9.241069423038717e-06,
+      "loss": 0.4639,
+      "step": 4202
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.0,
+      "learning_rate": 9.236930307542654e-06,
+      "loss": 0.3349,
+      "step": 4203
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.0,
+      "learning_rate": 9.232791323539313e-06,
+      "loss": 0.3663,
+      "step": 4204
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.0,
+      "learning_rate": 9.228652471741917e-06,
+      "loss": 0.4037,
+      "step": 4205
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.0,
+      "learning_rate": 9.224513752863678e-06,
+      "loss": 0.3304,
+      "step": 4206
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.0,
+      "learning_rate": 9.22037516761779e-06,
+      "loss": 0.4115,
+      "step": 4207
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.0,
+      "learning_rate": 9.216236716717411e-06,
+      "loss": 0.3433,
+      "step": 4208
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.0,
+      "learning_rate": 9.212098400875678e-06,
+      "loss": 0.4478,
+      "step": 4209
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.0,
+      "learning_rate": 9.207960220805719e-06,
+      "loss": 0.4499,
+      "step": 4210
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.0,
+      "learning_rate": 9.203822177220621e-06,
+      "loss": 0.3943,
+      "step": 4211
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.0,
+      "learning_rate": 9.199684270833453e-06,
+      "loss": 0.5067,
+      "step": 4212
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.0,
+      "learning_rate": 9.195546502357268e-06,
+      "loss": 0.5688,
+      "step": 4213
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.0,
+      "learning_rate": 9.191408872505086e-06,
+      "loss": 0.5052,
+      "step": 4214
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.0,
+      "learning_rate": 9.187271381989906e-06,
+      "loss": 0.4543,
+      "step": 4215
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.0,
+      "learning_rate": 9.183134031524706e-06,
+      "loss": 0.4573,
+      "step": 4216
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.0,
+      "learning_rate": 9.17899682182243e-06,
+      "loss": 0.4578,
+      "step": 4217
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.0,
+      "learning_rate": 9.174859753596013e-06,
+      "loss": 0.3906,
+      "step": 4218
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.0,
+      "learning_rate": 9.170722827558357e-06,
+      "loss": 0.3374,
+      "step": 4219
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.0,
+      "learning_rate": 9.166586044422334e-06,
+      "loss": 0.2798,
+      "step": 4220
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.0,
+      "learning_rate": 9.162449404900803e-06,
+      "loss": 0.3429,
+      "step": 4221
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.0,
+      "learning_rate": 9.158312909706593e-06,
+      "loss": 0.3776,
+      "step": 4222
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.0,
+      "learning_rate": 9.154176559552502e-06,
+      "loss": 0.3516,
+      "step": 4223
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.0,
+      "learning_rate": 9.150040355151314e-06,
+      "loss": 0.3726,
+      "step": 4224
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.0,
+      "learning_rate": 9.145904297215785e-06,
+      "loss": 0.5461,
+      "step": 4225
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.0,
+      "learning_rate": 9.14176838645863e-06,
+      "loss": 0.3716,
+      "step": 4226
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.13763262359257e-06,
+      "loss": 0.3696,
+      "step": 4227
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.133497009330271e-06,
+      "loss": 0.3932,
+      "step": 4228
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.12936154438439e-06,
+      "loss": 0.2956,
+      "step": 4229
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.125226229467553e-06,
+      "loss": 0.4662,
+      "step": 4230
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.121091065292356e-06,
+      "loss": 0.3138,
+      "step": 4231
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.116956052571377e-06,
+      "loss": 0.3344,
+      "step": 4232
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.112821192017168e-06,
+      "loss": 0.4101,
+      "step": 4233
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.108686484342241e-06,
+      "loss": 0.3683,
+      "step": 4234
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.104551930259101e-06,
+      "loss": 0.4664,
+      "step": 4235
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.0,
+      "learning_rate": 9.100417530480216e-06,
+      "loss": 0.3188,
+      "step": 4236
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.0,
+      "learning_rate": 9.096283285718025e-06,
+      "loss": 0.2923,
+      "step": 4237
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.0,
+      "learning_rate": 9.092149196684945e-06,
+      "loss": 0.39,
+      "step": 4238
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.0,
+      "learning_rate": 9.088015264093365e-06,
+      "loss": 0.3317,
+      "step": 4239
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.0,
+      "learning_rate": 9.083881488655653e-06,
+      "loss": 0.3657,
+      "step": 4240
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.0,
+      "learning_rate": 9.079747871084137e-06,
+      "loss": 0.4406,
+      "step": 4241
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.0,
+      "learning_rate": 9.075614412091126e-06,
+      "loss": 0.4401,
+      "step": 4242
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.0,
+      "learning_rate": 9.071481112388905e-06,
+      "loss": 0.3379,
+      "step": 4243
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.0,
+      "learning_rate": 9.067347972689722e-06,
+      "loss": 0.4593,
+      "step": 4244
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.0,
+      "learning_rate": 9.063214993705804e-06,
+      "loss": 0.3987,
+      "step": 4245
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.0,
+      "learning_rate": 9.059082176149353e-06,
+      "loss": 0.3495,
+      "step": 4246
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.0,
+      "learning_rate": 9.054949520732535e-06,
+      "loss": 0.4097,
+      "step": 4247
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.0,
+      "learning_rate": 9.050817028167489e-06,
+      "loss": 0.4397,
+      "step": 4248
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.0,
+      "learning_rate": 9.046684699166338e-06,
+      "loss": 0.3599,
+      "step": 4249
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.0,
+      "learning_rate": 9.04255253444116e-06,
+      "loss": 0.3957,
+      "step": 4250
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.0,
+      "learning_rate": 9.038420534704015e-06,
+      "loss": 0.4337,
+      "step": 4251
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.0,
+      "learning_rate": 9.034288700666936e-06,
+      "loss": 0.3554,
+      "step": 4252
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.0,
+      "learning_rate": 9.030157033041918e-06,
+      "loss": 0.3438,
+      "step": 4253
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.0,
+      "learning_rate": 9.026025532540935e-06,
+      "loss": 0.3242,
+      "step": 4254
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.0,
+      "learning_rate": 9.021894199875934e-06,
+      "loss": 0.3944,
+      "step": 4255
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.0,
+      "learning_rate": 9.017763035758818e-06,
+      "loss": 0.3925,
+      "step": 4256
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.0,
+      "learning_rate": 9.013632040901485e-06,
+      "loss": 0.3146,
+      "step": 4257
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.0,
+      "learning_rate": 9.009501216015785e-06,
+      "loss": 0.3661,
+      "step": 4258
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.0,
+      "learning_rate": 9.005370561813545e-06,
+      "loss": 0.3584,
+      "step": 4259
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.0,
+      "learning_rate": 9.001240079006565e-06,
+      "loss": 0.3335,
+      "step": 4260
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.0,
+      "learning_rate": 8.997109768306607e-06,
+      "loss": 0.3576,
+      "step": 4261
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.0,
+      "learning_rate": 8.992979630425412e-06,
+      "loss": 0.4226,
+      "step": 4262
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.0,
+      "learning_rate": 8.988849666074694e-06,
+      "loss": 0.3431,
+      "step": 4263
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.0,
+      "learning_rate": 8.984719875966121e-06,
+      "loss": 0.5265,
+      "step": 4264
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.0,
+      "learning_rate": 8.980590260811349e-06,
+      "loss": 0.4766,
+      "step": 4265
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.0,
+      "learning_rate": 8.976460821321996e-06,
+      "loss": 0.3397,
+      "step": 4266
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.0,
+      "learning_rate": 8.972331558209644e-06,
+      "loss": 0.3902,
+      "step": 4267
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.0,
+      "learning_rate": 8.968202472185856e-06,
+      "loss": 0.3975,
+      "step": 4268
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.0,
+      "learning_rate": 8.964073563962158e-06,
+      "loss": 0.3981,
+      "step": 4269
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.0,
+      "learning_rate": 8.959944834250045e-06,
+      "loss": 0.2591,
+      "step": 4270
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.0,
+      "learning_rate": 8.955816283760984e-06,
+      "loss": 0.3937,
+      "step": 4271
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.0,
+      "learning_rate": 8.95168791320641e-06,
+      "loss": 0.2805,
+      "step": 4272
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.0,
+      "learning_rate": 8.947559723297726e-06,
+      "loss": 0.4221,
+      "step": 4273
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.0,
+      "learning_rate": 8.943431714746304e-06,
+      "loss": 0.3637,
+      "step": 4274
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.0,
+      "learning_rate": 8.939303888263485e-06,
+      "loss": 0.4687,
+      "step": 4275
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.0,
+      "learning_rate": 8.935176244560584e-06,
+      "loss": 0.3358,
+      "step": 4276
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.0,
+      "learning_rate": 8.931048784348875e-06,
+      "loss": 0.3875,
+      "step": 4277
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.0,
+      "learning_rate": 8.926921508339601e-06,
+      "loss": 0.4396,
+      "step": 4278
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.0,
+      "learning_rate": 8.922794417243989e-06,
+      "loss": 0.3911,
+      "step": 4279
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.0,
+      "learning_rate": 8.91866751177321e-06,
+      "loss": 0.4251,
+      "step": 4280
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.0,
+      "learning_rate": 8.91454079263842e-06,
+      "loss": 0.3432,
+      "step": 4281
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.0,
+      "learning_rate": 8.910414260550743e-06,
+      "loss": 0.3222,
+      "step": 4282
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.0,
+      "learning_rate": 8.906287916221259e-06,
+      "loss": 0.4267,
+      "step": 4283
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.0,
+      "learning_rate": 8.902161760361024e-06,
+      "loss": 0.3101,
+      "step": 4284
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.0,
+      "learning_rate": 8.898035793681064e-06,
+      "loss": 0.4769,
+      "step": 4285
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.0,
+      "learning_rate": 8.89391001689236e-06,
+      "loss": 0.3937,
+      "step": 4286
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.0,
+      "learning_rate": 8.889784430705881e-06,
+      "loss": 0.2469,
+      "step": 4287
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.0,
+      "learning_rate": 8.885659035832543e-06,
+      "loss": 0.3076,
+      "step": 4288
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.0,
+      "learning_rate": 8.881533832983235e-06,
+      "loss": 0.4776,
+      "step": 4289
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.0,
+      "learning_rate": 8.877408822868825e-06,
+      "loss": 0.2886,
+      "step": 4290
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.0,
+      "learning_rate": 8.873284006200129e-06,
+      "loss": 0.4405,
+      "step": 4291
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.0,
+      "learning_rate": 8.869159383687937e-06,
+      "loss": 0.66,
+      "step": 4292
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.0,
+      "learning_rate": 8.865034956043017e-06,
+      "loss": 0.4455,
+      "step": 4293
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.0,
+      "learning_rate": 8.860910723976083e-06,
+      "loss": 0.3319,
+      "step": 4294
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.0,
+      "learning_rate": 8.856786688197827e-06,
+      "loss": 0.286,
+      "step": 4295
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.0,
+      "learning_rate": 8.852662849418914e-06,
+      "loss": 0.5035,
+      "step": 4296
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.0,
+      "learning_rate": 8.848539208349956e-06,
+      "loss": 0.335,
+      "step": 4297
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.0,
+      "learning_rate": 8.844415765701549e-06,
+      "loss": 0.33,
+      "step": 4298
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.0,
+      "learning_rate": 8.840292522184247e-06,
+      "loss": 0.3852,
+      "step": 4299
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.0,
+      "learning_rate": 8.836169478508565e-06,
+      "loss": 0.4512,
+      "step": 4300
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.0,
+      "learning_rate": 8.832046635384996e-06,
+      "loss": 0.3521,
+      "step": 4301
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.0,
+      "learning_rate": 8.827923993523989e-06,
+      "loss": 0.4342,
+      "step": 4302
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.0,
+      "learning_rate": 8.823801553635955e-06,
+      "loss": 0.5099,
+      "step": 4303
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.0,
+      "learning_rate": 8.819679316431282e-06,
+      "loss": 0.3086,
+      "step": 4304
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.0,
+      "learning_rate": 8.81555728262032e-06,
+      "loss": 0.3159,
+      "step": 4305
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.0,
+      "learning_rate": 8.81143545291337e-06,
+      "loss": 0.3095,
+      "step": 4306
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.0,
+      "learning_rate": 8.807313828020715e-06,
+      "loss": 0.389,
+      "step": 4307
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.0,
+      "learning_rate": 8.803192408652601e-06,
+      "loss": 0.4705,
+      "step": 4308
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.0,
+      "learning_rate": 8.799071195519224e-06,
+      "loss": 0.4032,
+      "step": 4309
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.0,
+      "learning_rate": 8.79495018933076e-06,
+      "loss": 0.4698,
+      "step": 4310
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.0,
+      "learning_rate": 8.790829390797342e-06,
+      "loss": 0.4592,
+      "step": 4311
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.0,
+      "learning_rate": 8.786708800629074e-06,
+      "loss": 0.3785,
+      "step": 4312
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.0,
+      "learning_rate": 8.782588419536015e-06,
+      "loss": 0.3488,
+      "step": 4313
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.0,
+      "learning_rate": 8.778468248228187e-06,
+      "loss": 0.3558,
+      "step": 4314
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.0,
+      "learning_rate": 8.774348287415589e-06,
+      "loss": 0.297,
+      "step": 4315
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.0,
+      "learning_rate": 8.770228537808177e-06,
+      "loss": 0.3246,
+      "step": 4316
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.0,
+      "learning_rate": 8.766109000115857e-06,
+      "loss": 0.3862,
+      "step": 4317
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.0,
+      "learning_rate": 8.761989675048523e-06,
+      "loss": 0.4279,
+      "step": 4318
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.0,
+      "learning_rate": 8.757870563316018e-06,
+      "loss": 0.3437,
+      "step": 4319
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.0,
+      "learning_rate": 8.753751665628141e-06,
+      "loss": 0.3274,
+      "step": 4320
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.0,
+      "learning_rate": 8.749632982694675e-06,
+      "loss": 0.3507,
+      "step": 4321
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.0,
+      "learning_rate": 8.745514515225347e-06,
+      "loss": 0.3279,
+      "step": 4322
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.0,
+      "learning_rate": 8.74139626392986e-06,
+      "loss": 0.3934,
+      "step": 4323
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.0,
+      "learning_rate": 8.73727822951787e-06,
+      "loss": 0.3504,
+      "step": 4324
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.0,
+      "learning_rate": 8.733160412698996e-06,
+      "loss": 0.5422,
+      "step": 4325
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.0,
+      "learning_rate": 8.729042814182835e-06,
+      "loss": 0.4112,
+      "step": 4326
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.0,
+      "learning_rate": 8.724925434678923e-06,
+      "loss": 0.3887,
+      "step": 4327
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.0,
+      "learning_rate": 8.720808274896772e-06,
+      "loss": 0.3078,
+      "step": 4328
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.0,
+      "learning_rate": 8.716691335545862e-06,
+      "loss": 0.3673,
+      "step": 4329
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.0,
+      "learning_rate": 8.712574617335617e-06,
+      "loss": 0.4145,
+      "step": 4330
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.708458120975436e-06,
+      "loss": 0.2913,
+      "step": 4331
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.70434184717468e-06,
+      "loss": 0.4014,
+      "step": 4332
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.700225796642663e-06,
+      "loss": 0.4308,
+      "step": 4333
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.696109970088668e-06,
+      "loss": 0.3986,
+      "step": 4334
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.691994368221942e-06,
+      "loss": 0.2945,
+      "step": 4335
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.687878991751679e-06,
+      "loss": 0.2454,
+      "step": 4336
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.683763841387051e-06,
+      "loss": 0.4299,
+      "step": 4337
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.679648917837184e-06,
+      "loss": 0.3805,
+      "step": 4338
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.675534221811156e-06,
+      "loss": 0.2977,
+      "step": 4339
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 0.0,
+      "learning_rate": 8.671419754018021e-06,
+      "loss": 0.371,
+      "step": 4340
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 0.0,
+      "learning_rate": 8.667305515166792e-06,
+      "loss": 0.3336,
+      "step": 4341
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 0.0,
+      "learning_rate": 8.663191505966426e-06,
+      "loss": 0.2799,
+      "step": 4342
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 0.0,
+      "learning_rate": 8.659077727125859e-06,
+      "loss": 0.3964,
+      "step": 4343
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 0.0,
+      "learning_rate": 8.654964179353983e-06,
+      "loss": 0.3166,
+      "step": 4344
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 0.0,
+      "learning_rate": 8.650850863359636e-06,
+      "loss": 0.2919,
+      "step": 4345
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 0.0,
+      "learning_rate": 8.646737779851643e-06,
+      "loss": 0.2737,
+      "step": 4346
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 0.0,
+      "learning_rate": 8.64262492953876e-06,
+      "loss": 0.4306,
+      "step": 4347
+    },
+    {
+      "epoch": 5.01,
+      "grad_norm": 0.0,
+      "learning_rate": 8.638512313129725e-06,
+      "loss": 0.4247,
+      "step": 4348
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 0.0,
+      "learning_rate": 8.634399931333226e-06,
+      "loss": 0.3569,
+      "step": 4349
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 0.0,
+      "learning_rate": 8.630287784857905e-06,
+      "loss": 0.2967,
+      "step": 4350
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 0.0,
+      "learning_rate": 8.626175874412376e-06,
+      "loss": 0.3264,
+      "step": 4351
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 0.0,
+      "learning_rate": 8.622064200705209e-06,
+      "loss": 0.2697,
+      "step": 4352
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 0.0,
+      "learning_rate": 8.617952764444922e-06,
+      "loss": 0.2953,
+      "step": 4353
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 0.0,
+      "learning_rate": 8.613841566340007e-06,
+      "loss": 0.2661,
+      "step": 4354
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 0.0,
+      "learning_rate": 8.60973060709891e-06,
+      "loss": 0.3118,
+      "step": 4355
+    },
+    {
+      "epoch": 5.02,
+      "grad_norm": 0.0,
+      "learning_rate": 8.605619887430025e-06,
+      "loss": 0.3402,
+      "step": 4356
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 0.0,
+      "learning_rate": 8.601509408041722e-06,
+      "loss": 0.2841,
+      "step": 4357
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 0.0,
+      "learning_rate": 8.59739916964232e-06,
+      "loss": 0.2404,
+      "step": 4358
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 0.0,
+      "learning_rate": 8.593289172940102e-06,
+      "loss": 0.3085,
+      "step": 4359
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 0.0,
+      "learning_rate": 8.589179418643302e-06,
+      "loss": 0.3267,
+      "step": 4360
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 0.0,
+      "learning_rate": 8.58506990746011e-06,
+      "loss": 0.3826,
+      "step": 4361
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 0.0,
+      "learning_rate": 8.580960640098692e-06,
+      "loss": 0.3136,
+      "step": 4362
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 0.0,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.3306,
+      "step": 4363
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 0.0,
+      "learning_rate": 8.572742839673557e-06,
+      "loss": 0.2557,
+      "step": 4364
+    },
+    {
+      "epoch": 5.03,
+      "grad_norm": 0.0,
+      "learning_rate": 8.568634308025945e-06,
+      "loss": 0.3728,
+      "step": 4365
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 0.0,
+      "learning_rate": 8.564526023032291e-06,
+      "loss": 0.2991,
+      "step": 4366
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 0.0,
+      "learning_rate": 8.560417985400541e-06,
+      "loss": 0.3454,
+      "step": 4367
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 0.0,
+      "learning_rate": 8.5563101958386e-06,
+      "loss": 0.3259,
+      "step": 4368
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 0.0,
+      "learning_rate": 8.552202655054315e-06,
+      "loss": 0.3703,
+      "step": 4369
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 0.0,
+      "learning_rate": 8.54809536375551e-06,
+      "loss": 0.2954,
+      "step": 4370
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 0.0,
+      "learning_rate": 8.543988322649954e-06,
+      "loss": 0.3287,
+      "step": 4371
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 0.0,
+      "learning_rate": 8.53988153244537e-06,
+      "loss": 0.242,
+      "step": 4372
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 0.0,
+      "learning_rate": 8.535774993849449e-06,
+      "loss": 0.2361,
+      "step": 4373
+    },
+    {
+      "epoch": 5.04,
+      "grad_norm": 0.0,
+      "learning_rate": 8.531668707569832e-06,
+      "loss": 0.2853,
+      "step": 4374
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 0.0,
+      "learning_rate": 8.527562674314109e-06,
+      "loss": 0.35,
+      "step": 4375
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 0.0,
+      "learning_rate": 8.52345689478985e-06,
+      "loss": 0.3399,
+      "step": 4376
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 0.0,
+      "learning_rate": 8.519351369704553e-06,
+      "loss": 0.3178,
+      "step": 4377
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 0.0,
+      "learning_rate": 8.515246099765687e-06,
+      "loss": 0.2985,
+      "step": 4378
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 0.0,
+      "learning_rate": 8.511141085680684e-06,
+      "loss": 0.2923,
+      "step": 4379
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 0.0,
+      "learning_rate": 8.507036328156908e-06,
+      "loss": 0.3862,
+      "step": 4380
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 0.0,
+      "learning_rate": 8.502931827901706e-06,
+      "loss": 0.2659,
+      "step": 4381
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 0.0,
+      "learning_rate": 8.498827585622367e-06,
+      "loss": 0.2802,
+      "step": 4382
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 0.0,
+      "learning_rate": 8.494723602026131e-06,
+      "loss": 0.4285,
+      "step": 4383
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 0.0,
+      "learning_rate": 8.490619877820205e-06,
+      "loss": 0.3324,
+      "step": 4384
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 0.0,
+      "learning_rate": 8.486516413711743e-06,
+      "loss": 0.3712,
+      "step": 4385
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 0.0,
+      "learning_rate": 8.482413210407855e-06,
+      "loss": 0.3549,
+      "step": 4386
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 0.0,
+      "learning_rate": 8.478310268615612e-06,
+      "loss": 0.2807,
+      "step": 4387
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 0.0,
+      "learning_rate": 8.474207589042038e-06,
+      "loss": 0.2468,
+      "step": 4388
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 0.0,
+      "learning_rate": 8.4701051723941e-06,
+      "loss": 0.3361,
+      "step": 4389
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 0.0,
+      "learning_rate": 8.466003019378736e-06,
+      "loss": 0.3312,
+      "step": 4390
+    },
+    {
+      "epoch": 5.06,
+      "grad_norm": 0.0,
+      "learning_rate": 8.461901130702836e-06,
+      "loss": 0.321,
+      "step": 4391
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 0.0,
+      "learning_rate": 8.457799507073229e-06,
+      "loss": 0.2782,
+      "step": 4392
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 0.0,
+      "learning_rate": 8.45369814919672e-06,
+      "loss": 0.2914,
+      "step": 4393
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 0.0,
+      "learning_rate": 8.449597057780052e-06,
+      "loss": 0.3255,
+      "step": 4394
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 0.0,
+      "learning_rate": 8.445496233529934e-06,
+      "loss": 0.2869,
+      "step": 4395
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 0.0,
+      "learning_rate": 8.441395677153017e-06,
+      "loss": 0.358,
+      "step": 4396
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 0.0,
+      "learning_rate": 8.437295389355912e-06,
+      "loss": 0.3817,
+      "step": 4397
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 0.0,
+      "learning_rate": 8.43319537084519e-06,
+      "loss": 0.2723,
+      "step": 4398
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 0.0,
+      "learning_rate": 8.429095622327363e-06,
+      "loss": 0.3323,
+      "step": 4399
+    },
+    {
+      "epoch": 5.07,
+      "grad_norm": 0.0,
+      "learning_rate": 8.424996144508902e-06,
+      "loss": 0.3039,
+      "step": 4400
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 0.0,
+      "learning_rate": 8.42089693809624e-06,
+      "loss": 0.4348,
+      "step": 4401
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 0.0,
+      "learning_rate": 8.416798003795744e-06,
+      "loss": 0.3222,
+      "step": 4402
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 0.0,
+      "learning_rate": 8.41269934231375e-06,
+      "loss": 0.3279,
+      "step": 4403
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 0.0,
+      "learning_rate": 8.408600954356548e-06,
+      "loss": 0.2839,
+      "step": 4404
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 0.0,
+      "learning_rate": 8.404502840630363e-06,
+      "loss": 0.3375,
+      "step": 4405
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 0.0,
+      "learning_rate": 8.4004050018414e-06,
+      "loss": 0.2967,
+      "step": 4406
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 0.0,
+      "learning_rate": 8.39630743869579e-06,
+      "loss": 0.3221,
+      "step": 4407
+    },
+    {
+      "epoch": 5.08,
+      "grad_norm": 0.0,
+      "learning_rate": 8.39221015189963e-06,
+      "loss": 0.3345,
+      "step": 4408
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 0.0,
+      "learning_rate": 8.388113142158972e-06,
+      "loss": 0.3228,
+      "step": 4409
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 0.0,
+      "learning_rate": 8.384016410179811e-06,
+      "loss": 0.3964,
+      "step": 4410
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 0.0,
+      "learning_rate": 8.3799199566681e-06,
+      "loss": 0.3122,
+      "step": 4411
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 0.0,
+      "learning_rate": 8.375823782329747e-06,
+      "loss": 0.2993,
+      "step": 4412
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 0.0,
+      "learning_rate": 8.3717278878706e-06,
+      "loss": 0.2214,
+      "step": 4413
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 0.0,
+      "learning_rate": 8.36763227399647e-06,
+      "loss": 0.3755,
+      "step": 4414
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 0.0,
+      "learning_rate": 8.363536941413121e-06,
+      "loss": 0.3086,
+      "step": 4415
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 0.0,
+      "learning_rate": 8.359441890826254e-06,
+      "loss": 0.2707,
+      "step": 4416
+    },
+    {
+      "epoch": 5.09,
+      "grad_norm": 0.0,
+      "learning_rate": 8.355347122941538e-06,
+      "loss": 0.4353,
+      "step": 4417
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 0.0,
+      "learning_rate": 8.351252638464587e-06,
+      "loss": 0.2659,
+      "step": 4418
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 0.0,
+      "learning_rate": 8.34715843810096e-06,
+      "loss": 0.3267,
+      "step": 4419
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 0.0,
+      "learning_rate": 8.343064522556177e-06,
+      "loss": 0.2956,
+      "step": 4420
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 0.0,
+      "learning_rate": 8.338970892535705e-06,
+      "loss": 0.3805,
+      "step": 4421
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 0.0,
+      "learning_rate": 8.334877548744954e-06,
+      "loss": 0.3413,
+      "step": 4422
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 0.0,
+      "learning_rate": 8.330784491889303e-06,
+      "loss": 0.3594,
+      "step": 4423
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 0.0,
+      "learning_rate": 8.326691722674063e-06,
+      "loss": 0.2569,
+      "step": 4424
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 0.0,
+      "learning_rate": 8.322599241804502e-06,
+      "loss": 0.3653,
+      "step": 4425
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 0.0,
+      "learning_rate": 8.318507049985844e-06,
+      "loss": 0.3746,
+      "step": 4426
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.314415147923254e-06,
+      "loss": 0.408,
+      "step": 4427
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.310323536321859e-06,
+      "loss": 0.2635,
+      "step": 4428
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.306232215886722e-06,
+      "loss": 0.2599,
+      "step": 4429
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.302141187322858e-06,
+      "loss": 0.3382,
+      "step": 4430
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.298050451335249e-06,
+      "loss": 0.289,
+      "step": 4431
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.293960008628804e-06,
+      "loss": 0.2619,
+      "step": 4432
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.289869859908389e-06,
+      "loss": 0.2975,
+      "step": 4433
+    },
+    {
+      "epoch": 5.11,
+      "grad_norm": 0.0,
+      "learning_rate": 8.285780005878829e-06,
+      "loss": 0.4375,
+      "step": 4434
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.281690447244887e-06,
+      "loss": 0.26,
+      "step": 4435
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.277601184711277e-06,
+      "loss": 0.2841,
+      "step": 4436
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.273512218982666e-06,
+      "loss": 0.325,
+      "step": 4437
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.26942355076367e-06,
+      "loss": 0.3368,
+      "step": 4438
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.265335180758843e-06,
+      "loss": 0.3528,
+      "step": 4439
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.261247109672705e-06,
+      "loss": 0.3219,
+      "step": 4440
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.257159338209708e-06,
+      "loss": 0.2878,
+      "step": 4441
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.253071867074267e-06,
+      "loss": 0.2906,
+      "step": 4442
+    },
+    {
+      "epoch": 5.12,
+      "grad_norm": 0.0,
+      "learning_rate": 8.248984696970732e-06,
+      "loss": 0.4186,
+      "step": 4443
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.0,
+      "learning_rate": 8.244897828603406e-06,
+      "loss": 0.2679,
+      "step": 4444
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.0,
+      "learning_rate": 8.24081126267655e-06,
+      "loss": 0.3157,
+      "step": 4445
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.0,
+      "learning_rate": 8.236724999894355e-06,
+      "loss": 0.3251,
+      "step": 4446
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.0,
+      "learning_rate": 8.23263904096097e-06,
+      "loss": 0.3355,
+      "step": 4447
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.0,
+      "learning_rate": 8.228553386580495e-06,
+      "loss": 0.3807,
+      "step": 4448
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.0,
+      "learning_rate": 8.224468037456969e-06,
+      "loss": 0.2191,
+      "step": 4449
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.0,
+      "learning_rate": 8.220382994294383e-06,
+      "loss": 0.3535,
+      "step": 4450
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.0,
+      "learning_rate": 8.216298257796677e-06,
+      "loss": 0.3834,
+      "step": 4451
+    },
+    {
+      "epoch": 5.13,
+      "grad_norm": 0.0,
+      "learning_rate": 8.21221382866773e-06,
+      "loss": 0.3715,
+      "step": 4452
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 0.0,
+      "learning_rate": 8.208129707611381e-06,
+      "loss": 0.3738,
+      "step": 4453
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 0.0,
+      "learning_rate": 8.204045895331406e-06,
+      "loss": 0.2731,
+      "step": 4454
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 0.0,
+      "learning_rate": 8.199962392531526e-06,
+      "loss": 0.2926,
+      "step": 4455
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 0.0,
+      "learning_rate": 8.195879199915417e-06,
+      "loss": 0.2714,
+      "step": 4456
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 0.0,
+      "learning_rate": 8.191796318186701e-06,
+      "loss": 0.3061,
+      "step": 4457
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 0.0,
+      "learning_rate": 8.187713748048935e-06,
+      "loss": 0.3276,
+      "step": 4458
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 0.0,
+      "learning_rate": 8.183631490205636e-06,
+      "loss": 0.3261,
+      "step": 4459
+    },
+    {
+      "epoch": 5.14,
+      "grad_norm": 0.0,
+      "learning_rate": 8.179549545360265e-06,
+      "loss": 0.3146,
+      "step": 4460
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 0.0,
+      "learning_rate": 8.175467914216216e-06,
+      "loss": 0.4106,
+      "step": 4461
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 0.0,
+      "learning_rate": 8.171386597476845e-06,
+      "loss": 0.2859,
+      "step": 4462
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 0.0,
+      "learning_rate": 8.167305595845445e-06,
+      "loss": 0.3165,
+      "step": 4463
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 0.0,
+      "learning_rate": 8.163224910025259e-06,
+      "loss": 0.3101,
+      "step": 4464
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 0.0,
+      "learning_rate": 8.159144540719475e-06,
+      "loss": 0.2641,
+      "step": 4465
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 0.0,
+      "learning_rate": 8.15506448863122e-06,
+      "loss": 0.35,
+      "step": 4466
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 0.0,
+      "learning_rate": 8.150984754463578e-06,
+      "loss": 0.3255,
+      "step": 4467
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 0.0,
+      "learning_rate": 8.14690533891957e-06,
+      "loss": 0.4204,
+      "step": 4468
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 0.0,
+      "learning_rate": 8.142826242702157e-06,
+      "loss": 0.3559,
+      "step": 4469
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.0,
+      "learning_rate": 8.138747466514259e-06,
+      "loss": 0.3804,
+      "step": 4470
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.0,
+      "learning_rate": 8.134669011058734e-06,
+      "loss": 0.2832,
+      "step": 4471
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.0,
+      "learning_rate": 8.130590877038379e-06,
+      "loss": 0.2747,
+      "step": 4472
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.0,
+      "learning_rate": 8.126513065155944e-06,
+      "loss": 0.295,
+      "step": 4473
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.0,
+      "learning_rate": 8.122435576114124e-06,
+      "loss": 0.2391,
+      "step": 4474
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.0,
+      "learning_rate": 8.118358410615545e-06,
+      "loss": 0.2497,
+      "step": 4475
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.0,
+      "learning_rate": 8.114281569362799e-06,
+      "loss": 0.3379,
+      "step": 4476
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.0,
+      "learning_rate": 8.110205053058399e-06,
+      "loss": 0.2719,
+      "step": 4477
+    },
+    {
+      "epoch": 5.16,
+      "grad_norm": 0.0,
+      "learning_rate": 8.106128862404825e-06,
+      "loss": 0.4063,
+      "step": 4478
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 0.0,
+      "learning_rate": 8.10205299810448e-06,
+      "loss": 0.3113,
+      "step": 4479
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 0.0,
+      "learning_rate": 8.09797746085972e-06,
+      "loss": 0.3072,
+      "step": 4480
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 0.0,
+      "learning_rate": 8.093902251372854e-06,
+      "loss": 0.2526,
+      "step": 4481
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 0.0,
+      "learning_rate": 8.089827370346114e-06,
+      "loss": 0.263,
+      "step": 4482
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 0.0,
+      "learning_rate": 8.08575281848169e-06,
+      "loss": 0.2847,
+      "step": 4483
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 0.0,
+      "learning_rate": 8.081678596481715e-06,
+      "loss": 0.3568,
+      "step": 4484
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 0.0,
+      "learning_rate": 8.077604705048258e-06,
+      "loss": 0.4295,
+      "step": 4485
+    },
+    {
+      "epoch": 5.17,
+      "grad_norm": 0.0,
+      "learning_rate": 8.073531144883334e-06,
+      "loss": 0.3378,
+      "step": 4486
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 0.0,
+      "learning_rate": 8.069457916688907e-06,
+      "loss": 0.2768,
+      "step": 4487
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 0.0,
+      "learning_rate": 8.065385021166872e-06,
+      "loss": 0.2713,
+      "step": 4488
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 0.0,
+      "learning_rate": 8.061312459019078e-06,
+      "loss": 0.3768,
+      "step": 4489
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 0.0,
+      "learning_rate": 8.057240230947313e-06,
+      "loss": 0.4049,
+      "step": 4490
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 0.0,
+      "learning_rate": 8.0531683376533e-06,
+      "loss": 0.2794,
+      "step": 4491
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 0.0,
+      "learning_rate": 8.04909677983872e-06,
+      "loss": 0.4143,
+      "step": 4492
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 0.0,
+      "learning_rate": 8.04502555820518e-06,
+      "loss": 0.3819,
+      "step": 4493
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 0.0,
+      "learning_rate": 8.04095467345423e-06,
+      "loss": 0.2532,
+      "step": 4494
+    },
+    {
+      "epoch": 5.18,
+      "grad_norm": 0.0,
+      "learning_rate": 8.036884126287382e-06,
+      "loss": 0.3341,
+      "step": 4495
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.0,
+      "learning_rate": 8.032813917406067e-06,
+      "loss": 0.3351,
+      "step": 4496
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.0,
+      "learning_rate": 8.028744047511666e-06,
+      "loss": 0.3437,
+      "step": 4497
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.0,
+      "learning_rate": 8.024674517305505e-06,
+      "loss": 0.3446,
+      "step": 4498
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.0,
+      "learning_rate": 8.020605327488846e-06,
+      "loss": 0.2984,
+      "step": 4499
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.0,
+      "learning_rate": 8.016536478762895e-06,
+      "loss": 0.2509,
+      "step": 4500
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.0,
+      "learning_rate": 8.012467971828803e-06,
+      "loss": 0.3212,
+      "step": 4501
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.0,
+      "learning_rate": 8.008399807387652e-06,
+      "loss": 0.3002,
+      "step": 4502
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.0,
+      "learning_rate": 8.004331986140474e-06,
+      "loss": 0.3343,
+      "step": 4503
+    },
+    {
+      "epoch": 5.19,
+      "grad_norm": 0.0,
+      "learning_rate": 8.000264508788241e-06,
+      "loss": 0.2946,
+      "step": 4504
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.0,
+      "learning_rate": 7.996197376031857e-06,
+      "loss": 0.3234,
+      "step": 4505
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.0,
+      "learning_rate": 7.99213058857218e-06,
+      "loss": 0.3518,
+      "step": 4506
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.0,
+      "learning_rate": 7.988064147110001e-06,
+      "loss": 0.3565,
+      "step": 4507
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.0,
+      "learning_rate": 7.983998052346045e-06,
+      "loss": 0.2792,
+      "step": 4508
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.0,
+      "learning_rate": 7.979932304980993e-06,
+      "loss": 0.422,
+      "step": 4509
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.0,
+      "learning_rate": 7.975866905715452e-06,
+      "loss": 0.3248,
+      "step": 4510
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.0,
+      "learning_rate": 7.97180185524998e-06,
+      "loss": 0.3242,
+      "step": 4511
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.0,
+      "learning_rate": 7.967737154285065e-06,
+      "loss": 0.3146,
+      "step": 4512
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 0.0,
+      "learning_rate": 7.963672803521136e-06,
+      "loss": 0.3074,
+      "step": 4513
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 0.0,
+      "learning_rate": 7.959608803658575e-06,
+      "loss": 0.2994,
+      "step": 4514
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 0.0,
+      "learning_rate": 7.955545155397684e-06,
+      "loss": 0.3379,
+      "step": 4515
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 0.0,
+      "learning_rate": 7.951481859438717e-06,
+      "loss": 0.2387,
+      "step": 4516
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 0.0,
+      "learning_rate": 7.947418916481867e-06,
+      "loss": 0.3427,
+      "step": 4517
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 0.0,
+      "learning_rate": 7.943356327227258e-06,
+      "loss": 0.2799,
+      "step": 4518
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 0.0,
+      "learning_rate": 7.939294092374961e-06,
+      "loss": 0.2462,
+      "step": 4519
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 0.0,
+      "learning_rate": 7.935232212624984e-06,
+      "loss": 0.2629,
+      "step": 4520
+    },
+    {
+      "epoch": 5.21,
+      "grad_norm": 0.0,
+      "learning_rate": 7.931170688677272e-06,
+      "loss": 0.384,
+      "step": 4521
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 0.0,
+      "learning_rate": 7.927109521231708e-06,
+      "loss": 0.3392,
+      "step": 4522
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 0.0,
+      "learning_rate": 7.923048710988119e-06,
+      "loss": 0.3077,
+      "step": 4523
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 0.0,
+      "learning_rate": 7.918988258646261e-06,
+      "loss": 0.4056,
+      "step": 4524
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 0.0,
+      "learning_rate": 7.914928164905844e-06,
+      "loss": 0.2765,
+      "step": 4525
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 0.0,
+      "learning_rate": 7.910868430466497e-06,
+      "loss": 0.3545,
+      "step": 4526
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 0.0,
+      "learning_rate": 7.906809056027797e-06,
+      "loss": 0.2433,
+      "step": 4527
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 0.0,
+      "learning_rate": 7.902750042289267e-06,
+      "loss": 0.3116,
+      "step": 4528
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 0.0,
+      "learning_rate": 7.898691389950348e-06,
+      "loss": 0.3459,
+      "step": 4529
+    },
+    {
+      "epoch": 5.22,
+      "grad_norm": 0.0,
+      "learning_rate": 7.894633099710434e-06,
+      "loss": 0.2621,
+      "step": 4530
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 0.0,
+      "learning_rate": 7.890575172268858e-06,
+      "loss": 0.3697,
+      "step": 4531
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 0.0,
+      "learning_rate": 7.886517608324878e-06,
+      "loss": 0.3621,
+      "step": 4532
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 0.0,
+      "learning_rate": 7.882460408577696e-06,
+      "loss": 0.2993,
+      "step": 4533
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 0.0,
+      "learning_rate": 7.878403573726458e-06,
+      "loss": 0.3563,
+      "step": 4534
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 0.0,
+      "learning_rate": 7.874347104470234e-06,
+      "loss": 0.3712,
+      "step": 4535
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 0.0,
+      "learning_rate": 7.87029100150804e-06,
+      "loss": 0.2987,
+      "step": 4536
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 0.0,
+      "learning_rate": 7.866235265538834e-06,
+      "loss": 0.3336,
+      "step": 4537
+    },
+    {
+      "epoch": 5.23,
+      "grad_norm": 0.0,
+      "learning_rate": 7.862179897261488e-06,
+      "loss": 0.3268,
+      "step": 4538
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 0.0,
+      "learning_rate": 7.858124897374837e-06,
+      "loss": 0.3016,
+      "step": 4539
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 0.0,
+      "learning_rate": 7.854070266577643e-06,
+      "loss": 0.3242,
+      "step": 4540
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 0.0,
+      "learning_rate": 7.850016005568593e-06,
+      "loss": 0.2858,
+      "step": 4541
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 0.0,
+      "learning_rate": 7.845962115046327e-06,
+      "loss": 0.2877,
+      "step": 4542
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 0.0,
+      "learning_rate": 7.841908595709416e-06,
+      "loss": 0.3549,
+      "step": 4543
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 0.0,
+      "learning_rate": 7.83785544825636e-06,
+      "loss": 0.3239,
+      "step": 4544
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 0.0,
+      "learning_rate": 7.833802673385604e-06,
+      "loss": 0.2715,
+      "step": 4545
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 0.0,
+      "learning_rate": 7.829750271795523e-06,
+      "loss": 0.3507,
+      "step": 4546
+    },
+    {
+      "epoch": 5.24,
+      "grad_norm": 0.0,
+      "learning_rate": 7.825698244184432e-06,
+      "loss": 0.2517,
+      "step": 4547
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 0.0,
+      "learning_rate": 7.821646591250579e-06,
+      "loss": 0.2908,
+      "step": 4548
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 0.0,
+      "learning_rate": 7.817595313692143e-06,
+      "loss": 0.4161,
+      "step": 4549
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 0.0,
+      "learning_rate": 7.81354441220725e-06,
+      "loss": 0.3195,
+      "step": 4550
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 0.0,
+      "learning_rate": 7.80949388749395e-06,
+      "loss": 0.3509,
+      "step": 4551
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 0.0,
+      "learning_rate": 7.805443740250232e-06,
+      "loss": 0.3374,
+      "step": 4552
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 0.0,
+      "learning_rate": 7.801393971174024e-06,
+      "loss": 0.2814,
+      "step": 4553
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 0.0,
+      "learning_rate": 7.797344580963183e-06,
+      "loss": 0.2564,
+      "step": 4554
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 0.0,
+      "learning_rate": 7.7932955703155e-06,
+      "loss": 0.3611,
+      "step": 4555
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 0.0,
+      "learning_rate": 7.789246939928706e-06,
+      "loss": 0.2767,
+      "step": 4556
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 0.0,
+      "learning_rate": 7.785198690500468e-06,
+      "loss": 0.2769,
+      "step": 4557
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 0.0,
+      "learning_rate": 7.781150822728374e-06,
+      "loss": 0.3019,
+      "step": 4558
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 0.0,
+      "learning_rate": 7.777103337309963e-06,
+      "loss": 0.3358,
+      "step": 4559
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 0.0,
+      "learning_rate": 7.773056234942696e-06,
+      "loss": 0.2799,
+      "step": 4560
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 0.0,
+      "learning_rate": 7.76900951632398e-06,
+      "loss": 0.4081,
+      "step": 4561
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 0.0,
+      "learning_rate": 7.76496318215114e-06,
+      "loss": 0.3047,
+      "step": 4562
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 0.0,
+      "learning_rate": 7.760917233121443e-06,
+      "loss": 0.3177,
+      "step": 4563
+    },
+    {
+      "epoch": 5.26,
+      "grad_norm": 0.0,
+      "learning_rate": 7.7568716699321e-06,
+      "loss": 0.3971,
+      "step": 4564
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 0.0,
+      "learning_rate": 7.752826493280236e-06,
+      "loss": 0.2205,
+      "step": 4565
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 0.0,
+      "learning_rate": 7.748781703862918e-06,
+      "loss": 0.2342,
+      "step": 4566
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 0.0,
+      "learning_rate": 7.744737302377156e-06,
+      "loss": 0.2964,
+      "step": 4567
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 0.0,
+      "learning_rate": 7.740693289519876e-06,
+      "loss": 0.3024,
+      "step": 4568
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 0.0,
+      "learning_rate": 7.736649665987944e-06,
+      "loss": 0.31,
+      "step": 4569
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 0.0,
+      "learning_rate": 7.73260643247817e-06,
+      "loss": 0.3644,
+      "step": 4570
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 0.0,
+      "learning_rate": 7.728563589687275e-06,
+      "loss": 0.3674,
+      "step": 4571
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 0.0,
+      "learning_rate": 7.724521138311933e-06,
+      "loss": 0.2559,
+      "step": 4572
+    },
+    {
+      "epoch": 5.27,
+      "grad_norm": 0.0,
+      "learning_rate": 7.720479079048741e-06,
+      "loss": 0.3007,
+      "step": 4573
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 0.0,
+      "learning_rate": 7.716437412594225e-06,
+      "loss": 0.2886,
+      "step": 4574
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 0.0,
+      "learning_rate": 7.712396139644854e-06,
+      "loss": 0.265,
+      "step": 4575
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 0.0,
+      "learning_rate": 7.70835526089702e-06,
+      "loss": 0.2755,
+      "step": 4576
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 0.0,
+      "learning_rate": 7.704314777047045e-06,
+      "loss": 0.2827,
+      "step": 4577
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 0.0,
+      "learning_rate": 7.700274688791196e-06,
+      "loss": 0.3093,
+      "step": 4578
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 0.0,
+      "learning_rate": 7.696234996825663e-06,
+      "loss": 0.3479,
+      "step": 4579
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 0.0,
+      "learning_rate": 7.69219570184656e-06,
+      "loss": 0.3203,
+      "step": 4580
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 0.0,
+      "learning_rate": 7.688156804549956e-06,
+      "loss": 0.3711,
+      "step": 4581
+    },
+    {
+      "epoch": 5.28,
+      "grad_norm": 0.0,
+      "learning_rate": 7.684118305631823e-06,
+      "loss": 0.3289,
+      "step": 4582
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 0.0,
+      "learning_rate": 7.680080205788086e-06,
+      "loss": 0.3742,
+      "step": 4583
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 0.0,
+      "learning_rate": 7.676042505714591e-06,
+      "loss": 0.3404,
+      "step": 4584
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 0.0,
+      "learning_rate": 7.672005206107115e-06,
+      "loss": 0.3341,
+      "step": 4585
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 0.0,
+      "learning_rate": 7.667968307661371e-06,
+      "loss": 0.2714,
+      "step": 4586
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 0.0,
+      "learning_rate": 7.663931811073003e-06,
+      "loss": 0.2431,
+      "step": 4587
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 0.0,
+      "learning_rate": 7.659895717037575e-06,
+      "loss": 0.3605,
+      "step": 4588
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 0.0,
+      "learning_rate": 7.655860026250596e-06,
+      "loss": 0.3259,
+      "step": 4589
+    },
+    {
+      "epoch": 5.29,
+      "grad_norm": 0.0,
+      "learning_rate": 7.6518247394075e-06,
+      "loss": 0.3408,
+      "step": 4590
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.0,
+      "learning_rate": 7.647789857203644e-06,
+      "loss": 0.3574,
+      "step": 4591
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.0,
+      "learning_rate": 7.643755380334327e-06,
+      "loss": 0.3169,
+      "step": 4592
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.0,
+      "learning_rate": 7.639721309494771e-06,
+      "loss": 0.2752,
+      "step": 4593
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.0,
+      "learning_rate": 7.635687645380134e-06,
+      "loss": 0.3839,
+      "step": 4594
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.0,
+      "learning_rate": 7.631654388685496e-06,
+      "loss": 0.2786,
+      "step": 4595
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.0,
+      "learning_rate": 7.6276215401058675e-06,
+      "loss": 0.2833,
+      "step": 4596
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.0,
+      "learning_rate": 7.623589100336199e-06,
+      "loss": 0.2819,
+      "step": 4597
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.0,
+      "learning_rate": 7.619557070071359e-06,
+      "loss": 0.3123,
+      "step": 4598
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 0.0,
+      "learning_rate": 7.6155254500061494e-06,
+      "loss": 0.3019,
+      "step": 4599
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 0.0,
+      "learning_rate": 7.611494240835308e-06,
+      "loss": 0.3731,
+      "step": 4600
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 0.0,
+      "learning_rate": 7.607463443253487e-06,
+      "loss": 0.3489,
+      "step": 4601
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 0.0,
+      "learning_rate": 7.603433057955279e-06,
+      "loss": 0.3085,
+      "step": 4602
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 0.0,
+      "learning_rate": 7.599403085635208e-06,
+      "loss": 0.294,
+      "step": 4603
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 0.0,
+      "learning_rate": 7.5953735269877175e-06,
+      "loss": 0.3318,
+      "step": 4604
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 0.0,
+      "learning_rate": 7.591344382707181e-06,
+      "loss": 0.2933,
+      "step": 4605
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 0.0,
+      "learning_rate": 7.587315653487911e-06,
+      "loss": 0.362,
+      "step": 4606
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 0.0,
+      "learning_rate": 7.583287340024132e-06,
+      "loss": 0.2723,
+      "step": 4607
+    },
+    {
+      "epoch": 5.31,
+      "grad_norm": 0.0,
+      "learning_rate": 7.579259443010014e-06,
+      "loss": 0.3446,
+      "step": 4608
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 0.0,
+      "learning_rate": 7.575231963139645e-06,
+      "loss": 0.3575,
+      "step": 4609
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 0.0,
+      "learning_rate": 7.5712049011070365e-06,
+      "loss": 0.4244,
+      "step": 4610
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 0.0,
+      "learning_rate": 7.567178257606147e-06,
+      "loss": 0.3203,
+      "step": 4611
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 0.0,
+      "learning_rate": 7.563152033330841e-06,
+      "loss": 0.2143,
+      "step": 4612
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 0.0,
+      "learning_rate": 7.559126228974921e-06,
+      "loss": 0.3619,
+      "step": 4613
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 0.0,
+      "learning_rate": 7.555100845232124e-06,
+      "loss": 0.3404,
+      "step": 4614
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 0.0,
+      "learning_rate": 7.551075882796098e-06,
+      "loss": 0.3224,
+      "step": 4615
+    },
+    {
+      "epoch": 5.32,
+      "grad_norm": 0.0,
+      "learning_rate": 7.547051342360431e-06,
+      "loss": 0.4012,
+      "step": 4616
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 0.0,
+      "learning_rate": 7.543027224618637e-06,
+      "loss": 0.2585,
+      "step": 4617
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 0.0,
+      "learning_rate": 7.539003530264151e-06,
+      "loss": 0.325,
+      "step": 4618
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 0.0,
+      "learning_rate": 7.534980259990341e-06,
+      "loss": 0.3869,
+      "step": 4619
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 0.0,
+      "learning_rate": 7.530957414490501e-06,
+      "loss": 0.3165,
+      "step": 4620
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 0.0,
+      "learning_rate": 7.5269349944578454e-06,
+      "loss": 0.3961,
+      "step": 4621
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 0.0,
+      "learning_rate": 7.522913000585527e-06,
+      "loss": 0.321,
+      "step": 4622
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 0.0,
+      "learning_rate": 7.518891433566617e-06,
+      "loss": 0.3058,
+      "step": 4623
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 0.0,
+      "learning_rate": 7.5148702940941094e-06,
+      "loss": 0.3089,
+      "step": 4624
+    },
+    {
+      "epoch": 5.33,
+      "grad_norm": 0.0,
+      "learning_rate": 7.510849582860936e-06,
+      "loss": 0.2798,
+      "step": 4625
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 0.0,
+      "learning_rate": 7.506829300559949e-06,
+      "loss": 0.2923,
+      "step": 4626
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 0.0,
+      "learning_rate": 7.50280944788392e-06,
+      "loss": 0.3008,
+      "step": 4627
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 0.0,
+      "learning_rate": 7.4987900255255575e-06,
+      "loss": 0.3705,
+      "step": 4628
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 0.0,
+      "learning_rate": 7.49477103417749e-06,
+      "loss": 0.26,
+      "step": 4629
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 0.0,
+      "learning_rate": 7.490752474532277e-06,
+      "loss": 0.321,
+      "step": 4630
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 0.0,
+      "learning_rate": 7.486734347282394e-06,
+      "loss": 0.2845,
+      "step": 4631
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 0.0,
+      "learning_rate": 7.482716653120248e-06,
+      "loss": 0.3641,
+      "step": 4632
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 0.0,
+      "learning_rate": 7.478699392738176e-06,
+      "loss": 0.302,
+      "step": 4633
+    },
+    {
+      "epoch": 5.34,
+      "grad_norm": 0.0,
+      "learning_rate": 7.474682566828431e-06,
+      "loss": 0.2839,
+      "step": 4634
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 0.0,
+      "learning_rate": 7.470666176083193e-06,
+      "loss": 0.3327,
+      "step": 4635
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 0.0,
+      "learning_rate": 7.466650221194577e-06,
+      "loss": 0.3064,
+      "step": 4636
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 0.0,
+      "learning_rate": 7.462634702854607e-06,
+      "loss": 0.2978,
+      "step": 4637
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 0.0,
+      "learning_rate": 7.458619621755243e-06,
+      "loss": 0.342,
+      "step": 4638
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 0.0,
+      "learning_rate": 7.454604978588373e-06,
+      "loss": 0.3391,
+      "step": 4639
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 0.0,
+      "learning_rate": 7.45059077404579e-06,
+      "loss": 0.3825,
+      "step": 4640
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 0.0,
+      "learning_rate": 7.446577008819238e-06,
+      "loss": 0.3162,
+      "step": 4641
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 0.0,
+      "learning_rate": 7.442563683600364e-06,
+      "loss": 0.3432,
+      "step": 4642
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 0.0,
+      "learning_rate": 7.438550799080746e-06,
+      "loss": 0.3016,
+      "step": 4643
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 0.0,
+      "learning_rate": 7.434538355951896e-06,
+      "loss": 0.3474,
+      "step": 4644
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 0.0,
+      "learning_rate": 7.430526354905232e-06,
+      "loss": 0.4564,
+      "step": 4645
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 0.0,
+      "learning_rate": 7.4265147966321075e-06,
+      "loss": 0.3383,
+      "step": 4646
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 0.0,
+      "learning_rate": 7.4225036818238e-06,
+      "loss": 0.3017,
+      "step": 4647
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 0.0,
+      "learning_rate": 7.418493011171506e-06,
+      "loss": 0.24,
+      "step": 4648
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 0.0,
+      "learning_rate": 7.414482785366342e-06,
+      "loss": 0.2629,
+      "step": 4649
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 0.0,
+      "learning_rate": 7.4104730050993636e-06,
+      "loss": 0.2949,
+      "step": 4650
+    },
+    {
+      "epoch": 5.36,
+      "grad_norm": 0.0,
+      "learning_rate": 7.40646367106153e-06,
+      "loss": 0.3319,
+      "step": 4651
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 0.0,
+      "learning_rate": 7.4024547839437335e-06,
+      "loss": 0.2663,
+      "step": 4652
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 0.0,
+      "learning_rate": 7.398446344436795e-06,
+      "loss": 0.2465,
+      "step": 4653
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 0.0,
+      "learning_rate": 7.394438353231444e-06,
+      "loss": 0.3456,
+      "step": 4654
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 0.0,
+      "learning_rate": 7.390430811018343e-06,
+      "loss": 0.2642,
+      "step": 4655
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 0.0,
+      "learning_rate": 7.386423718488078e-06,
+      "loss": 0.3124,
+      "step": 4656
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 0.0,
+      "learning_rate": 7.382417076331148e-06,
+      "loss": 0.4527,
+      "step": 4657
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 0.0,
+      "learning_rate": 7.378410885237986e-06,
+      "loss": 0.3789,
+      "step": 4658
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 0.0,
+      "learning_rate": 7.3744051458989395e-06,
+      "loss": 0.4148,
+      "step": 4659
+    },
+    {
+      "epoch": 5.37,
+      "grad_norm": 0.0,
+      "learning_rate": 7.370399859004276e-06,
+      "loss": 0.332,
+      "step": 4660
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 0.0,
+      "learning_rate": 7.3663950252441974e-06,
+      "loss": 0.3032,
+      "step": 4661
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 0.0,
+      "learning_rate": 7.362390645308817e-06,
+      "loss": 0.3208,
+      "step": 4662
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 0.0,
+      "learning_rate": 7.358386719888166e-06,
+      "loss": 0.2824,
+      "step": 4663
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 0.0,
+      "learning_rate": 7.354383249672212e-06,
+      "loss": 0.2824,
+      "step": 4664
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 0.0,
+      "learning_rate": 7.35038023535083e-06,
+      "loss": 0.2821,
+      "step": 4665
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 0.0,
+      "learning_rate": 7.3463776776138294e-06,
+      "loss": 0.2557,
+      "step": 4666
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 0.0,
+      "learning_rate": 7.342375577150928e-06,
+      "loss": 0.3087,
+      "step": 4667
+    },
+    {
+      "epoch": 5.38,
+      "grad_norm": 0.0,
+      "learning_rate": 7.338373934651769e-06,
+      "loss": 0.2937,
+      "step": 4668
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 0.0,
+      "learning_rate": 7.334372750805926e-06,
+      "loss": 0.2705,
+      "step": 4669
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 0.0,
+      "learning_rate": 7.330372026302879e-06,
+      "loss": 0.3134,
+      "step": 4670
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 0.0,
+      "learning_rate": 7.326371761832035e-06,
+      "loss": 0.3085,
+      "step": 4671
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 0.0,
+      "learning_rate": 7.322371958082728e-06,
+      "loss": 0.419,
+      "step": 4672
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 0.0,
+      "learning_rate": 7.318372615744207e-06,
+      "loss": 0.3363,
+      "step": 4673
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 0.0,
+      "learning_rate": 7.314373735505633e-06,
+      "loss": 0.2755,
+      "step": 4674
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 0.0,
+      "learning_rate": 7.310375318056107e-06,
+      "loss": 0.3528,
+      "step": 4675
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 0.0,
+      "learning_rate": 7.306377364084629e-06,
+      "loss": 0.3022,
+      "step": 4676
+    },
+    {
+      "epoch": 5.39,
+      "grad_norm": 0.0,
+      "learning_rate": 7.30237987428014e-06,
+      "loss": 0.4227,
+      "step": 4677
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 0.0,
+      "learning_rate": 7.298382849331482e-06,
+      "loss": 0.4126,
+      "step": 4678
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 0.0,
+      "learning_rate": 7.294386289927425e-06,
+      "loss": 0.2986,
+      "step": 4679
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 0.0,
+      "learning_rate": 7.290390196756667e-06,
+      "loss": 0.4435,
+      "step": 4680
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 0.0,
+      "learning_rate": 7.286394570507811e-06,
+      "loss": 0.2691,
+      "step": 4681
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 0.0,
+      "learning_rate": 7.282399411869383e-06,
+      "loss": 0.3174,
+      "step": 4682
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 0.0,
+      "learning_rate": 7.278404721529843e-06,
+      "loss": 0.3198,
+      "step": 4683
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 0.0,
+      "learning_rate": 7.274410500177549e-06,
+      "loss": 0.2521,
+      "step": 4684
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 0.0,
+      "learning_rate": 7.270416748500788e-06,
+      "loss": 0.3592,
+      "step": 4685
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 0.0,
+      "learning_rate": 7.2664234671877755e-06,
+      "loss": 0.3156,
+      "step": 4686
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 0.0,
+      "learning_rate": 7.2624306569266265e-06,
+      "loss": 0.3385,
+      "step": 4687
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 0.0,
+      "learning_rate": 7.2584383184053876e-06,
+      "loss": 0.3819,
+      "step": 4688
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 0.0,
+      "learning_rate": 7.254446452312027e-06,
+      "loss": 0.3456,
+      "step": 4689
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 0.0,
+      "learning_rate": 7.250455059334418e-06,
+      "loss": 0.3216,
+      "step": 4690
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 0.0,
+      "learning_rate": 7.246464140160365e-06,
+      "loss": 0.3814,
+      "step": 4691
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 0.0,
+      "learning_rate": 7.242473695477587e-06,
+      "loss": 0.3421,
+      "step": 4692
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 0.0,
+      "learning_rate": 7.238483725973715e-06,
+      "loss": 0.2465,
+      "step": 4693
+    },
+    {
+      "epoch": 5.41,
+      "grad_norm": 0.0,
+      "learning_rate": 7.234494232336308e-06,
+      "loss": 0.3679,
+      "step": 4694
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 0.0,
+      "learning_rate": 7.230505215252839e-06,
+      "loss": 0.3409,
+      "step": 4695
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 0.0,
+      "learning_rate": 7.226516675410693e-06,
+      "loss": 0.3493,
+      "step": 4696
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 0.0,
+      "learning_rate": 7.2225286134971815e-06,
+      "loss": 0.2737,
+      "step": 4697
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 0.0,
+      "learning_rate": 7.218541030199535e-06,
+      "loss": 0.2693,
+      "step": 4698
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 0.0,
+      "learning_rate": 7.214553926204884e-06,
+      "loss": 0.331,
+      "step": 4699
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 0.0,
+      "learning_rate": 7.210567302200303e-06,
+      "loss": 0.44,
+      "step": 4700
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 0.0,
+      "learning_rate": 7.206581158872761e-06,
+      "loss": 0.3631,
+      "step": 4701
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 0.0,
+      "learning_rate": 7.202595496909156e-06,
+      "loss": 0.5433,
+      "step": 4702
+    },
+    {
+      "epoch": 5.42,
+      "grad_norm": 0.0,
+      "learning_rate": 7.198610316996303e-06,
+      "loss": 0.2532,
+      "step": 4703
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 0.0,
+      "learning_rate": 7.194625619820924e-06,
+      "loss": 0.3607,
+      "step": 4704
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 0.0,
+      "learning_rate": 7.190641406069672e-06,
+      "loss": 0.4287,
+      "step": 4705
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 0.0,
+      "learning_rate": 7.186657676429107e-06,
+      "loss": 0.2681,
+      "step": 4706
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 0.0,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.3242,
+      "step": 4707
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 0.0,
+      "learning_rate": 7.178691672225864e-06,
+      "loss": 0.3287,
+      "step": 4708
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 0.0,
+      "learning_rate": 7.1747093990359e-06,
+      "loss": 0.3363,
+      "step": 4709
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 0.0,
+      "learning_rate": 7.170727612702033e-06,
+      "loss": 0.3129,
+      "step": 4710
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 0.0,
+      "learning_rate": 7.166746313910413e-06,
+      "loss": 0.3073,
+      "step": 4711
+    },
+    {
+      "epoch": 5.43,
+      "grad_norm": 0.0,
+      "learning_rate": 7.162765503347097e-06,
+      "loss": 0.3869,
+      "step": 4712
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.0,
+      "learning_rate": 7.158785181698069e-06,
+      "loss": 0.3106,
+      "step": 4713
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.0,
+      "learning_rate": 7.154805349649212e-06,
+      "loss": 0.4004,
+      "step": 4714
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.0,
+      "learning_rate": 7.150826007886334e-06,
+      "loss": 0.363,
+      "step": 4715
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.0,
+      "learning_rate": 7.146847157095165e-06,
+      "loss": 0.2549,
+      "step": 4716
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.0,
+      "learning_rate": 7.142868797961335e-06,
+      "loss": 0.3393,
+      "step": 4717
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.0,
+      "learning_rate": 7.138890931170402e-06,
+      "loss": 0.3436,
+      "step": 4718
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.0,
+      "learning_rate": 7.134913557407838e-06,
+      "loss": 0.3391,
+      "step": 4719
+    },
+    {
+      "epoch": 5.44,
+      "grad_norm": 0.0,
+      "learning_rate": 7.1309366773590206e-06,
+      "loss": 0.3714,
+      "step": 4720
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 0.0,
+      "learning_rate": 7.126960291709248e-06,
+      "loss": 0.3472,
+      "step": 4721
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 0.0,
+      "learning_rate": 7.122984401143745e-06,
+      "loss": 0.2727,
+      "step": 4722
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 0.0,
+      "learning_rate": 7.119009006347625e-06,
+      "loss": 0.4406,
+      "step": 4723
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 0.0,
+      "learning_rate": 7.11503410800594e-06,
+      "loss": 0.276,
+      "step": 4724
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 0.0,
+      "learning_rate": 7.111059706803649e-06,
+      "loss": 0.3817,
+      "step": 4725
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 0.0,
+      "learning_rate": 7.107085803425615e-06,
+      "loss": 0.4502,
+      "step": 4726
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 0.0,
+      "learning_rate": 7.103112398556631e-06,
+      "loss": 0.3244,
+      "step": 4727
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 0.0,
+      "learning_rate": 7.099139492881397e-06,
+      "loss": 0.4242,
+      "step": 4728
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 0.0,
+      "learning_rate": 7.095167087084518e-06,
+      "loss": 0.3144,
+      "step": 4729
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 0.0,
+      "learning_rate": 7.0911951818505364e-06,
+      "loss": 0.3174,
+      "step": 4730
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 0.0,
+      "learning_rate": 7.087223777863883e-06,
+      "loss": 0.2467,
+      "step": 4731
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 0.0,
+      "learning_rate": 7.083252875808913e-06,
+      "loss": 0.3147,
+      "step": 4732
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 0.0,
+      "learning_rate": 7.079282476369902e-06,
+      "loss": 0.2749,
+      "step": 4733
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 0.0,
+      "learning_rate": 7.0753125802310275e-06,
+      "loss": 0.2974,
+      "step": 4734
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 0.0,
+      "learning_rate": 7.071343188076383e-06,
+      "loss": 0.266,
+      "step": 4735
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 0.0,
+      "learning_rate": 7.067374300589983e-06,
+      "loss": 0.2206,
+      "step": 4736
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 0.0,
+      "learning_rate": 7.0634059184557415e-06,
+      "loss": 0.272,
+      "step": 4737
+    },
+    {
+      "epoch": 5.46,
+      "grad_norm": 0.0,
+      "learning_rate": 7.059438042357499e-06,
+      "loss": 0.2525,
+      "step": 4738
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 0.0,
+      "learning_rate": 7.055470672979003e-06,
+      "loss": 0.2803,
+      "step": 4739
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 0.0,
+      "learning_rate": 7.051503811003908e-06,
+      "loss": 0.2464,
+      "step": 4740
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 0.0,
+      "learning_rate": 7.04753745711579e-06,
+      "loss": 0.2847,
+      "step": 4741
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 0.0,
+      "learning_rate": 7.043571611998137e-06,
+      "loss": 0.4339,
+      "step": 4742
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 0.0,
+      "learning_rate": 7.039606276334336e-06,
+      "loss": 0.2867,
+      "step": 4743
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 0.0,
+      "learning_rate": 7.035641450807709e-06,
+      "loss": 0.2663,
+      "step": 4744
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 0.0,
+      "learning_rate": 7.031677136101471e-06,
+      "loss": 0.3589,
+      "step": 4745
+    },
+    {
+      "epoch": 5.47,
+      "grad_norm": 0.0,
+      "learning_rate": 7.027713332898755e-06,
+      "loss": 0.2666,
+      "step": 4746
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 0.0,
+      "learning_rate": 7.023750041882609e-06,
+      "loss": 0.3779,
+      "step": 4747
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 0.0,
+      "learning_rate": 7.019787263735985e-06,
+      "loss": 0.2885,
+      "step": 4748
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 0.0,
+      "learning_rate": 7.015824999141762e-06,
+      "loss": 0.3358,
+      "step": 4749
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 0.0,
+      "learning_rate": 7.011863248782711e-06,
+      "loss": 0.2913,
+      "step": 4750
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 0.0,
+      "learning_rate": 7.007902013341525e-06,
+      "loss": 0.2901,
+      "step": 4751
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 0.0,
+      "learning_rate": 7.003941293500813e-06,
+      "loss": 0.2607,
+      "step": 4752
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 0.0,
+      "learning_rate": 6.999981089943083e-06,
+      "loss": 0.3261,
+      "step": 4753
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 0.0,
+      "learning_rate": 6.996021403350759e-06,
+      "loss": 0.3708,
+      "step": 4754
+    },
+    {
+      "epoch": 5.48,
+      "grad_norm": 0.0,
+      "learning_rate": 6.992062234406185e-06,
+      "loss": 0.3827,
+      "step": 4755
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 0.0,
+      "learning_rate": 6.9881035837916e-06,
+      "loss": 0.3485,
+      "step": 4756
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 0.0,
+      "learning_rate": 6.984145452189162e-06,
+      "loss": 0.2931,
+      "step": 4757
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 0.0,
+      "learning_rate": 6.980187840280945e-06,
+      "loss": 0.3652,
+      "step": 4758
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 0.0,
+      "learning_rate": 6.976230748748919e-06,
+      "loss": 0.4172,
+      "step": 4759
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 0.0,
+      "learning_rate": 6.972274178274983e-06,
+      "loss": 0.2475,
+      "step": 4760
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 0.0,
+      "learning_rate": 6.968318129540929e-06,
+      "loss": 0.3662,
+      "step": 4761
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 0.0,
+      "learning_rate": 6.964362603228465e-06,
+      "loss": 0.4439,
+      "step": 4762
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 0.0,
+      "learning_rate": 6.960407600019217e-06,
+      "loss": 0.2776,
+      "step": 4763
+    },
+    {
+      "epoch": 5.49,
+      "grad_norm": 0.0,
+      "learning_rate": 6.956453120594709e-06,
+      "loss": 0.3859,
+      "step": 4764
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 0.0,
+      "learning_rate": 6.952499165636379e-06,
+      "loss": 0.4489,
+      "step": 4765
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 0.0,
+      "learning_rate": 6.94854573582558e-06,
+      "loss": 0.4327,
+      "step": 4766
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 0.0,
+      "learning_rate": 6.944592831843566e-06,
+      "loss": 0.3385,
+      "step": 4767
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 0.0,
+      "learning_rate": 6.940640454371502e-06,
+      "loss": 0.3486,
+      "step": 4768
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 0.0,
+      "learning_rate": 6.936688604090472e-06,
+      "loss": 0.4275,
+      "step": 4769
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 0.0,
+      "learning_rate": 6.932737281681454e-06,
+      "loss": 0.2944,
+      "step": 4770
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 0.0,
+      "learning_rate": 6.9287864878253475e-06,
+      "loss": 0.3454,
+      "step": 4771
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 0.0,
+      "learning_rate": 6.9248362232029575e-06,
+      "loss": 0.4015,
+      "step": 4772
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 0.0,
+      "learning_rate": 6.920886488494989e-06,
+      "loss": 0.3654,
+      "step": 4773
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 0.0,
+      "learning_rate": 6.91693728438207e-06,
+      "loss": 0.409,
+      "step": 4774
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 0.0,
+      "learning_rate": 6.9129886115447294e-06,
+      "loss": 0.336,
+      "step": 4775
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 0.0,
+      "learning_rate": 6.9090404706633995e-06,
+      "loss": 0.2689,
+      "step": 4776
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 0.0,
+      "learning_rate": 6.905092862418434e-06,
+      "loss": 0.3314,
+      "step": 4777
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 0.0,
+      "learning_rate": 6.901145787490087e-06,
+      "loss": 0.2864,
+      "step": 4778
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 0.0,
+      "learning_rate": 6.897199246558515e-06,
+      "loss": 0.3105,
+      "step": 4779
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 0.0,
+      "learning_rate": 6.893253240303795e-06,
+      "loss": 0.2969,
+      "step": 4780
+    },
+    {
+      "epoch": 5.51,
+      "grad_norm": 0.0,
+      "learning_rate": 6.889307769405904e-06,
+      "loss": 0.3681,
+      "step": 4781
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 0.0,
+      "learning_rate": 6.885362834544724e-06,
+      "loss": 0.4001,
+      "step": 4782
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 0.0,
+      "learning_rate": 6.881418436400056e-06,
+      "loss": 0.5642,
+      "step": 4783
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 0.0,
+      "learning_rate": 6.877474575651596e-06,
+      "loss": 0.252,
+      "step": 4784
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 0.0,
+      "learning_rate": 6.873531252978959e-06,
+      "loss": 0.3958,
+      "step": 4785
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 0.0,
+      "learning_rate": 6.8695884690616545e-06,
+      "loss": 0.3217,
+      "step": 4786
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 0.0,
+      "learning_rate": 6.865646224579108e-06,
+      "loss": 0.2951,
+      "step": 4787
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 0.0,
+      "learning_rate": 6.861704520210652e-06,
+      "loss": 0.2699,
+      "step": 4788
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 0.0,
+      "learning_rate": 6.857763356635525e-06,
+      "loss": 0.4012,
+      "step": 4789
+    },
+    {
+      "epoch": 5.52,
+      "grad_norm": 0.0,
+      "learning_rate": 6.853822734532864e-06,
+      "loss": 0.5039,
+      "step": 4790
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 0.0,
+      "learning_rate": 6.849882654581726e-06,
+      "loss": 0.3648,
+      "step": 4791
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 0.0,
+      "learning_rate": 6.845943117461069e-06,
+      "loss": 0.3446,
+      "step": 4792
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 0.0,
+      "learning_rate": 6.8420041238497525e-06,
+      "loss": 0.3028,
+      "step": 4793
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 0.0,
+      "learning_rate": 6.8380656744265484e-06,
+      "loss": 0.2723,
+      "step": 4794
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 0.0,
+      "learning_rate": 6.834127769870134e-06,
+      "loss": 0.3553,
+      "step": 4795
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 0.0,
+      "learning_rate": 6.830190410859092e-06,
+      "loss": 0.411,
+      "step": 4796
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 0.0,
+      "learning_rate": 6.826253598071911e-06,
+      "loss": 0.4085,
+      "step": 4797
+    },
+    {
+      "epoch": 5.53,
+      "grad_norm": 0.0,
+      "learning_rate": 6.822317332186983e-06,
+      "loss": 0.4491,
+      "step": 4798
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 0.0,
+      "learning_rate": 6.818381613882612e-06,
+      "loss": 0.229,
+      "step": 4799
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 0.0,
+      "learning_rate": 6.814446443837002e-06,
+      "loss": 0.3195,
+      "step": 4800
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 0.0,
+      "learning_rate": 6.8105118227282605e-06,
+      "loss": 0.3611,
+      "step": 4801
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 0.0,
+      "learning_rate": 6.806577751234413e-06,
+      "loss": 0.268,
+      "step": 4802
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 0.0,
+      "learning_rate": 6.802644230033373e-06,
+      "loss": 0.3112,
+      "step": 4803
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 0.0,
+      "learning_rate": 6.7987112598029695e-06,
+      "loss": 0.2828,
+      "step": 4804
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 0.0,
+      "learning_rate": 6.79477884122094e-06,
+      "loss": 0.3314,
+      "step": 4805
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 0.0,
+      "learning_rate": 6.790846974964914e-06,
+      "loss": 0.3756,
+      "step": 4806
+    },
+    {
+      "epoch": 5.54,
+      "grad_norm": 0.0,
+      "learning_rate": 6.786915661712439e-06,
+      "loss": 0.3061,
+      "step": 4807
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 0.0,
+      "learning_rate": 6.782984902140963e-06,
+      "loss": 0.3622,
+      "step": 4808
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 0.0,
+      "learning_rate": 6.779054696927831e-06,
+      "loss": 0.2478,
+      "step": 4809
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 0.0,
+      "learning_rate": 6.775125046750304e-06,
+      "loss": 0.3197,
+      "step": 4810
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 0.0,
+      "learning_rate": 6.771195952285541e-06,
+      "loss": 0.3087,
+      "step": 4811
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 0.0,
+      "learning_rate": 6.767267414210603e-06,
+      "loss": 0.4744,
+      "step": 4812
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 0.0,
+      "learning_rate": 6.763339433202462e-06,
+      "loss": 0.3386,
+      "step": 4813
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 0.0,
+      "learning_rate": 6.759412009937993e-06,
+      "loss": 0.2785,
+      "step": 4814
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 0.0,
+      "learning_rate": 6.755485145093965e-06,
+      "loss": 0.2995,
+      "step": 4815
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 0.0,
+      "learning_rate": 6.7515588393470634e-06,
+      "loss": 0.3429,
+      "step": 4816
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 0.0,
+      "learning_rate": 6.747633093373873e-06,
+      "loss": 0.2907,
+      "step": 4817
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 0.0,
+      "learning_rate": 6.743707907850874e-06,
+      "loss": 0.2886,
+      "step": 4818
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 0.0,
+      "learning_rate": 6.739783283454469e-06,
+      "loss": 0.2897,
+      "step": 4819
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 0.0,
+      "learning_rate": 6.73585922086094e-06,
+      "loss": 0.373,
+      "step": 4820
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 0.0,
+      "learning_rate": 6.731935720746492e-06,
+      "loss": 0.2571,
+      "step": 4821
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 0.0,
+      "learning_rate": 6.728012783787224e-06,
+      "loss": 0.3692,
+      "step": 4822
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 0.0,
+      "learning_rate": 6.724090410659137e-06,
+      "loss": 0.3312,
+      "step": 4823
+    },
+    {
+      "epoch": 5.56,
+      "grad_norm": 0.0,
+      "learning_rate": 6.7201686020381375e-06,
+      "loss": 0.3979,
+      "step": 4824
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 0.0,
+      "learning_rate": 6.716247358600038e-06,
+      "loss": 0.3295,
+      "step": 4825
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 0.0,
+      "learning_rate": 6.7123266810205446e-06,
+      "loss": 0.288,
+      "step": 4826
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 0.0,
+      "learning_rate": 6.708406569975274e-06,
+      "loss": 0.3296,
+      "step": 4827
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 0.0,
+      "learning_rate": 6.704487026139747e-06,
+      "loss": 0.2896,
+      "step": 4828
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 0.0,
+      "learning_rate": 6.700568050189371e-06,
+      "loss": 0.3209,
+      "step": 4829
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 0.0,
+      "learning_rate": 6.696649642799478e-06,
+      "loss": 0.3549,
+      "step": 4830
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 0.0,
+      "learning_rate": 6.6927318046452825e-06,
+      "loss": 0.3946,
+      "step": 4831
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 0.0,
+      "learning_rate": 6.688814536401918e-06,
+      "loss": 0.3131,
+      "step": 4832
+    },
+    {
+      "epoch": 5.57,
+      "grad_norm": 0.0,
+      "learning_rate": 6.684897838744403e-06,
+      "loss": 0.3191,
+      "step": 4833
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 0.0,
+      "learning_rate": 6.680981712347667e-06,
+      "loss": 0.3018,
+      "step": 4834
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 0.0,
+      "learning_rate": 6.6770661578865444e-06,
+      "loss": 0.4013,
+      "step": 4835
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 0.0,
+      "learning_rate": 6.673151176035762e-06,
+      "loss": 0.427,
+      "step": 4836
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 0.0,
+      "learning_rate": 6.669236767469952e-06,
+      "loss": 0.3371,
+      "step": 4837
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 0.0,
+      "learning_rate": 6.6653229328636536e-06,
+      "loss": 0.2885,
+      "step": 4838
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 0.0,
+      "learning_rate": 6.6614096728912945e-06,
+      "loss": 0.3129,
+      "step": 4839
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 0.0,
+      "learning_rate": 6.657496988227212e-06,
+      "loss": 0.2792,
+      "step": 4840
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 0.0,
+      "learning_rate": 6.6535848795456485e-06,
+      "loss": 0.3188,
+      "step": 4841
+    },
+    {
+      "epoch": 5.58,
+      "grad_norm": 0.0,
+      "learning_rate": 6.649673347520733e-06,
+      "loss": 0.309,
+      "step": 4842
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 0.0,
+      "learning_rate": 6.645762392826509e-06,
+      "loss": 0.3483,
+      "step": 4843
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 0.0,
+      "learning_rate": 6.641852016136916e-06,
+      "loss": 0.3076,
+      "step": 4844
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 0.0,
+      "learning_rate": 6.637942218125786e-06,
+      "loss": 0.271,
+      "step": 4845
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 0.0,
+      "learning_rate": 6.634032999466866e-06,
+      "loss": 0.28,
+      "step": 4846
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 0.0,
+      "learning_rate": 6.630124360833794e-06,
+      "loss": 0.265,
+      "step": 4847
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 0.0,
+      "learning_rate": 6.626216302900099e-06,
+      "loss": 0.2645,
+      "step": 4848
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 0.0,
+      "learning_rate": 6.622308826339237e-06,
+      "loss": 0.3915,
+      "step": 4849
+    },
+    {
+      "epoch": 5.59,
+      "grad_norm": 0.0,
+      "learning_rate": 6.618401931824538e-06,
+      "loss": 0.2992,
+      "step": 4850
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.0,
+      "learning_rate": 6.614495620029238e-06,
+      "loss": 0.4185,
+      "step": 4851
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.0,
+      "learning_rate": 6.610589891626484e-06,
+      "loss": 0.3206,
+      "step": 4852
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.0,
+      "learning_rate": 6.6066847472893046e-06,
+      "loss": 0.3431,
+      "step": 4853
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.0,
+      "learning_rate": 6.602780187690645e-06,
+      "loss": 0.4078,
+      "step": 4854
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.0,
+      "learning_rate": 6.5988762135033405e-06,
+      "loss": 0.2552,
+      "step": 4855
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.0,
+      "learning_rate": 6.594972825400119e-06,
+      "loss": 0.3692,
+      "step": 4856
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.0,
+      "learning_rate": 6.591070024053623e-06,
+      "loss": 0.2954,
+      "step": 4857
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.0,
+      "learning_rate": 6.587167810136388e-06,
+      "loss": 0.32,
+      "step": 4858
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 0.0,
+      "learning_rate": 6.583266184320836e-06,
+      "loss": 0.3183,
+      "step": 4859
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 0.0,
+      "learning_rate": 6.5793651472793065e-06,
+      "loss": 0.3388,
+      "step": 4860
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 0.0,
+      "learning_rate": 6.575464699684029e-06,
+      "loss": 0.305,
+      "step": 4861
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 0.0,
+      "learning_rate": 6.571564842207124e-06,
+      "loss": 0.2834,
+      "step": 4862
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 0.0,
+      "learning_rate": 6.567665575520625e-06,
+      "loss": 0.2691,
+      "step": 4863
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 0.0,
+      "learning_rate": 6.563766900296458e-06,
+      "loss": 0.3611,
+      "step": 4864
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 0.0,
+      "learning_rate": 6.559868817206435e-06,
+      "loss": 0.3155,
+      "step": 4865
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 0.0,
+      "learning_rate": 6.555971326922287e-06,
+      "loss": 0.3502,
+      "step": 4866
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 0.0,
+      "learning_rate": 6.552074430115624e-06,
+      "loss": 0.338,
+      "step": 4867
+    },
+    {
+      "epoch": 5.61,
+      "grad_norm": 0.0,
+      "learning_rate": 6.548178127457974e-06,
+      "loss": 0.3802,
+      "step": 4868
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 0.0,
+      "learning_rate": 6.54428241962074e-06,
+      "loss": 0.4428,
+      "step": 4869
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 0.0,
+      "learning_rate": 6.540387307275233e-06,
+      "loss": 0.2986,
+      "step": 4870
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 0.0,
+      "learning_rate": 6.53649279109267e-06,
+      "loss": 0.3897,
+      "step": 4871
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 0.0,
+      "learning_rate": 6.532598871744149e-06,
+      "loss": 0.2533,
+      "step": 4872
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 0.0,
+      "learning_rate": 6.528705549900674e-06,
+      "loss": 0.3311,
+      "step": 4873
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 0.0,
+      "learning_rate": 6.524812826233151e-06,
+      "loss": 0.3461,
+      "step": 4874
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 0.0,
+      "learning_rate": 6.520920701412371e-06,
+      "loss": 0.3231,
+      "step": 4875
+    },
+    {
+      "epoch": 5.62,
+      "grad_norm": 0.0,
+      "learning_rate": 6.517029176109026e-06,
+      "loss": 0.2984,
+      "step": 4876
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 0.0,
+      "learning_rate": 6.513138250993716e-06,
+      "loss": 0.2976,
+      "step": 4877
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 0.0,
+      "learning_rate": 6.509247926736914e-06,
+      "loss": 0.2975,
+      "step": 4878
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 0.0,
+      "learning_rate": 6.505358204009018e-06,
+      "loss": 0.242,
+      "step": 4879
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 0.0,
+      "learning_rate": 6.501469083480298e-06,
+      "loss": 0.3103,
+      "step": 4880
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 0.0,
+      "learning_rate": 6.4975805658209324e-06,
+      "loss": 0.4048,
+      "step": 4881
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 0.0,
+      "learning_rate": 6.493692651700998e-06,
+      "loss": 0.2647,
+      "step": 4882
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 0.0,
+      "learning_rate": 6.489805341790456e-06,
+      "loss": 0.2884,
+      "step": 4883
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 0.0,
+      "learning_rate": 6.485918636759172e-06,
+      "loss": 0.3384,
+      "step": 4884
+    },
+    {
+      "epoch": 5.63,
+      "grad_norm": 0.0,
+      "learning_rate": 6.4820325372769126e-06,
+      "loss": 0.229,
+      "step": 4885
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 0.0,
+      "learning_rate": 6.478147044013323e-06,
+      "loss": 0.2375,
+      "step": 4886
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 0.0,
+      "learning_rate": 6.474262157637958e-06,
+      "loss": 0.2978,
+      "step": 4887
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 0.0,
+      "learning_rate": 6.470377878820271e-06,
+      "loss": 0.2962,
+      "step": 4888
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 0.0,
+      "learning_rate": 6.4664942082295925e-06,
+      "loss": 0.2784,
+      "step": 4889
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 0.0,
+      "learning_rate": 6.4626111465351674e-06,
+      "loss": 0.3057,
+      "step": 4890
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 0.0,
+      "learning_rate": 6.458728694406124e-06,
+      "loss": 0.4264,
+      "step": 4891
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 0.0,
+      "learning_rate": 6.454846852511489e-06,
+      "loss": 0.3402,
+      "step": 4892
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 0.0,
+      "learning_rate": 6.450965621520187e-06,
+      "loss": 0.3579,
+      "step": 4893
+    },
+    {
+      "epoch": 5.64,
+      "grad_norm": 0.0,
+      "learning_rate": 6.447085002101034e-06,
+      "loss": 0.4901,
+      "step": 4894
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 0.0,
+      "learning_rate": 6.443204994922738e-06,
+      "loss": 0.266,
+      "step": 4895
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 0.0,
+      "learning_rate": 6.439325600653907e-06,
+      "loss": 0.4019,
+      "step": 4896
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 0.0,
+      "learning_rate": 6.4354468199630425e-06,
+      "loss": 0.4198,
+      "step": 4897
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 0.0,
+      "learning_rate": 6.431568653518534e-06,
+      "loss": 0.3507,
+      "step": 4898
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 0.0,
+      "learning_rate": 6.427691101988673e-06,
+      "loss": 0.3377,
+      "step": 4899
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 0.0,
+      "learning_rate": 6.423814166041642e-06,
+      "loss": 0.4129,
+      "step": 4900
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 0.0,
+      "learning_rate": 6.419937846345519e-06,
+      "loss": 0.2847,
+      "step": 4901
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 0.0,
+      "learning_rate": 6.416062143568271e-06,
+      "loss": 0.2472,
+      "step": 4902
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 0.0,
+      "learning_rate": 6.412187058377762e-06,
+      "loss": 0.3392,
+      "step": 4903
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 0.0,
+      "learning_rate": 6.4083125914417545e-06,
+      "loss": 0.2694,
+      "step": 4904
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 0.0,
+      "learning_rate": 6.404438743427894e-06,
+      "loss": 0.3789,
+      "step": 4905
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 0.0,
+      "learning_rate": 6.400565515003722e-06,
+      "loss": 0.2954,
+      "step": 4906
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 0.0,
+      "learning_rate": 6.396692906836686e-06,
+      "loss": 0.2993,
+      "step": 4907
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 0.0,
+      "learning_rate": 6.392820919594109e-06,
+      "loss": 0.3066,
+      "step": 4908
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 0.0,
+      "learning_rate": 6.388949553943216e-06,
+      "loss": 0.4057,
+      "step": 4909
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 0.0,
+      "learning_rate": 6.385078810551124e-06,
+      "loss": 0.264,
+      "step": 4910
+    },
+    {
+      "epoch": 5.66,
+      "grad_norm": 0.0,
+      "learning_rate": 6.3812086900848456e-06,
+      "loss": 0.2866,
+      "step": 4911
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.377339193211274e-06,
+      "loss": 0.358,
+      "step": 4912
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.373470320597212e-06,
+      "loss": 0.2984,
+      "step": 4913
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.369602072909342e-06,
+      "loss": 0.3573,
+      "step": 4914
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.3657344508142495e-06,
+      "loss": 0.4032,
+      "step": 4915
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.3618674549783975e-06,
+      "loss": 0.2993,
+      "step": 4916
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.358001086068153e-06,
+      "loss": 0.342,
+      "step": 4917
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.354135344749774e-06,
+      "loss": 0.3774,
+      "step": 4918
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.350270231689405e-06,
+      "loss": 0.2809,
+      "step": 4919
+    },
+    {
+      "epoch": 5.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.346405747553083e-06,
+      "loss": 0.3299,
+      "step": 4920
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.342541893006746e-06,
+      "loss": 0.3156,
+      "step": 4921
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.33867866871621e-06,
+      "loss": 0.2893,
+      "step": 4922
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.334816075347185e-06,
+      "loss": 0.2904,
+      "step": 4923
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.3309541135652885e-06,
+      "loss": 0.2948,
+      "step": 4924
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.327092784036002e-06,
+      "loss": 0.2685,
+      "step": 4925
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.323232087424724e-06,
+      "loss": 0.3611,
+      "step": 4926
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.319372024396728e-06,
+      "loss": 0.3279,
+      "step": 4927
+    },
+    {
+      "epoch": 5.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.31551259561718e-06,
+      "loss": 0.3017,
+      "step": 4928
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.311653801751145e-06,
+      "loss": 0.3544,
+      "step": 4929
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.307795643463571e-06,
+      "loss": 0.3304,
+      "step": 4930
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.303938121419295e-06,
+      "loss": 0.2789,
+      "step": 4931
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.300081236283053e-06,
+      "loss": 0.3627,
+      "step": 4932
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.296224988719466e-06,
+      "loss": 0.284,
+      "step": 4933
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.2923693793930395e-06,
+      "loss": 0.2658,
+      "step": 4934
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.288514408968183e-06,
+      "loss": 0.3159,
+      "step": 4935
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.2846600781091795e-06,
+      "loss": 0.2848,
+      "step": 4936
+    },
+    {
+      "epoch": 5.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.2808063874802165e-06,
+      "loss": 0.315,
+      "step": 4937
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.0,
+      "learning_rate": 6.276953337745365e-06,
+      "loss": 0.2679,
+      "step": 4938
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.0,
+      "learning_rate": 6.273100929568579e-06,
+      "loss": 0.3308,
+      "step": 4939
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.0,
+      "learning_rate": 6.269249163613713e-06,
+      "loss": 0.3533,
+      "step": 4940
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.0,
+      "learning_rate": 6.265398040544508e-06,
+      "loss": 0.3666,
+      "step": 4941
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.0,
+      "learning_rate": 6.261547561024586e-06,
+      "loss": 0.2871,
+      "step": 4942
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.0,
+      "learning_rate": 6.257697725717469e-06,
+      "loss": 0.3178,
+      "step": 4943
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.0,
+      "learning_rate": 6.2538485352865645e-06,
+      "loss": 0.2709,
+      "step": 4944
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.0,
+      "learning_rate": 6.249999990395162e-06,
+      "loss": 0.3278,
+      "step": 4945
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 0.0,
+      "learning_rate": 6.24615209170645e-06,
+      "loss": 0.299,
+      "step": 4946
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 0.0,
+      "learning_rate": 6.242304839883502e-06,
+      "loss": 0.2782,
+      "step": 4947
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 0.0,
+      "learning_rate": 6.238458235589274e-06,
+      "loss": 0.3459,
+      "step": 4948
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 0.0,
+      "learning_rate": 6.23461227948662e-06,
+      "loss": 0.3508,
+      "step": 4949
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 0.0,
+      "learning_rate": 6.230766972238277e-06,
+      "loss": 0.2334,
+      "step": 4950
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 0.0,
+      "learning_rate": 6.226922314506874e-06,
+      "loss": 0.3226,
+      "step": 4951
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 0.0,
+      "learning_rate": 6.223078306954919e-06,
+      "loss": 0.2886,
+      "step": 4952
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 0.0,
+      "learning_rate": 6.219234950244816e-06,
+      "loss": 0.324,
+      "step": 4953
+    },
+    {
+      "epoch": 5.71,
+      "grad_norm": 0.0,
+      "learning_rate": 6.2153922450388606e-06,
+      "loss": 0.3138,
+      "step": 4954
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 0.0,
+      "learning_rate": 6.211550191999223e-06,
+      "loss": 0.3754,
+      "step": 4955
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 0.0,
+      "learning_rate": 6.207708791787971e-06,
+      "loss": 0.2789,
+      "step": 4956
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 0.0,
+      "learning_rate": 6.2038680450670605e-06,
+      "loss": 0.2925,
+      "step": 4957
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 0.0,
+      "learning_rate": 6.200027952498329e-06,
+      "loss": 0.2897,
+      "step": 4958
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 0.0,
+      "learning_rate": 6.1961885147435015e-06,
+      "loss": 0.3003,
+      "step": 4959
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 0.0,
+      "learning_rate": 6.1923497324642e-06,
+      "loss": 0.3171,
+      "step": 4960
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 0.0,
+      "learning_rate": 6.188511606321918e-06,
+      "loss": 0.3694,
+      "step": 4961
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 0.0,
+      "learning_rate": 6.184674136978049e-06,
+      "loss": 0.262,
+      "step": 4962
+    },
+    {
+      "epoch": 5.72,
+      "grad_norm": 0.0,
+      "learning_rate": 6.18083732509387e-06,
+      "loss": 0.3803,
+      "step": 4963
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 0.0,
+      "learning_rate": 6.177001171330535e-06,
+      "loss": 0.4335,
+      "step": 4964
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 0.0,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 0.3073,
+      "step": 4965
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 0.0,
+      "learning_rate": 6.169330840810506e-06,
+      "loss": 0.3552,
+      "step": 4966
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 0.0,
+      "learning_rate": 6.165496665375556e-06,
+      "loss": 0.3615,
+      "step": 4967
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 0.0,
+      "learning_rate": 6.161663150704977e-06,
+      "loss": 0.3763,
+      "step": 4968
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 0.0,
+      "learning_rate": 6.1578302974593526e-06,
+      "loss": 0.3006,
+      "step": 4969
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 0.0,
+      "learning_rate": 6.153998106299164e-06,
+      "loss": 0.4222,
+      "step": 4970
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 0.0,
+      "learning_rate": 6.150166577884781e-06,
+      "loss": 0.2417,
+      "step": 4971
+    },
+    {
+      "epoch": 5.73,
+      "grad_norm": 0.0,
+      "learning_rate": 6.146335712876448e-06,
+      "loss": 0.3458,
+      "step": 4972
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 0.0,
+      "learning_rate": 6.1425055119343115e-06,
+      "loss": 0.3844,
+      "step": 4973
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 0.0,
+      "learning_rate": 6.13867597571839e-06,
+      "loss": 0.4158,
+      "step": 4974
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 0.0,
+      "learning_rate": 6.134847104888588e-06,
+      "loss": 0.311,
+      "step": 4975
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 0.0,
+      "learning_rate": 6.131018900104706e-06,
+      "loss": 0.4108,
+      "step": 4976
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 0.0,
+      "learning_rate": 6.1271913620264225e-06,
+      "loss": 0.361,
+      "step": 4977
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 0.0,
+      "learning_rate": 6.123364491313295e-06,
+      "loss": 0.3925,
+      "step": 4978
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 0.0,
+      "learning_rate": 6.119538288624778e-06,
+      "loss": 0.3689,
+      "step": 4979
+    },
+    {
+      "epoch": 5.74,
+      "grad_norm": 0.0,
+      "learning_rate": 6.115712754620208e-06,
+      "loss": 0.2692,
+      "step": 4980
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 0.0,
+      "learning_rate": 6.111887889958795e-06,
+      "loss": 0.4414,
+      "step": 4981
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 0.0,
+      "learning_rate": 6.108063695299651e-06,
+      "loss": 0.3455,
+      "step": 4982
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 0.0,
+      "learning_rate": 6.104240171301756e-06,
+      "loss": 0.2528,
+      "step": 4983
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 0.0,
+      "learning_rate": 6.100417318623994e-06,
+      "loss": 0.3811,
+      "step": 4984
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 0.0,
+      "learning_rate": 6.096595137925111e-06,
+      "loss": 0.2319,
+      "step": 4985
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 0.0,
+      "learning_rate": 6.092773629863751e-06,
+      "loss": 0.2705,
+      "step": 4986
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 0.0,
+      "learning_rate": 6.088952795098442e-06,
+      "loss": 0.3111,
+      "step": 4987
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 0.0,
+      "learning_rate": 6.08513263428759e-06,
+      "loss": 0.2924,
+      "step": 4988
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 0.0,
+      "learning_rate": 6.081313148089485e-06,
+      "loss": 0.2943,
+      "step": 4989
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 0.0,
+      "learning_rate": 6.077494337162311e-06,
+      "loss": 0.3402,
+      "step": 4990
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 0.0,
+      "learning_rate": 6.073676202164123e-06,
+      "loss": 0.4197,
+      "step": 4991
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 0.0,
+      "learning_rate": 6.069858743752863e-06,
+      "loss": 0.304,
+      "step": 4992
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 0.0,
+      "learning_rate": 6.066041962586367e-06,
+      "loss": 0.3042,
+      "step": 4993
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 0.0,
+      "learning_rate": 6.062225859322337e-06,
+      "loss": 0.2908,
+      "step": 4994
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 0.0,
+      "learning_rate": 6.058410434618367e-06,
+      "loss": 0.3365,
+      "step": 4995
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 0.0,
+      "learning_rate": 6.054595689131939e-06,
+      "loss": 0.3057,
+      "step": 4996
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 0.0,
+      "learning_rate": 6.0507816235204065e-06,
+      "loss": 0.2565,
+      "step": 4997
+    },
+    {
+      "epoch": 5.76,
+      "grad_norm": 0.0,
+      "learning_rate": 6.04696823844102e-06,
+      "loss": 0.3186,
+      "step": 4998
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 0.0,
+      "learning_rate": 6.043155534550897e-06,
+      "loss": 0.3713,
+      "step": 4999
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 0.0,
+      "learning_rate": 6.039343512507048e-06,
+      "loss": 0.3742,
+      "step": 5000
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 0.0,
+      "learning_rate": 6.035532172966367e-06,
+      "loss": 0.3472,
+      "step": 5001
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 0.0,
+      "learning_rate": 6.03172151658562e-06,
+      "loss": 0.2804,
+      "step": 5002
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 0.0,
+      "learning_rate": 6.027911544021465e-06,
+      "loss": 0.3599,
+      "step": 5003
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 0.0,
+      "learning_rate": 6.024102255930444e-06,
+      "loss": 0.2376,
+      "step": 5004
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 0.0,
+      "learning_rate": 6.020293652968969e-06,
+      "loss": 0.3029,
+      "step": 5005
+    },
+    {
+      "epoch": 5.77,
+      "grad_norm": 0.0,
+      "learning_rate": 6.0164857357933424e-06,
+      "loss": 0.2809,
+      "step": 5006
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 0.0,
+      "learning_rate": 6.012678505059755e-06,
+      "loss": 0.4572,
+      "step": 5007
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 0.0,
+      "learning_rate": 6.008871961424259e-06,
+      "loss": 0.3215,
+      "step": 5008
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 0.0,
+      "learning_rate": 6.00506610554281e-06,
+      "loss": 0.3679,
+      "step": 5009
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 0.0,
+      "learning_rate": 6.001260938071235e-06,
+      "loss": 0.3553,
+      "step": 5010
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 0.0,
+      "learning_rate": 5.997456459665237e-06,
+      "loss": 0.282,
+      "step": 5011
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 0.0,
+      "learning_rate": 5.993652670980415e-06,
+      "loss": 0.3067,
+      "step": 5012
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 0.0,
+      "learning_rate": 5.989849572672235e-06,
+      "loss": 0.3251,
+      "step": 5013
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 0.0,
+      "learning_rate": 5.986047165396049e-06,
+      "loss": 0.2981,
+      "step": 5014
+    },
+    {
+      "epoch": 5.78,
+      "grad_norm": 0.0,
+      "learning_rate": 5.982245449807094e-06,
+      "loss": 0.2705,
+      "step": 5015
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 0.0,
+      "learning_rate": 5.978444426560484e-06,
+      "loss": 0.2919,
+      "step": 5016
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 0.0,
+      "learning_rate": 5.9746440963112095e-06,
+      "loss": 0.3235,
+      "step": 5017
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 0.0,
+      "learning_rate": 5.970844459714151e-06,
+      "loss": 0.4295,
+      "step": 5018
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 0.0,
+      "learning_rate": 5.967045517424062e-06,
+      "loss": 0.2564,
+      "step": 5019
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 0.0,
+      "learning_rate": 5.963247270095585e-06,
+      "loss": 0.406,
+      "step": 5020
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 0.0,
+      "learning_rate": 5.9594497183832276e-06,
+      "loss": 0.3748,
+      "step": 5021
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 0.0,
+      "learning_rate": 5.955652862941389e-06,
+      "loss": 0.2718,
+      "step": 5022
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 0.0,
+      "learning_rate": 5.951856704424353e-06,
+      "loss": 0.3969,
+      "step": 5023
+    },
+    {
+      "epoch": 5.79,
+      "grad_norm": 0.0,
+      "learning_rate": 5.948061243486268e-06,
+      "loss": 0.2432,
+      "step": 5024
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 0.0,
+      "learning_rate": 5.944266480781173e-06,
+      "loss": 0.4246,
+      "step": 5025
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 0.0,
+      "learning_rate": 5.940472416962986e-06,
+      "loss": 0.2942,
+      "step": 5026
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 0.0,
+      "learning_rate": 5.936679052685505e-06,
+      "loss": 0.3118,
+      "step": 5027
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 0.0,
+      "learning_rate": 5.932886388602397e-06,
+      "loss": 0.3287,
+      "step": 5028
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 0.0,
+      "learning_rate": 5.929094425367224e-06,
+      "loss": 0.3349,
+      "step": 5029
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 0.0,
+      "learning_rate": 5.925303163633419e-06,
+      "loss": 0.255,
+      "step": 5030
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 0.0,
+      "learning_rate": 5.921512604054289e-06,
+      "loss": 0.3026,
+      "step": 5031
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 0.0,
+      "learning_rate": 5.917722747283035e-06,
+      "loss": 0.3064,
+      "step": 5032
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 0.0,
+      "learning_rate": 5.913933593972719e-06,
+      "loss": 0.281,
+      "step": 5033
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 0.0,
+      "learning_rate": 5.910145144776298e-06,
+      "loss": 0.2666,
+      "step": 5034
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 0.0,
+      "learning_rate": 5.906357400346596e-06,
+      "loss": 0.2622,
+      "step": 5035
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 0.0,
+      "learning_rate": 5.902570361336319e-06,
+      "loss": 0.3915,
+      "step": 5036
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 0.0,
+      "learning_rate": 5.898784028398058e-06,
+      "loss": 0.2358,
+      "step": 5037
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 0.0,
+      "learning_rate": 5.894998402184271e-06,
+      "loss": 0.4293,
+      "step": 5038
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8912134833473e-06,
+      "loss": 0.2542,
+      "step": 5039
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 0.0,
+      "learning_rate": 5.887429272539369e-06,
+      "loss": 0.3214,
+      "step": 5040
+    },
+    {
+      "epoch": 5.81,
+      "grad_norm": 0.0,
+      "learning_rate": 5.883645770412574e-06,
+      "loss": 0.4223,
+      "step": 5041
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 0.0,
+      "learning_rate": 5.879862977618886e-06,
+      "loss": 0.3523,
+      "step": 5042
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 0.0,
+      "learning_rate": 5.876080894810167e-06,
+      "loss": 0.2983,
+      "step": 5043
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8722995226381405e-06,
+      "loss": 0.2718,
+      "step": 5044
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 0.0,
+      "learning_rate": 5.868518861754422e-06,
+      "loss": 0.2453,
+      "step": 5045
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 0.0,
+      "learning_rate": 5.864738912810495e-06,
+      "loss": 0.4162,
+      "step": 5046
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 0.0,
+      "learning_rate": 5.860959676457718e-06,
+      "loss": 0.2648,
+      "step": 5047
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8571811533473396e-06,
+      "loss": 0.3035,
+      "step": 5048
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 0.0,
+      "learning_rate": 5.853403344130476e-06,
+      "loss": 0.3368,
+      "step": 5049
+    },
+    {
+      "epoch": 5.82,
+      "grad_norm": 0.0,
+      "learning_rate": 5.849626249458116e-06,
+      "loss": 0.3731,
+      "step": 5050
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 0.0,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.2555,
+      "step": 5051
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 0.0,
+      "learning_rate": 5.842074206350288e-06,
+      "loss": 0.2771,
+      "step": 5052
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 0.0,
+      "learning_rate": 5.838299259216187e-06,
+      "loss": 0.3793,
+      "step": 5053
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 0.0,
+      "learning_rate": 5.834525029229345e-06,
+      "loss": 0.2814,
+      "step": 5054
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 0.0,
+      "learning_rate": 5.830751517040133e-06,
+      "loss": 0.322,
+      "step": 5055
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 0.0,
+      "learning_rate": 5.826978723298811e-06,
+      "loss": 0.3667,
+      "step": 5056
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 0.0,
+      "learning_rate": 5.823206648655503e-06,
+      "loss": 0.2904,
+      "step": 5057
+    },
+    {
+      "epoch": 5.83,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8194352937602186e-06,
+      "loss": 0.3641,
+      "step": 5058
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.0,
+      "learning_rate": 5.815664659262845e-06,
+      "loss": 0.3261,
+      "step": 5059
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.0,
+      "learning_rate": 5.811894745813136e-06,
+      "loss": 0.279,
+      "step": 5060
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.0,
+      "learning_rate": 5.808125554060723e-06,
+      "loss": 0.3581,
+      "step": 5061
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.0,
+      "learning_rate": 5.804357084655124e-06,
+      "loss": 0.3414,
+      "step": 5062
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8005893382457125e-06,
+      "loss": 0.3848,
+      "step": 5063
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.0,
+      "learning_rate": 5.796822315481758e-06,
+      "loss": 0.3907,
+      "step": 5064
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7930560170123985e-06,
+      "loss": 0.3169,
+      "step": 5065
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.0,
+      "learning_rate": 5.789290443486637e-06,
+      "loss": 0.2702,
+      "step": 5066
+    },
+    {
+      "epoch": 5.84,
+      "grad_norm": 0.0,
+      "learning_rate": 5.78552559555337e-06,
+      "loss": 0.2995,
+      "step": 5067
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 0.0,
+      "learning_rate": 5.781761473861352e-06,
+      "loss": 0.2417,
+      "step": 5068
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 0.0,
+      "learning_rate": 5.777998079059215e-06,
+      "loss": 0.2979,
+      "step": 5069
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 0.0,
+      "learning_rate": 5.774235411795477e-06,
+      "loss": 0.2635,
+      "step": 5070
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 0.0,
+      "learning_rate": 5.770473472718525e-06,
+      "loss": 0.4248,
+      "step": 5071
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 0.0,
+      "learning_rate": 5.766712262476611e-06,
+      "loss": 0.2675,
+      "step": 5072
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 0.0,
+      "learning_rate": 5.762951781717879e-06,
+      "loss": 0.3484,
+      "step": 5073
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 0.0,
+      "learning_rate": 5.759192031090333e-06,
+      "loss": 0.3849,
+      "step": 5074
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 0.0,
+      "learning_rate": 5.755433011241851e-06,
+      "loss": 0.2629,
+      "step": 5075
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7516747228201955e-06,
+      "loss": 0.3178,
+      "step": 5076
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 0.0,
+      "learning_rate": 5.747917166473e-06,
+      "loss": 0.3631,
+      "step": 5077
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7441603428477634e-06,
+      "loss": 0.3904,
+      "step": 5078
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 0.0,
+      "learning_rate": 5.74040425259187e-06,
+      "loss": 0.2596,
+      "step": 5079
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 0.0,
+      "learning_rate": 5.736648896352565e-06,
+      "loss": 0.316,
+      "step": 5080
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7328942747769824e-06,
+      "loss": 0.3202,
+      "step": 5081
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 0.0,
+      "learning_rate": 5.729140388512112e-06,
+      "loss": 0.2926,
+      "step": 5082
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 0.0,
+      "learning_rate": 5.725387238204831e-06,
+      "loss": 0.387,
+      "step": 5083
+    },
+    {
+      "epoch": 5.86,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7216348245018896e-06,
+      "loss": 0.3019,
+      "step": 5084
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 0.0,
+      "learning_rate": 5.717883148049901e-06,
+      "loss": 0.3262,
+      "step": 5085
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7141322094953545e-06,
+      "loss": 0.3746,
+      "step": 5086
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 0.0,
+      "learning_rate": 5.710382009484616e-06,
+      "loss": 0.255,
+      "step": 5087
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 0.0,
+      "learning_rate": 5.70663254866393e-06,
+      "loss": 0.4064,
+      "step": 5088
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 0.0,
+      "learning_rate": 5.702883827679397e-06,
+      "loss": 0.2238,
+      "step": 5089
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 0.0,
+      "learning_rate": 5.699135847177004e-06,
+      "loss": 0.3307,
+      "step": 5090
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 0.0,
+      "learning_rate": 5.695388607802603e-06,
+      "loss": 0.3641,
+      "step": 5091
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 0.0,
+      "learning_rate": 5.6916421102019245e-06,
+      "loss": 0.2493,
+      "step": 5092
+    },
+    {
+      "epoch": 5.87,
+      "grad_norm": 0.0,
+      "learning_rate": 5.687896355020563e-06,
+      "loss": 0.3443,
+      "step": 5093
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 0.0,
+      "learning_rate": 5.684151342903992e-06,
+      "loss": 0.2802,
+      "step": 5094
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 0.0,
+      "learning_rate": 5.680407074497558e-06,
+      "loss": 0.407,
+      "step": 5095
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 0.0,
+      "learning_rate": 5.676663550446475e-06,
+      "loss": 0.2887,
+      "step": 5096
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 0.0,
+      "learning_rate": 5.672920771395822e-06,
+      "loss": 0.2991,
+      "step": 5097
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 0.0,
+      "learning_rate": 5.669178737990568e-06,
+      "loss": 0.3471,
+      "step": 5098
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 0.0,
+      "learning_rate": 5.665437450875534e-06,
+      "loss": 0.3123,
+      "step": 5099
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 0.0,
+      "learning_rate": 5.661696910695425e-06,
+      "loss": 0.2752,
+      "step": 5100
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 0.0,
+      "learning_rate": 5.657957118094818e-06,
+      "loss": 0.3892,
+      "step": 5101
+    },
+    {
+      "epoch": 5.88,
+      "grad_norm": 0.0,
+      "learning_rate": 5.654218073718147e-06,
+      "loss": 0.303,
+      "step": 5102
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 0.0,
+      "learning_rate": 5.6504797782097386e-06,
+      "loss": 0.2678,
+      "step": 5103
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 0.0,
+      "learning_rate": 5.646742232213771e-06,
+      "loss": 0.4268,
+      "step": 5104
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 0.0,
+      "learning_rate": 5.6430054363742995e-06,
+      "loss": 0.3288,
+      "step": 5105
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 0.0,
+      "learning_rate": 5.639269391335254e-06,
+      "loss": 0.3367,
+      "step": 5106
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 0.0,
+      "learning_rate": 5.635534097740435e-06,
+      "loss": 0.2714,
+      "step": 5107
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 0.0,
+      "learning_rate": 5.631799556233506e-06,
+      "loss": 0.499,
+      "step": 5108
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 0.0,
+      "learning_rate": 5.628065767458011e-06,
+      "loss": 0.3283,
+      "step": 5109
+    },
+    {
+      "epoch": 5.89,
+      "grad_norm": 0.0,
+      "learning_rate": 5.624332732057357e-06,
+      "loss": 0.3639,
+      "step": 5110
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 0.0,
+      "learning_rate": 5.620600450674819e-06,
+      "loss": 0.2656,
+      "step": 5111
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 0.0,
+      "learning_rate": 5.616868923953551e-06,
+      "loss": 0.3064,
+      "step": 5112
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 0.0,
+      "learning_rate": 5.61313815253657e-06,
+      "loss": 0.2828,
+      "step": 5113
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 0.0,
+      "learning_rate": 5.609408137066772e-06,
+      "loss": 0.3802,
+      "step": 5114
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 0.0,
+      "learning_rate": 5.605678878186911e-06,
+      "loss": 0.3797,
+      "step": 5115
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 0.0,
+      "learning_rate": 5.601950376539611e-06,
+      "loss": 0.3586,
+      "step": 5116
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 0.0,
+      "learning_rate": 5.598222632767374e-06,
+      "loss": 0.3256,
+      "step": 5117
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 0.0,
+      "learning_rate": 5.594495647512572e-06,
+      "loss": 0.4164,
+      "step": 5118
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 0.0,
+      "learning_rate": 5.590769421417435e-06,
+      "loss": 0.2894,
+      "step": 5119
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.587043955124074e-06,
+      "loss": 0.3447,
+      "step": 5120
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.5833192492744635e-06,
+      "loss": 0.3121,
+      "step": 5121
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.57959530451044e-06,
+      "loss": 0.3463,
+      "step": 5122
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.575872121473722e-06,
+      "loss": 0.2404,
+      "step": 5123
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.572149700805896e-06,
+      "loss": 0.2429,
+      "step": 5124
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.568428043148402e-06,
+      "loss": 0.3174,
+      "step": 5125
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.564707149142568e-06,
+      "loss": 0.3166,
+      "step": 5126
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.560987019429575e-06,
+      "loss": 0.3261,
+      "step": 5127
+    },
+    {
+      "epoch": 5.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.557267654650482e-06,
+      "loss": 0.3576,
+      "step": 5128
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.0,
+      "learning_rate": 5.553549055446209e-06,
+      "loss": 0.3631,
+      "step": 5129
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.0,
+      "learning_rate": 5.549831222457549e-06,
+      "loss": 0.3668,
+      "step": 5130
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.0,
+      "learning_rate": 5.546114156325166e-06,
+      "loss": 0.3029,
+      "step": 5131
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.0,
+      "learning_rate": 5.542397857689585e-06,
+      "loss": 0.3533,
+      "step": 5132
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.0,
+      "learning_rate": 5.538682327191195e-06,
+      "loss": 0.3537,
+      "step": 5133
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.0,
+      "learning_rate": 5.534967565470268e-06,
+      "loss": 0.2877,
+      "step": 5134
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.0,
+      "learning_rate": 5.531253573166928e-06,
+      "loss": 0.2879,
+      "step": 5135
+    },
+    {
+      "epoch": 5.92,
+      "grad_norm": 0.0,
+      "learning_rate": 5.527540350921173e-06,
+      "loss": 0.396,
+      "step": 5136
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 0.0,
+      "learning_rate": 5.523827899372876e-06,
+      "loss": 0.3127,
+      "step": 5137
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 0.0,
+      "learning_rate": 5.52011621916176e-06,
+      "loss": 0.3938,
+      "step": 5138
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 0.0,
+      "learning_rate": 5.516405310927431e-06,
+      "loss": 0.316,
+      "step": 5139
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 0.0,
+      "learning_rate": 5.512695175309353e-06,
+      "loss": 0.4027,
+      "step": 5140
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 0.0,
+      "learning_rate": 5.50898581294685e-06,
+      "loss": 0.3815,
+      "step": 5141
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 0.0,
+      "learning_rate": 5.505277224479139e-06,
+      "loss": 0.3073,
+      "step": 5142
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 0.0,
+      "learning_rate": 5.501569410545278e-06,
+      "loss": 0.2276,
+      "step": 5143
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 0.0,
+      "learning_rate": 5.497862371784197e-06,
+      "loss": 0.3279,
+      "step": 5144
+    },
+    {
+      "epoch": 5.93,
+      "grad_norm": 0.0,
+      "learning_rate": 5.494156108834703e-06,
+      "loss": 0.34,
+      "step": 5145
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 0.0,
+      "learning_rate": 5.490450622335454e-06,
+      "loss": 0.3765,
+      "step": 5146
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4867459129249846e-06,
+      "loss": 0.4168,
+      "step": 5147
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4830419812417e-06,
+      "loss": 0.3194,
+      "step": 5148
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 0.0,
+      "learning_rate": 5.479338827923854e-06,
+      "loss": 0.2526,
+      "step": 5149
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 0.0,
+      "learning_rate": 5.475636453609585e-06,
+      "loss": 0.3009,
+      "step": 5150
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 0.0,
+      "learning_rate": 5.471934858936887e-06,
+      "loss": 0.2728,
+      "step": 5151
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4682340445436146e-06,
+      "loss": 0.3364,
+      "step": 5152
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 0.0,
+      "learning_rate": 5.464534011067499e-06,
+      "loss": 0.2941,
+      "step": 5153
+    },
+    {
+      "epoch": 5.94,
+      "grad_norm": 0.0,
+      "learning_rate": 5.460834759146138e-06,
+      "loss": 0.3871,
+      "step": 5154
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4571362894169795e-06,
+      "loss": 0.3525,
+      "step": 5155
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 0.0,
+      "learning_rate": 5.453438602517358e-06,
+      "loss": 0.2481,
+      "step": 5156
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4497416990844545e-06,
+      "loss": 0.2586,
+      "step": 5157
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 0.0,
+      "learning_rate": 5.446045579755318e-06,
+      "loss": 0.3219,
+      "step": 5158
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 0.0,
+      "learning_rate": 5.442350245166872e-06,
+      "loss": 0.3358,
+      "step": 5159
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 0.0,
+      "learning_rate": 5.438655695955901e-06,
+      "loss": 0.2822,
+      "step": 5160
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4349619327590474e-06,
+      "loss": 0.2609,
+      "step": 5161
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 0.0,
+      "learning_rate": 5.431268956212829e-06,
+      "loss": 0.2928,
+      "step": 5162
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.0,
+      "learning_rate": 5.427576766953615e-06,
+      "loss": 0.3138,
+      "step": 5163
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.0,
+      "learning_rate": 5.423885365617654e-06,
+      "loss": 0.3266,
+      "step": 5164
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4201947528410425e-06,
+      "loss": 0.3236,
+      "step": 5165
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.0,
+      "learning_rate": 5.416504929259754e-06,
+      "loss": 0.364,
+      "step": 5166
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.0,
+      "learning_rate": 5.412815895509624e-06,
+      "loss": 0.2473,
+      "step": 5167
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.0,
+      "learning_rate": 5.409127652226349e-06,
+      "loss": 0.2493,
+      "step": 5168
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.0,
+      "learning_rate": 5.405440200045482e-06,
+      "loss": 0.2788,
+      "step": 5169
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.0,
+      "learning_rate": 5.401753539602457e-06,
+      "loss": 0.2832,
+      "step": 5170
+    },
+    {
+      "epoch": 5.96,
+      "grad_norm": 0.0,
+      "learning_rate": 5.398067671532554e-06,
+      "loss": 0.3422,
+      "step": 5171
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 0.0,
+      "learning_rate": 5.394382596470929e-06,
+      "loss": 0.2898,
+      "step": 5172
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 0.0,
+      "learning_rate": 5.390698315052598e-06,
+      "loss": 0.3369,
+      "step": 5173
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 0.0,
+      "learning_rate": 5.387014827912435e-06,
+      "loss": 0.4427,
+      "step": 5174
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 0.0,
+      "learning_rate": 5.383332135685186e-06,
+      "loss": 0.2973,
+      "step": 5175
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 0.0,
+      "learning_rate": 5.37965023900545e-06,
+      "loss": 0.2805,
+      "step": 5176
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 0.0,
+      "learning_rate": 5.375969138507695e-06,
+      "loss": 0.4444,
+      "step": 5177
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 0.0,
+      "learning_rate": 5.372288834826257e-06,
+      "loss": 0.347,
+      "step": 5178
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 0.0,
+      "learning_rate": 5.368609328595323e-06,
+      "loss": 0.4688,
+      "step": 5179
+    },
+    {
+      "epoch": 5.97,
+      "grad_norm": 0.0,
+      "learning_rate": 5.364930620448946e-06,
+      "loss": 0.3623,
+      "step": 5180
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 0.0,
+      "learning_rate": 5.3612527110210495e-06,
+      "loss": 0.2751,
+      "step": 5181
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 0.0,
+      "learning_rate": 5.357575600945408e-06,
+      "loss": 0.3442,
+      "step": 5182
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 0.0,
+      "learning_rate": 5.353899290855666e-06,
+      "loss": 0.3265,
+      "step": 5183
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 0.0,
+      "learning_rate": 5.350223781385333e-06,
+      "loss": 0.3087,
+      "step": 5184
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 0.0,
+      "learning_rate": 5.346549073167766e-06,
+      "loss": 0.3523,
+      "step": 5185
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 0.0,
+      "learning_rate": 5.342875166836202e-06,
+      "loss": 0.2597,
+      "step": 5186
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 0.0,
+      "learning_rate": 5.339202063023727e-06,
+      "loss": 0.2614,
+      "step": 5187
+    },
+    {
+      "epoch": 5.98,
+      "grad_norm": 0.0,
+      "learning_rate": 5.335529762363289e-06,
+      "loss": 0.2688,
+      "step": 5188
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 0.0,
+      "learning_rate": 5.3318582654877074e-06,
+      "loss": 0.2962,
+      "step": 5189
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 0.0,
+      "learning_rate": 5.328187573029657e-06,
+      "loss": 0.2783,
+      "step": 5190
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 0.0,
+      "learning_rate": 5.324517685621671e-06,
+      "loss": 0.2922,
+      "step": 5191
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 0.0,
+      "learning_rate": 5.320848603896152e-06,
+      "loss": 0.3147,
+      "step": 5192
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 0.0,
+      "learning_rate": 5.3171803284853544e-06,
+      "loss": 0.3429,
+      "step": 5193
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 0.0,
+      "learning_rate": 5.313512860021395e-06,
+      "loss": 0.2788,
+      "step": 5194
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 0.0,
+      "learning_rate": 5.309846199136258e-06,
+      "loss": 0.2959,
+      "step": 5195
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 0.0,
+      "learning_rate": 5.306180346461787e-06,
+      "loss": 0.3023,
+      "step": 5196
+    },
+    {
+      "epoch": 5.99,
+      "grad_norm": 0.0,
+      "learning_rate": 5.302515302629686e-06,
+      "loss": 0.3396,
+      "step": 5197
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5.298851068271516e-06,
+      "loss": 0.2675,
+      "step": 5198
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5.295187644018694e-06,
+      "loss": 0.3882,
+      "step": 5199
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5.291525030502515e-06,
+      "loss": 0.3122,
+      "step": 5200
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5.287863228354113e-06,
+      "loss": 0.4704,
+      "step": 5201
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5.284202238204497e-06,
+      "loss": 0.2708,
+      "step": 5202
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5.280542060684535e-06,
+      "loss": 0.2649,
+      "step": 5203
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5.276882696424947e-06,
+      "loss": 0.2697,
+      "step": 5204
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5.273224146056315e-06,
+      "loss": 0.3529,
+      "step": 5205
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.0,
+      "learning_rate": 5.269566410209086e-06,
+      "loss": 0.2235,
+      "step": 5206
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 0.0,
+      "learning_rate": 5.265909489513568e-06,
+      "loss": 0.3205,
+      "step": 5207
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 0.0,
+      "learning_rate": 5.262253384599916e-06,
+      "loss": 0.2364,
+      "step": 5208
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 0.0,
+      "learning_rate": 5.25859809609816e-06,
+      "loss": 0.2811,
+      "step": 5209
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 0.0,
+      "learning_rate": 5.254943624638174e-06,
+      "loss": 0.2907,
+      "step": 5210
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 0.0,
+      "learning_rate": 5.2512899708497086e-06,
+      "loss": 0.324,
+      "step": 5211
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 0.0,
+      "learning_rate": 5.247637135362356e-06,
+      "loss": 0.344,
+      "step": 5212
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 0.0,
+      "learning_rate": 5.243985118805577e-06,
+      "loss": 0.2882,
+      "step": 5213
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 0.0,
+      "learning_rate": 5.2403339218086956e-06,
+      "loss": 0.2903,
+      "step": 5214
+    },
+    {
+      "epoch": 6.01,
+      "grad_norm": 0.0,
+      "learning_rate": 5.2366835450008836e-06,
+      "loss": 0.2811,
+      "step": 5215
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 0.0,
+      "learning_rate": 5.233033989011173e-06,
+      "loss": 0.2725,
+      "step": 5216
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 0.0,
+      "learning_rate": 5.229385254468469e-06,
+      "loss": 0.3371,
+      "step": 5217
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 0.0,
+      "learning_rate": 5.22573734200151e-06,
+      "loss": 0.3697,
+      "step": 5218
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 0.0,
+      "learning_rate": 5.222090252238916e-06,
+      "loss": 0.2562,
+      "step": 5219
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 0.0,
+      "learning_rate": 5.218443985809158e-06,
+      "loss": 0.301,
+      "step": 5220
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 0.0,
+      "learning_rate": 5.2147985433405555e-06,
+      "loss": 0.3148,
+      "step": 5221
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 0.0,
+      "learning_rate": 5.211153925461301e-06,
+      "loss": 0.2918,
+      "step": 5222
+    },
+    {
+      "epoch": 6.02,
+      "grad_norm": 0.0,
+      "learning_rate": 5.207510132799436e-06,
+      "loss": 0.3494,
+      "step": 5223
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.203867165982854e-06,
+      "loss": 0.3041,
+      "step": 5224
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.20022502563932e-06,
+      "loss": 0.2438,
+      "step": 5225
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.196583712396456e-06,
+      "loss": 0.2633,
+      "step": 5226
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.192943226881724e-06,
+      "loss": 0.243,
+      "step": 5227
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.189303569722463e-06,
+      "loss": 0.3567,
+      "step": 5228
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.185664741545862e-06,
+      "loss": 0.3156,
+      "step": 5229
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.182026742978959e-06,
+      "loss": 0.2926,
+      "step": 5230
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.178389574648659e-06,
+      "loss": 0.3388,
+      "step": 5231
+    },
+    {
+      "epoch": 6.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.174753237181725e-06,
+      "loss": 0.2969,
+      "step": 5232
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.171117731204775e-06,
+      "loss": 0.3363,
+      "step": 5233
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.167483057344278e-06,
+      "loss": 0.2897,
+      "step": 5234
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.163849216226562e-06,
+      "loss": 0.2535,
+      "step": 5235
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.160216208477813e-06,
+      "loss": 0.2656,
+      "step": 5236
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.156584034724081e-06,
+      "loss": 0.3284,
+      "step": 5237
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.152952695591257e-06,
+      "loss": 0.3261,
+      "step": 5238
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.149322191705102e-06,
+      "loss": 0.2275,
+      "step": 5239
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.145692523691223e-06,
+      "loss": 0.2498,
+      "step": 5240
+    },
+    {
+      "epoch": 6.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.142063692175084e-06,
+      "loss": 0.2972,
+      "step": 5241
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.1384356977820105e-06,
+      "loss": 0.2524,
+      "step": 5242
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.134808541137183e-06,
+      "loss": 0.258,
+      "step": 5243
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.131182222865638e-06,
+      "loss": 0.2787,
+      "step": 5244
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.127556743592263e-06,
+      "loss": 0.2447,
+      "step": 5245
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.123932103941799e-06,
+      "loss": 0.2685,
+      "step": 5246
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.120308304538852e-06,
+      "loss": 0.2958,
+      "step": 5247
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.116685346007873e-06,
+      "loss": 0.3321,
+      "step": 5248
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.113063228973174e-06,
+      "loss": 0.2543,
+      "step": 5249
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.1094419540589265e-06,
+      "loss": 0.2982,
+      "step": 5250
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.105821521889147e-06,
+      "loss": 0.2739,
+      "step": 5251
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.102201933087707e-06,
+      "loss": 0.2693,
+      "step": 5252
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.098583188278345e-06,
+      "loss": 0.2274,
+      "step": 5253
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.094965288084637e-06,
+      "loss": 0.3305,
+      "step": 5254
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0913482331300244e-06,
+      "loss": 0.3149,
+      "step": 5255
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.087732024037808e-06,
+      "loss": 0.2796,
+      "step": 5256
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0841166614311266e-06,
+      "loss": 0.3164,
+      "step": 5257
+    },
+    {
+      "epoch": 6.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0805021459329885e-06,
+      "loss": 0.2395,
+      "step": 5258
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.076888478166247e-06,
+      "loss": 0.3437,
+      "step": 5259
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0732756587536024e-06,
+      "loss": 0.3408,
+      "step": 5260
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.069663688317638e-06,
+      "loss": 0.2688,
+      "step": 5261
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.066052567480759e-06,
+      "loss": 0.3284,
+      "step": 5262
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.062442296865234e-06,
+      "loss": 0.3213,
+      "step": 5263
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.058832877093196e-06,
+      "loss": 0.3013,
+      "step": 5264
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.055224308786614e-06,
+      "loss": 0.2621,
+      "step": 5265
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.051616592567323e-06,
+      "loss": 0.3053,
+      "step": 5266
+    },
+    {
+      "epoch": 6.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.048009729057012e-06,
+      "loss": 0.2877,
+      "step": 5267
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.044403718877209e-06,
+      "loss": 0.2552,
+      "step": 5268
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.040798562649314e-06,
+      "loss": 0.2121,
+      "step": 5269
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.037194260994565e-06,
+      "loss": 0.3798,
+      "step": 5270
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.033590814534054e-06,
+      "loss": 0.2877,
+      "step": 5271
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0299882238887335e-06,
+      "loss": 0.3011,
+      "step": 5272
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.026386489679408e-06,
+      "loss": 0.2952,
+      "step": 5273
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0227856125267235e-06,
+      "loss": 0.2906,
+      "step": 5274
+    },
+    {
+      "epoch": 6.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0191855930511946e-06,
+      "loss": 0.2784,
+      "step": 5275
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0155864318731736e-06,
+      "loss": 0.2989,
+      "step": 5276
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.011988129612868e-06,
+      "loss": 0.2511,
+      "step": 5277
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.008390686890342e-06,
+      "loss": 0.2317,
+      "step": 5278
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.004794104325512e-06,
+      "loss": 0.3284,
+      "step": 5279
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0011983825381445e-06,
+      "loss": 0.3126,
+      "step": 5280
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.997603522147856e-06,
+      "loss": 0.2527,
+      "step": 5281
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994009523774109e-06,
+      "loss": 0.2338,
+      "step": 5282
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.990416388036233e-06,
+      "loss": 0.3164,
+      "step": 5283
+    },
+    {
+      "epoch": 6.09,
+      "grad_norm": 0.0,
+      "learning_rate": 4.986824115553393e-06,
+      "loss": 0.3671,
+      "step": 5284
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.983232706944613e-06,
+      "loss": 0.2775,
+      "step": 5285
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.979642162828774e-06,
+      "loss": 0.2532,
+      "step": 5286
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.976052483824595e-06,
+      "loss": 0.3082,
+      "step": 5287
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.972463670550649e-06,
+      "loss": 0.2417,
+      "step": 5288
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.968875723625371e-06,
+      "loss": 0.3399,
+      "step": 5289
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.96528864366703e-06,
+      "loss": 0.2502,
+      "step": 5290
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.961702431293759e-06,
+      "loss": 0.2265,
+      "step": 5291
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9581170871235404e-06,
+      "loss": 0.281,
+      "step": 5292
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 0.0,
+      "learning_rate": 4.954532611774195e-06,
+      "loss": 0.3135,
+      "step": 5293
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9509490058634104e-06,
+      "loss": 0.2337,
+      "step": 5294
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.947366270008708e-06,
+      "loss": 0.2672,
+      "step": 5295
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9437844048274705e-06,
+      "loss": 0.2532,
+      "step": 5296
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.940203410936934e-06,
+      "loss": 0.2966,
+      "step": 5297
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.936623288954171e-06,
+      "loss": 0.2927,
+      "step": 5298
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.933044039496107e-06,
+      "loss": 0.2786,
+      "step": 5299
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9294656631795325e-06,
+      "loss": 0.3429,
+      "step": 5300
+    },
+    {
+      "epoch": 6.11,
+      "grad_norm": 0.0,
+      "learning_rate": 4.925888160621064e-06,
+      "loss": 0.2474,
+      "step": 5301
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.922311532437185e-06,
+      "loss": 0.3302,
+      "step": 5302
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.918735779244226e-06,
+      "loss": 0.2413,
+      "step": 5303
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.915160901658355e-06,
+      "loss": 0.3099,
+      "step": 5304
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.911586900295606e-06,
+      "loss": 0.4115,
+      "step": 5305
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.908013775771849e-06,
+      "loss": 0.3175,
+      "step": 5306
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.904441528702806e-06,
+      "loss": 0.2732,
+      "step": 5307
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9008701597040485e-06,
+      "loss": 0.2543,
+      "step": 5308
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.897299669391006e-06,
+      "loss": 0.3141,
+      "step": 5309
+    },
+    {
+      "epoch": 6.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.893730058378937e-06,
+      "loss": 0.2574,
+      "step": 5310
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.890161327282968e-06,
+      "loss": 0.4018,
+      "step": 5311
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.886593476718062e-06,
+      "loss": 0.2379,
+      "step": 5312
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.883026507299028e-06,
+      "loss": 0.2619,
+      "step": 5313
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.879460419640535e-06,
+      "loss": 0.31,
+      "step": 5314
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.875895214357093e-06,
+      "loss": 0.3877,
+      "step": 5315
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.872330892063063e-06,
+      "loss": 0.2385,
+      "step": 5316
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.868767453372649e-06,
+      "loss": 0.3569,
+      "step": 5317
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.865204898899902e-06,
+      "loss": 0.3776,
+      "step": 5318
+    },
+    {
+      "epoch": 6.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8616432292587315e-06,
+      "loss": 0.2485,
+      "step": 5319
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8580824450628785e-06,
+      "loss": 0.2576,
+      "step": 5320
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8545225469259435e-06,
+      "loss": 0.3789,
+      "step": 5321
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.850963535461376e-06,
+      "loss": 0.2484,
+      "step": 5322
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.847405411282462e-06,
+      "loss": 0.2107,
+      "step": 5323
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.843848175002337e-06,
+      "loss": 0.2624,
+      "step": 5324
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8402918272339905e-06,
+      "loss": 0.2908,
+      "step": 5325
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.836736368590254e-06,
+      "loss": 0.3298,
+      "step": 5326
+    },
+    {
+      "epoch": 6.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.833181799683812e-06,
+      "loss": 0.2927,
+      "step": 5327
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.829628121127185e-06,
+      "loss": 0.2371,
+      "step": 5328
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.826075333532742e-06,
+      "loss": 0.3021,
+      "step": 5329
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.822523437512709e-06,
+      "loss": 0.2902,
+      "step": 5330
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.818972433679145e-06,
+      "loss": 0.3059,
+      "step": 5331
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.815422322643965e-06,
+      "loss": 0.2224,
+      "step": 5332
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.81187310501893e-06,
+      "loss": 0.3123,
+      "step": 5333
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.808324781415641e-06,
+      "loss": 0.3828,
+      "step": 5334
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.804777352445542e-06,
+      "loss": 0.2648,
+      "step": 5335
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.801230818719938e-06,
+      "loss": 0.2426,
+      "step": 5336
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.79768518084996e-06,
+      "loss": 0.2763,
+      "step": 5337
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.794140439446602e-06,
+      "loss": 0.3402,
+      "step": 5338
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.790596595120699e-06,
+      "loss": 0.2929,
+      "step": 5339
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.78705364848292e-06,
+      "loss": 0.2873,
+      "step": 5340
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.783511600143796e-06,
+      "loss": 0.3261,
+      "step": 5341
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.779970450713692e-06,
+      "loss": 0.2163,
+      "step": 5342
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.776430200802819e-06,
+      "loss": 0.2891,
+      "step": 5343
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7728908510212356e-06,
+      "loss": 0.2968,
+      "step": 5344
+    },
+    {
+      "epoch": 6.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7693524019788525e-06,
+      "loss": 0.3077,
+      "step": 5345
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.765814854285409e-06,
+      "loss": 0.2597,
+      "step": 5346
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.762278208550505e-06,
+      "loss": 0.2953,
+      "step": 5347
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.758742465383574e-06,
+      "loss": 0.2543,
+      "step": 5348
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7552076253938905e-06,
+      "loss": 0.2636,
+      "step": 5349
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7516736891905955e-06,
+      "loss": 0.265,
+      "step": 5350
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7481406573826496e-06,
+      "loss": 0.3065,
+      "step": 5351
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.744608530578872e-06,
+      "loss": 0.3671,
+      "step": 5352
+    },
+    {
+      "epoch": 6.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.741077309387921e-06,
+      "loss": 0.3246,
+      "step": 5353
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.737546994418292e-06,
+      "loss": 0.3075,
+      "step": 5354
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.734017586278337e-06,
+      "loss": 0.2945,
+      "step": 5355
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.73048908557625e-06,
+      "loss": 0.2457,
+      "step": 5356
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.726961492920057e-06,
+      "loss": 0.309,
+      "step": 5357
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.723434808917643e-06,
+      "loss": 0.255,
+      "step": 5358
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.719909034176724e-06,
+      "loss": 0.2707,
+      "step": 5359
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7163841693048615e-06,
+      "loss": 0.2858,
+      "step": 5360
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.712860214909466e-06,
+      "loss": 0.3549,
+      "step": 5361
+    },
+    {
+      "epoch": 6.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.709337171597788e-06,
+      "loss": 0.2736,
+      "step": 5362
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7058150399769245e-06,
+      "loss": 0.2418,
+      "step": 5363
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.702293820653807e-06,
+      "loss": 0.2773,
+      "step": 5364
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.698773514235212e-06,
+      "loss": 0.305,
+      "step": 5365
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.695254121327769e-06,
+      "loss": 0.3188,
+      "step": 5366
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.691735642537934e-06,
+      "loss": 0.2552,
+      "step": 5367
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.688218078472018e-06,
+      "loss": 0.2735,
+      "step": 5368
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6847014297361735e-06,
+      "loss": 0.4047,
+      "step": 5369
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.681185696936389e-06,
+      "loss": 0.3188,
+      "step": 5370
+    },
+    {
+      "epoch": 6.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.677670880678493e-06,
+      "loss": 0.3183,
+      "step": 5371
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.674156981568168e-06,
+      "loss": 0.3573,
+      "step": 5372
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6706440002109275e-06,
+      "loss": 0.2981,
+      "step": 5373
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6671319372121325e-06,
+      "loss": 0.351,
+      "step": 5374
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6636207931769884e-06,
+      "loss": 0.2671,
+      "step": 5375
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.66011056871053e-06,
+      "loss": 0.3219,
+      "step": 5376
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.65660126441765e-06,
+      "loss": 0.2902,
+      "step": 5377
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.653092880903068e-06,
+      "loss": 0.3047,
+      "step": 5378
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.649585418771348e-06,
+      "loss": 0.2959,
+      "step": 5379
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6460788786269095e-06,
+      "loss": 0.2914,
+      "step": 5380
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.642573261073996e-06,
+      "loss": 0.3484,
+      "step": 5381
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.639068566716696e-06,
+      "loss": 0.2754,
+      "step": 5382
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.635564796158946e-06,
+      "loss": 0.2539,
+      "step": 5383
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.632061950004511e-06,
+      "loss": 0.285,
+      "step": 5384
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.628560028857008e-06,
+      "loss": 0.2668,
+      "step": 5385
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6250590333198955e-06,
+      "loss": 0.2882,
+      "step": 5386
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.621558963996458e-06,
+      "loss": 0.2966,
+      "step": 5387
+    },
+    {
+      "epoch": 6.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.61805982148984e-06,
+      "loss": 0.2606,
+      "step": 5388
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6145616064030095e-06,
+      "loss": 0.3267,
+      "step": 5389
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.611064319338781e-06,
+      "loss": 0.2366,
+      "step": 5390
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.60756796089981e-06,
+      "loss": 0.3076,
+      "step": 5391
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.604072531688598e-06,
+      "loss": 0.3134,
+      "step": 5392
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.60057803230747e-06,
+      "loss": 0.2897,
+      "step": 5393
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.59708446335861e-06,
+      "loss": 0.2812,
+      "step": 5394
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.2802,
+      "step": 5395
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.590100119165573e-06,
+      "loss": 0.3617,
+      "step": 5396
+    },
+    {
+      "epoch": 6.22,
+      "grad_norm": 0.0,
+      "learning_rate": 4.586609345124944e-06,
+      "loss": 0.2511,
+      "step": 5397
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.583119503923672e-06,
+      "loss": 0.2482,
+      "step": 5398
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.579630596163134e-06,
+      "loss": 0.3081,
+      "step": 5399
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.576142622444537e-06,
+      "loss": 0.2641,
+      "step": 5400
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.572655583368927e-06,
+      "loss": 0.2778,
+      "step": 5401
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5691694795372e-06,
+      "loss": 0.2744,
+      "step": 5402
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.565684311550077e-06,
+      "loss": 0.2657,
+      "step": 5403
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.562200080008129e-06,
+      "loss": 0.2578,
+      "step": 5404
+    },
+    {
+      "epoch": 6.23,
+      "grad_norm": 0.0,
+      "learning_rate": 4.558716785511764e-06,
+      "loss": 0.3394,
+      "step": 5405
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.555234428661223e-06,
+      "loss": 0.2251,
+      "step": 5406
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.551753010056584e-06,
+      "loss": 0.3422,
+      "step": 5407
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.548272530297775e-06,
+      "loss": 0.2866,
+      "step": 5408
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.544792989984544e-06,
+      "loss": 0.306,
+      "step": 5409
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.541314389716501e-06,
+      "loss": 0.2891,
+      "step": 5410
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.537836730093077e-06,
+      "loss": 0.2918,
+      "step": 5411
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.534360011713538e-06,
+      "loss": 0.2929,
+      "step": 5412
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5308842351770026e-06,
+      "loss": 0.3074,
+      "step": 5413
+    },
+    {
+      "epoch": 6.24,
+      "grad_norm": 0.0,
+      "learning_rate": 4.527409401082412e-06,
+      "loss": 0.2912,
+      "step": 5414
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.523935510028556e-06,
+      "loss": 0.2874,
+      "step": 5415
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.520462562614063e-06,
+      "loss": 0.2358,
+      "step": 5416
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.516990559437388e-06,
+      "loss": 0.2811,
+      "step": 5417
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.513519501096826e-06,
+      "loss": 0.246,
+      "step": 5418
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.510049388190518e-06,
+      "loss": 0.3257,
+      "step": 5419
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.506580221316432e-06,
+      "loss": 0.3076,
+      "step": 5420
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.503112001072381e-06,
+      "loss": 0.2465,
+      "step": 5421
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4996447280560115e-06,
+      "loss": 0.2245,
+      "step": 5422
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4961784028648025e-06,
+      "loss": 0.2602,
+      "step": 5423
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.49271302609608e-06,
+      "loss": 0.3327,
+      "step": 5424
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.489248598346994e-06,
+      "loss": 0.2836,
+      "step": 5425
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.485785120214538e-06,
+      "loss": 0.296,
+      "step": 5426
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.482322592295541e-06,
+      "loss": 0.2355,
+      "step": 5427
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.478861015186674e-06,
+      "loss": 0.316,
+      "step": 5428
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.47540038948443e-06,
+      "loss": 0.2588,
+      "step": 5429
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4719407157851546e-06,
+      "loss": 0.3206,
+      "step": 5430
+    },
+    {
+      "epoch": 6.26,
+      "grad_norm": 0.0,
+      "learning_rate": 4.468481994685018e-06,
+      "loss": 0.2357,
+      "step": 5431
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.465024226780023e-06,
+      "loss": 0.2799,
+      "step": 5432
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4615674126660226e-06,
+      "loss": 0.2758,
+      "step": 5433
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.458111552938694e-06,
+      "loss": 0.3238,
+      "step": 5434
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.454656648193559e-06,
+      "loss": 0.2888,
+      "step": 5435
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.451202699025966e-06,
+      "loss": 0.33,
+      "step": 5436
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.447749706031096e-06,
+      "loss": 0.2827,
+      "step": 5437
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.444297669803981e-06,
+      "loss": 0.264,
+      "step": 5438
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.440846590939469e-06,
+      "loss": 0.2769,
+      "step": 5439
+    },
+    {
+      "epoch": 6.27,
+      "grad_norm": 0.0,
+      "learning_rate": 4.437396470032257e-06,
+      "loss": 0.252,
+      "step": 5440
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4339473076768765e-06,
+      "loss": 0.2523,
+      "step": 5441
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4304991044676845e-06,
+      "loss": 0.373,
+      "step": 5442
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.427051860998877e-06,
+      "loss": 0.251,
+      "step": 5443
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.423605577864485e-06,
+      "loss": 0.3619,
+      "step": 5444
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4201602556583775e-06,
+      "loss": 0.2813,
+      "step": 5445
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.416715894974258e-06,
+      "loss": 0.418,
+      "step": 5446
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4132724964056576e-06,
+      "loss": 0.3534,
+      "step": 5447
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4098300605459406e-06,
+      "loss": 0.2316,
+      "step": 5448
+    },
+    {
+      "epoch": 6.28,
+      "grad_norm": 0.0,
+      "learning_rate": 4.406388587988318e-06,
+      "loss": 0.2504,
+      "step": 5449
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.402948079325821e-06,
+      "loss": 0.2913,
+      "step": 5450
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.399508535151321e-06,
+      "loss": 0.3078,
+      "step": 5451
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.396069956057528e-06,
+      "loss": 0.3678,
+      "step": 5452
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.392632342636978e-06,
+      "loss": 0.2776,
+      "step": 5453
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.389195695482037e-06,
+      "loss": 0.2633,
+      "step": 5454
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.385760015184919e-06,
+      "loss": 0.2506,
+      "step": 5455
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.382325302337657e-06,
+      "loss": 0.2823,
+      "step": 5456
+    },
+    {
+      "epoch": 6.29,
+      "grad_norm": 0.0,
+      "learning_rate": 4.378891557532123e-06,
+      "loss": 0.2727,
+      "step": 5457
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.37545878136003e-06,
+      "loss": 0.2837,
+      "step": 5458
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.372026974412907e-06,
+      "loss": 0.2691,
+      "step": 5459
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.368596137282134e-06,
+      "loss": 0.2825,
+      "step": 5460
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3651662705589104e-06,
+      "loss": 0.3482,
+      "step": 5461
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.36173737483427e-06,
+      "loss": 0.3918,
+      "step": 5462
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.358309450699087e-06,
+      "loss": 0.2927,
+      "step": 5463
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.354882498744066e-06,
+      "loss": 0.3133,
+      "step": 5464
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3514565195597356e-06,
+      "loss": 0.2508,
+      "step": 5465
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3480315137364706e-06,
+      "loss": 0.3135,
+      "step": 5466
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.344607481864466e-06,
+      "loss": 0.337,
+      "step": 5467
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.341184424533746e-06,
+      "loss": 0.3044,
+      "step": 5468
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.337762342334189e-06,
+      "loss": 0.2822,
+      "step": 5469
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.33434123585548e-06,
+      "loss": 0.3003,
+      "step": 5470
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.330921105687155e-06,
+      "loss": 0.2742,
+      "step": 5471
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.327501952418567e-06,
+      "loss": 0.2672,
+      "step": 5472
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.324083776638905e-06,
+      "loss": 0.3358,
+      "step": 5473
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.320666578937196e-06,
+      "loss": 0.2762,
+      "step": 5474
+    },
+    {
+      "epoch": 6.31,
+      "grad_norm": 0.0,
+      "learning_rate": 4.317250359902295e-06,
+      "loss": 0.307,
+      "step": 5475
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.0,
+      "learning_rate": 4.313835120122881e-06,
+      "loss": 0.3684,
+      "step": 5476
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.0,
+      "learning_rate": 4.31042086018748e-06,
+      "loss": 0.2423,
+      "step": 5477
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3070075806844335e-06,
+      "loss": 0.2695,
+      "step": 5478
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.0,
+      "learning_rate": 4.303595282201918e-06,
+      "loss": 0.2489,
+      "step": 5479
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.0,
+      "learning_rate": 4.300183965327945e-06,
+      "loss": 0.304,
+      "step": 5480
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.0,
+      "learning_rate": 4.296773630650358e-06,
+      "loss": 0.2957,
+      "step": 5481
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2933642787568295e-06,
+      "loss": 0.2993,
+      "step": 5482
+    },
+    {
+      "epoch": 6.32,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2899559102348585e-06,
+      "loss": 0.2434,
+      "step": 5483
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 0.0,
+      "learning_rate": 4.286548525671775e-06,
+      "loss": 0.3238,
+      "step": 5484
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 0.0,
+      "learning_rate": 4.283142125654748e-06,
+      "loss": 0.2459,
+      "step": 5485
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 0.0,
+      "learning_rate": 4.279736710770765e-06,
+      "loss": 0.2369,
+      "step": 5486
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 0.0,
+      "learning_rate": 4.27633228160665e-06,
+      "loss": 0.3167,
+      "step": 5487
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 0.0,
+      "learning_rate": 4.272928838749063e-06,
+      "loss": 0.2706,
+      "step": 5488
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2695263827844835e-06,
+      "loss": 0.2515,
+      "step": 5489
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 0.0,
+      "learning_rate": 4.266124914299221e-06,
+      "loss": 0.2973,
+      "step": 5490
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 0.0,
+      "learning_rate": 4.262724433879427e-06,
+      "loss": 0.2708,
+      "step": 5491
+    },
+    {
+      "epoch": 6.33,
+      "grad_norm": 0.0,
+      "learning_rate": 4.259324942111067e-06,
+      "loss": 0.2629,
+      "step": 5492
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.0,
+      "learning_rate": 4.255926439579948e-06,
+      "loss": 0.2662,
+      "step": 5493
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.0,
+      "learning_rate": 4.252528926871705e-06,
+      "loss": 0.3621,
+      "step": 5494
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.0,
+      "learning_rate": 4.249132404571794e-06,
+      "loss": 0.3402,
+      "step": 5495
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.0,
+      "learning_rate": 4.245736873265513e-06,
+      "loss": 0.2714,
+      "step": 5496
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.0,
+      "learning_rate": 4.242342333537978e-06,
+      "loss": 0.2665,
+      "step": 5497
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.0,
+      "learning_rate": 4.238948785974131e-06,
+      "loss": 0.3292,
+      "step": 5498
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.0,
+      "learning_rate": 4.235556231158765e-06,
+      "loss": 0.2675,
+      "step": 5499
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.0,
+      "learning_rate": 4.232164669676483e-06,
+      "loss": 0.2721,
+      "step": 5500
+    },
+    {
+      "epoch": 6.34,
+      "grad_norm": 0.0,
+      "learning_rate": 4.228774102111714e-06,
+      "loss": 0.2624,
+      "step": 5501
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 0.0,
+      "learning_rate": 4.225384529048733e-06,
+      "loss": 0.3471,
+      "step": 5502
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 0.0,
+      "learning_rate": 4.221995951071623e-06,
+      "loss": 0.3918,
+      "step": 5503
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 0.0,
+      "learning_rate": 4.218608368764314e-06,
+      "loss": 0.2715,
+      "step": 5504
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 0.0,
+      "learning_rate": 4.215221782710559e-06,
+      "loss": 0.3642,
+      "step": 5505
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 0.0,
+      "learning_rate": 4.211836193493928e-06,
+      "loss": 0.3116,
+      "step": 5506
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 0.0,
+      "learning_rate": 4.208451601697836e-06,
+      "loss": 0.3051,
+      "step": 5507
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 0.0,
+      "learning_rate": 4.205068007905516e-06,
+      "loss": 0.3443,
+      "step": 5508
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 0.0,
+      "learning_rate": 4.201685412700026e-06,
+      "loss": 0.2432,
+      "step": 5509
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 0.0,
+      "learning_rate": 4.198303816664263e-06,
+      "loss": 0.2664,
+      "step": 5510
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 0.0,
+      "learning_rate": 4.194923220380947e-06,
+      "loss": 0.2699,
+      "step": 5511
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 0.0,
+      "learning_rate": 4.191543624432619e-06,
+      "loss": 0.2773,
+      "step": 5512
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1881650294016606e-06,
+      "loss": 0.3043,
+      "step": 5513
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 0.0,
+      "learning_rate": 4.184787435870271e-06,
+      "loss": 0.2971,
+      "step": 5514
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 0.0,
+      "learning_rate": 4.181410844420473e-06,
+      "loss": 0.2681,
+      "step": 5515
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1780352556341305e-06,
+      "loss": 0.281,
+      "step": 5516
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1746606700929245e-06,
+      "loss": 0.2683,
+      "step": 5517
+    },
+    {
+      "epoch": 6.36,
+      "grad_norm": 0.0,
+      "learning_rate": 4.171287088378372e-06,
+      "loss": 0.426,
+      "step": 5518
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 0.0,
+      "learning_rate": 4.167914511071805e-06,
+      "loss": 0.306,
+      "step": 5519
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 0.0,
+      "learning_rate": 4.164542938754386e-06,
+      "loss": 0.4544,
+      "step": 5520
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 0.0,
+      "learning_rate": 4.161172372007115e-06,
+      "loss": 0.365,
+      "step": 5521
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 0.0,
+      "learning_rate": 4.157802811410801e-06,
+      "loss": 0.3379,
+      "step": 5522
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 0.0,
+      "learning_rate": 4.154434257546095e-06,
+      "loss": 0.3872,
+      "step": 5523
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 0.0,
+      "learning_rate": 4.15106671099347e-06,
+      "loss": 0.2713,
+      "step": 5524
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 0.0,
+      "learning_rate": 4.147700172333222e-06,
+      "loss": 0.3429,
+      "step": 5525
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 0.0,
+      "learning_rate": 4.144334642145472e-06,
+      "loss": 0.318,
+      "step": 5526
+    },
+    {
+      "epoch": 6.37,
+      "grad_norm": 0.0,
+      "learning_rate": 4.140970121010177e-06,
+      "loss": 0.2941,
+      "step": 5527
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 0.0,
+      "learning_rate": 4.137606609507103e-06,
+      "loss": 0.2548,
+      "step": 5528
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 0.0,
+      "learning_rate": 4.134244108215867e-06,
+      "loss": 0.2712,
+      "step": 5529
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 0.0,
+      "learning_rate": 4.13088261771589e-06,
+      "loss": 0.2235,
+      "step": 5530
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 0.0,
+      "learning_rate": 4.127522138586424e-06,
+      "loss": 0.2998,
+      "step": 5531
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1241626714065555e-06,
+      "loss": 0.2544,
+      "step": 5532
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 0.0,
+      "learning_rate": 4.120804216755183e-06,
+      "loss": 0.2829,
+      "step": 5533
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 0.0,
+      "learning_rate": 4.117446775211042e-06,
+      "loss": 0.3059,
+      "step": 5534
+    },
+    {
+      "epoch": 6.38,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1140903473526935e-06,
+      "loss": 0.3184,
+      "step": 5535
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1107349337585145e-06,
+      "loss": 0.2964,
+      "step": 5536
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 0.0,
+      "learning_rate": 4.107380535006709e-06,
+      "loss": 0.2533,
+      "step": 5537
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 0.0,
+      "learning_rate": 4.104027151675318e-06,
+      "loss": 0.2775,
+      "step": 5538
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 0.0,
+      "learning_rate": 4.10067478434219e-06,
+      "loss": 0.3922,
+      "step": 5539
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 0.0,
+      "learning_rate": 4.097323433585011e-06,
+      "loss": 0.2474,
+      "step": 5540
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 0.0,
+      "learning_rate": 4.093973099981292e-06,
+      "loss": 0.2692,
+      "step": 5541
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0906237841083585e-06,
+      "loss": 0.2541,
+      "step": 5542
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 0.0,
+      "learning_rate": 4.087275486543374e-06,
+      "loss": 0.2988,
+      "step": 5543
+    },
+    {
+      "epoch": 6.39,
+      "grad_norm": 0.0,
+      "learning_rate": 4.083928207863315e-06,
+      "loss": 0.255,
+      "step": 5544
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.0,
+      "learning_rate": 4.080581948644985e-06,
+      "loss": 0.2913,
+      "step": 5545
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.0,
+      "learning_rate": 4.077236709465014e-06,
+      "loss": 0.3055,
+      "step": 5546
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.0,
+      "learning_rate": 4.073892490899865e-06,
+      "loss": 0.3249,
+      "step": 5547
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0705492935258036e-06,
+      "loss": 0.3679,
+      "step": 5548
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.0,
+      "learning_rate": 4.067207117918944e-06,
+      "loss": 0.2258,
+      "step": 5549
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.0,
+      "learning_rate": 4.063865964655204e-06,
+      "loss": 0.2527,
+      "step": 5550
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0605258343103315e-06,
+      "loss": 0.3181,
+      "step": 5551
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.0,
+      "learning_rate": 4.057186727459906e-06,
+      "loss": 0.2933,
+      "step": 5552
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 0.0,
+      "learning_rate": 4.053848644679322e-06,
+      "loss": 0.2851,
+      "step": 5553
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 0.0,
+      "learning_rate": 4.050511586543806e-06,
+      "loss": 0.3075,
+      "step": 5554
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 0.0,
+      "learning_rate": 4.047175553628397e-06,
+      "loss": 0.3383,
+      "step": 5555
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 0.0,
+      "learning_rate": 4.043840546507961e-06,
+      "loss": 0.2694,
+      "step": 5556
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 0.0,
+      "learning_rate": 4.04050656575719e-06,
+      "loss": 0.3467,
+      "step": 5557
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 0.0,
+      "learning_rate": 4.037173611950603e-06,
+      "loss": 0.2675,
+      "step": 5558
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 0.0,
+      "learning_rate": 4.033841685662529e-06,
+      "loss": 0.2693,
+      "step": 5559
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 0.0,
+      "learning_rate": 4.030510787467137e-06,
+      "loss": 0.2585,
+      "step": 5560
+    },
+    {
+      "epoch": 6.41,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0271809179384035e-06,
+      "loss": 0.28,
+      "step": 5561
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 0.0,
+      "learning_rate": 4.023852077650131e-06,
+      "loss": 0.3058,
+      "step": 5562
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 0.0,
+      "learning_rate": 4.020524267175954e-06,
+      "loss": 0.2696,
+      "step": 5563
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 0.0,
+      "learning_rate": 4.017197487089319e-06,
+      "loss": 0.3278,
+      "step": 5564
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 0.0,
+      "learning_rate": 4.013871737963506e-06,
+      "loss": 0.2516,
+      "step": 5565
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 0.0,
+      "learning_rate": 4.010547020371604e-06,
+      "loss": 0.2665,
+      "step": 5566
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 0.0,
+      "learning_rate": 4.007223334886531e-06,
+      "loss": 0.2731,
+      "step": 5567
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 0.0,
+      "learning_rate": 4.003900682081031e-06,
+      "loss": 0.3452,
+      "step": 5568
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 0.0,
+      "learning_rate": 4.00057906252766e-06,
+      "loss": 0.3465,
+      "step": 5569
+    },
+    {
+      "epoch": 6.42,
+      "grad_norm": 0.0,
+      "learning_rate": 3.997258476798804e-06,
+      "loss": 0.288,
+      "step": 5570
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.993938925466674e-06,
+      "loss": 0.2475,
+      "step": 5571
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.990620409103293e-06,
+      "loss": 0.2733,
+      "step": 5572
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.987302928280507e-06,
+      "loss": 0.325,
+      "step": 5573
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9839864835699935e-06,
+      "loss": 0.2424,
+      "step": 5574
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.980671075543238e-06,
+      "loss": 0.4788,
+      "step": 5575
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.977356704771556e-06,
+      "loss": 0.2365,
+      "step": 5576
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.974043371826087e-06,
+      "loss": 0.2426,
+      "step": 5577
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.970731077277781e-06,
+      "loss": 0.3145,
+      "step": 5578
+    },
+    {
+      "epoch": 6.43,
+      "grad_norm": 0.0,
+      "learning_rate": 3.96741982169742e-06,
+      "loss": 0.262,
+      "step": 5579
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.964109605655599e-06,
+      "loss": 0.3069,
+      "step": 5580
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.960800429722734e-06,
+      "loss": 0.3456,
+      "step": 5581
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.957492294469067e-06,
+      "loss": 0.322,
+      "step": 5582
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.954185200464663e-06,
+      "loss": 0.2973,
+      "step": 5583
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.950879148279395e-06,
+      "loss": 0.2596,
+      "step": 5584
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9475741384829715e-06,
+      "loss": 0.3402,
+      "step": 5585
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.944270171644906e-06,
+      "loss": 0.2699,
+      "step": 5586
+    },
+    {
+      "epoch": 6.44,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9409672483345465e-06,
+      "loss": 0.2957,
+      "step": 5587
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.0,
+      "learning_rate": 3.937665369121055e-06,
+      "loss": 0.2844,
+      "step": 5588
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.0,
+      "learning_rate": 3.93436453457341e-06,
+      "loss": 0.262,
+      "step": 5589
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.0,
+      "learning_rate": 3.93106474526042e-06,
+      "loss": 0.3322,
+      "step": 5590
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.0,
+      "learning_rate": 3.927766001750702e-06,
+      "loss": 0.2411,
+      "step": 5591
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.0,
+      "learning_rate": 3.924468304612697e-06,
+      "loss": 0.2715,
+      "step": 5592
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.0,
+      "learning_rate": 3.921171654414667e-06,
+      "loss": 0.291,
+      "step": 5593
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.0,
+      "learning_rate": 3.917876051724698e-06,
+      "loss": 0.2256,
+      "step": 5594
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.0,
+      "learning_rate": 3.914581497110684e-06,
+      "loss": 0.2673,
+      "step": 5595
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 0.0,
+      "learning_rate": 3.911287991140351e-06,
+      "loss": 0.2742,
+      "step": 5596
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 0.0,
+      "learning_rate": 3.907995534381237e-06,
+      "loss": 0.2955,
+      "step": 5597
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 0.0,
+      "learning_rate": 3.904704127400694e-06,
+      "loss": 0.2492,
+      "step": 5598
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 0.0,
+      "learning_rate": 3.901413770765905e-06,
+      "loss": 0.2502,
+      "step": 5599
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 0.0,
+      "learning_rate": 3.898124465043866e-06,
+      "loss": 0.2816,
+      "step": 5600
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 0.0,
+      "learning_rate": 3.894836210801397e-06,
+      "loss": 0.3501,
+      "step": 5601
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 0.0,
+      "learning_rate": 3.891549008605128e-06,
+      "loss": 0.2798,
+      "step": 5602
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 0.0,
+      "learning_rate": 3.888262859021508e-06,
+      "loss": 0.2765,
+      "step": 5603
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 0.0,
+      "learning_rate": 3.884977762616816e-06,
+      "loss": 0.3,
+      "step": 5604
+    },
+    {
+      "epoch": 6.46,
+      "grad_norm": 0.0,
+      "learning_rate": 3.881693719957136e-06,
+      "loss": 0.3298,
+      "step": 5605
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 0.0,
+      "learning_rate": 3.878410731608379e-06,
+      "loss": 0.2619,
+      "step": 5606
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 0.0,
+      "learning_rate": 3.875128798136276e-06,
+      "loss": 0.2586,
+      "step": 5607
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 0.0,
+      "learning_rate": 3.871847920106369e-06,
+      "loss": 0.3236,
+      "step": 5608
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 0.0,
+      "learning_rate": 3.868568098084017e-06,
+      "loss": 0.2929,
+      "step": 5609
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 0.0,
+      "learning_rate": 3.865289332634407e-06,
+      "loss": 0.3055,
+      "step": 5610
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 0.0,
+      "learning_rate": 3.862011624322534e-06,
+      "loss": 0.2702,
+      "step": 5611
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 0.0,
+      "learning_rate": 3.858734973713218e-06,
+      "loss": 0.2369,
+      "step": 5612
+    },
+    {
+      "epoch": 6.47,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8554593813710975e-06,
+      "loss": 0.2906,
+      "step": 5613
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.0,
+      "learning_rate": 3.852184847860615e-06,
+      "loss": 0.3059,
+      "step": 5614
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.0,
+      "learning_rate": 3.848911373746051e-06,
+      "loss": 0.2429,
+      "step": 5615
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.0,
+      "learning_rate": 3.845638959591485e-06,
+      "loss": 0.2456,
+      "step": 5616
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.0,
+      "learning_rate": 3.842367605960821e-06,
+      "loss": 0.2402,
+      "step": 5617
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.0,
+      "learning_rate": 3.83909731341779e-06,
+      "loss": 0.2637,
+      "step": 5618
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.0,
+      "learning_rate": 3.835828082525925e-06,
+      "loss": 0.2881,
+      "step": 5619
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8325599138485814e-06,
+      "loss": 0.2674,
+      "step": 5620
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.0,
+      "learning_rate": 3.829292807948937e-06,
+      "loss": 0.3446,
+      "step": 5621
+    },
+    {
+      "epoch": 6.48,
+      "grad_norm": 0.0,
+      "learning_rate": 3.826026765389974e-06,
+      "loss": 0.2646,
+      "step": 5622
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 0.0,
+      "learning_rate": 3.822761786734506e-06,
+      "loss": 0.3022,
+      "step": 5623
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 0.0,
+      "learning_rate": 3.819497872545156e-06,
+      "loss": 0.2809,
+      "step": 5624
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 0.0,
+      "learning_rate": 3.81623502338436e-06,
+      "loss": 0.2598,
+      "step": 5625
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8129732398143804e-06,
+      "loss": 0.2421,
+      "step": 5626
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8097125223972864e-06,
+      "loss": 0.2887,
+      "step": 5627
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8064528716949644e-06,
+      "loss": 0.2592,
+      "step": 5628
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 0.0,
+      "learning_rate": 3.803194288269122e-06,
+      "loss": 0.2449,
+      "step": 5629
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 0.0,
+      "learning_rate": 3.799936772681286e-06,
+      "loss": 0.2722,
+      "step": 5630
+    },
+    {
+      "epoch": 6.49,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7966803254927852e-06,
+      "loss": 0.2957,
+      "step": 5631
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.0,
+      "learning_rate": 3.79342494726478e-06,
+      "loss": 0.3484,
+      "step": 5632
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7901706385582372e-06,
+      "loss": 0.3297,
+      "step": 5633
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.0,
+      "learning_rate": 3.786917399933937e-06,
+      "loss": 0.297,
+      "step": 5634
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7836652319524835e-06,
+      "loss": 0.2807,
+      "step": 5635
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.0,
+      "learning_rate": 3.780414135174293e-06,
+      "loss": 0.2435,
+      "step": 5636
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7771641101596e-06,
+      "loss": 0.2777,
+      "step": 5637
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7739151574684484e-06,
+      "loss": 0.2537,
+      "step": 5638
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.0,
+      "learning_rate": 3.770667277660696e-06,
+      "loss": 0.3185,
+      "step": 5639
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 0.0,
+      "learning_rate": 3.767420471296026e-06,
+      "loss": 0.2752,
+      "step": 5640
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7641747389339232e-06,
+      "loss": 0.2877,
+      "step": 5641
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7609300811337004e-06,
+      "loss": 0.3358,
+      "step": 5642
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7576864984544814e-06,
+      "loss": 0.2457,
+      "step": 5643
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 0.0,
+      "learning_rate": 3.754443991455199e-06,
+      "loss": 0.2122,
+      "step": 5644
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7512025606946e-06,
+      "loss": 0.2479,
+      "step": 5645
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 0.0,
+      "learning_rate": 3.747962206731256e-06,
+      "loss": 0.2513,
+      "step": 5646
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7447229301235443e-06,
+      "loss": 0.3028,
+      "step": 5647
+    },
+    {
+      "epoch": 6.51,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7414847314296654e-06,
+      "loss": 0.3002,
+      "step": 5648
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7382476112076227e-06,
+      "loss": 0.2915,
+      "step": 5649
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 0.0,
+      "learning_rate": 3.735011570015237e-06,
+      "loss": 0.3396,
+      "step": 5650
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 0.0,
+      "learning_rate": 3.73177660841015e-06,
+      "loss": 0.2749,
+      "step": 5651
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7285427269498087e-06,
+      "loss": 0.2744,
+      "step": 5652
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7253099261914794e-06,
+      "loss": 0.2937,
+      "step": 5653
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7220782066922434e-06,
+      "loss": 0.2563,
+      "step": 5654
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 0.0,
+      "learning_rate": 3.718847569008992e-06,
+      "loss": 0.3219,
+      "step": 5655
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7156180136984253e-06,
+      "loss": 0.2728,
+      "step": 5656
+    },
+    {
+      "epoch": 6.52,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7123895413170706e-06,
+      "loss": 0.3267,
+      "step": 5657
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7091621524212527e-06,
+      "loss": 0.3127,
+      "step": 5658
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7059358475671225e-06,
+      "loss": 0.3012,
+      "step": 5659
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7027106273106415e-06,
+      "loss": 0.2903,
+      "step": 5660
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6994864922075756e-06,
+      "loss": 0.3527,
+      "step": 5661
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6962634428135157e-06,
+      "loss": 0.2466,
+      "step": 5662
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6930414796838587e-06,
+      "loss": 0.3805,
+      "step": 5663
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6898206033738106e-06,
+      "loss": 0.2952,
+      "step": 5664
+    },
+    {
+      "epoch": 6.53,
+      "grad_norm": 0.0,
+      "learning_rate": 3.686600814438399e-06,
+      "loss": 0.2595,
+      "step": 5665
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6833821134324632e-06,
+      "loss": 0.312,
+      "step": 5666
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 0.0,
+      "learning_rate": 3.680164500910646e-06,
+      "loss": 0.3225,
+      "step": 5667
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 0.0,
+      "learning_rate": 3.676947977427415e-06,
+      "loss": 0.34,
+      "step": 5668
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6737325435370376e-06,
+      "loss": 0.2368,
+      "step": 5669
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 0.0,
+      "learning_rate": 3.670518199793606e-06,
+      "loss": 0.3594,
+      "step": 5670
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 0.0,
+      "learning_rate": 3.667304946751011e-06,
+      "loss": 0.2616,
+      "step": 5671
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 0.0,
+      "learning_rate": 3.664092784962967e-06,
+      "loss": 0.3044,
+      "step": 5672
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 0.0,
+      "learning_rate": 3.660881714982999e-06,
+      "loss": 0.234,
+      "step": 5673
+    },
+    {
+      "epoch": 6.54,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6576717373644368e-06,
+      "loss": 0.2747,
+      "step": 5674
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 0.0,
+      "learning_rate": 3.654462852660423e-06,
+      "loss": 0.3706,
+      "step": 5675
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6512550614239197e-06,
+      "loss": 0.2466,
+      "step": 5676
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 0.0,
+      "learning_rate": 3.648048364207698e-06,
+      "loss": 0.317,
+      "step": 5677
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6448427615643313e-06,
+      "loss": 0.3071,
+      "step": 5678
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 0.0,
+      "learning_rate": 3.641638254046218e-06,
+      "loss": 0.3019,
+      "step": 5679
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 0.0,
+      "learning_rate": 3.638434842205558e-06,
+      "loss": 0.3668,
+      "step": 5680
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6352325265943634e-06,
+      "loss": 0.2531,
+      "step": 5681
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 0.0,
+      "learning_rate": 3.632031307764461e-06,
+      "loss": 0.2782,
+      "step": 5682
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6288311862674885e-06,
+      "loss": 0.2451,
+      "step": 5683
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 0.0,
+      "learning_rate": 3.625632162654895e-06,
+      "loss": 0.3366,
+      "step": 5684
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6224342374779365e-06,
+      "loss": 0.2392,
+      "step": 5685
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 0.0,
+      "learning_rate": 3.619237411287678e-06,
+      "loss": 0.358,
+      "step": 5686
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6160416846350056e-06,
+      "loss": 0.3143,
+      "step": 5687
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 0.0,
+      "learning_rate": 3.612847058070602e-06,
+      "loss": 0.275,
+      "step": 5688
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 0.0,
+      "learning_rate": 3.609653532144972e-06,
+      "loss": 0.322,
+      "step": 5689
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 0.0,
+      "learning_rate": 3.606461107408429e-06,
+      "loss": 0.324,
+      "step": 5690
+    },
+    {
+      "epoch": 6.56,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6032697844110896e-06,
+      "loss": 0.3543,
+      "step": 5691
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 0.0,
+      "learning_rate": 3.600079563702883e-06,
+      "loss": 0.4028,
+      "step": 5692
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5968904458335573e-06,
+      "loss": 0.2659,
+      "step": 5693
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5937024313526546e-06,
+      "loss": 0.2981,
+      "step": 5694
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 0.0,
+      "learning_rate": 3.590515520809542e-06,
+      "loss": 0.2644,
+      "step": 5695
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5873297147533913e-06,
+      "loss": 0.271,
+      "step": 5696
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 0.0,
+      "learning_rate": 3.584145013733178e-06,
+      "loss": 0.2414,
+      "step": 5697
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 0.0,
+      "learning_rate": 3.580961418297697e-06,
+      "loss": 0.2981,
+      "step": 5698
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5777789289955454e-06,
+      "loss": 0.2422,
+      "step": 5699
+    },
+    {
+      "epoch": 6.57,
+      "grad_norm": 0.0,
+      "learning_rate": 3.574597546375129e-06,
+      "loss": 0.2627,
+      "step": 5700
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5714172709846686e-06,
+      "loss": 0.2478,
+      "step": 5701
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5682381033721946e-06,
+      "loss": 0.2484,
+      "step": 5702
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5650600440855375e-06,
+      "loss": 0.2249,
+      "step": 5703
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5618830936723503e-06,
+      "loss": 0.262,
+      "step": 5704
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 0.0,
+      "learning_rate": 3.558707252680078e-06,
+      "loss": 0.261,
+      "step": 5705
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 0.0,
+      "learning_rate": 3.55553252165599e-06,
+      "loss": 0.3283,
+      "step": 5706
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5523589011471592e-06,
+      "loss": 0.2698,
+      "step": 5707
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 0.0,
+      "learning_rate": 3.549186391700461e-06,
+      "loss": 0.3213,
+      "step": 5708
+    },
+    {
+      "epoch": 6.58,
+      "grad_norm": 0.0,
+      "learning_rate": 3.546014993862591e-06,
+      "loss": 0.2765,
+      "step": 5709
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5428447081800434e-06,
+      "loss": 0.3379,
+      "step": 5710
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5396755351991197e-06,
+      "loss": 0.3665,
+      "step": 5711
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 0.0,
+      "learning_rate": 3.536507475465939e-06,
+      "loss": 0.2908,
+      "step": 5712
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5333405295264255e-06,
+      "loss": 0.2506,
+      "step": 5713
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 0.0,
+      "learning_rate": 3.530174697926304e-06,
+      "loss": 0.263,
+      "step": 5714
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 0.0,
+      "learning_rate": 3.527009981211119e-06,
+      "loss": 0.3642,
+      "step": 5715
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 0.0,
+      "learning_rate": 3.52384637992621e-06,
+      "loss": 0.3089,
+      "step": 5716
+    },
+    {
+      "epoch": 6.59,
+      "grad_norm": 0.0,
+      "learning_rate": 3.520683894616739e-06,
+      "loss": 0.3018,
+      "step": 5717
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5175225258276593e-06,
+      "loss": 0.3256,
+      "step": 5718
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5143622741037432e-06,
+      "loss": 0.3064,
+      "step": 5719
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.0,
+      "learning_rate": 3.511203139989573e-06,
+      "loss": 0.2782,
+      "step": 5720
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5080451240295288e-06,
+      "loss": 0.2918,
+      "step": 5721
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5048882267677985e-06,
+      "loss": 0.258,
+      "step": 5722
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5017324487483873e-06,
+      "loss": 0.3319,
+      "step": 5723
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.0,
+      "learning_rate": 3.498577790515095e-06,
+      "loss": 0.2875,
+      "step": 5724
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.0,
+      "learning_rate": 3.495424252611538e-06,
+      "loss": 0.3033,
+      "step": 5725
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 0.0,
+      "learning_rate": 3.492271835581139e-06,
+      "loss": 0.3297,
+      "step": 5726
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4891205399671226e-06,
+      "loss": 0.252,
+      "step": 5727
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4859703663125166e-06,
+      "loss": 0.2591,
+      "step": 5728
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4828213151601695e-06,
+      "loss": 0.2671,
+      "step": 5729
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4796733870527223e-06,
+      "loss": 0.3353,
+      "step": 5730
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 0.0,
+      "learning_rate": 3.47652658253263e-06,
+      "loss": 0.3538,
+      "step": 5731
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 0.0,
+      "learning_rate": 3.473380902142156e-06,
+      "loss": 0.32,
+      "step": 5732
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4702363464233604e-06,
+      "loss": 0.2963,
+      "step": 5733
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 0.0,
+      "learning_rate": 3.46709291591812e-06,
+      "loss": 0.2546,
+      "step": 5734
+    },
+    {
+      "epoch": 6.61,
+      "grad_norm": 0.0,
+      "learning_rate": 3.463950611168111e-06,
+      "loss": 0.3105,
+      "step": 5735
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 0.0,
+      "learning_rate": 3.460809432714811e-06,
+      "loss": 0.2331,
+      "step": 5736
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 0.0,
+      "learning_rate": 3.457669381099523e-06,
+      "loss": 0.2981,
+      "step": 5737
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4545304568633363e-06,
+      "loss": 0.2723,
+      "step": 5738
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.3124,
+      "step": 5739
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 0.0,
+      "learning_rate": 3.448255992691677e-06,
+      "loss": 0.2955,
+      "step": 5740
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 0.0,
+      "learning_rate": 3.445120453837424e-06,
+      "loss": 0.2622,
+      "step": 5741
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4419860445247132e-06,
+      "loss": 0.2668,
+      "step": 5742
+    },
+    {
+      "epoch": 6.62,
+      "grad_norm": 0.0,
+      "learning_rate": 3.438852765293671e-06,
+      "loss": 0.258,
+      "step": 5743
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 0.0,
+      "learning_rate": 3.435720616684218e-06,
+      "loss": 0.2398,
+      "step": 5744
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 0.0,
+      "learning_rate": 3.432589599236099e-06,
+      "loss": 0.2361,
+      "step": 5745
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4294597134888462e-06,
+      "loss": 0.2888,
+      "step": 5746
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4263309599818017e-06,
+      "loss": 0.3206,
+      "step": 5747
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4232033392541176e-06,
+      "loss": 0.3661,
+      "step": 5748
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4200768518447524e-06,
+      "loss": 0.3603,
+      "step": 5749
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4169514982924556e-06,
+      "loss": 0.2645,
+      "step": 5750
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4138272791357995e-06,
+      "loss": 0.2908,
+      "step": 5751
+    },
+    {
+      "epoch": 6.63,
+      "grad_norm": 0.0,
+      "learning_rate": 3.410704194913144e-06,
+      "loss": 0.2596,
+      "step": 5752
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4075822461626663e-06,
+      "loss": 0.335,
+      "step": 5753
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.0,
+      "learning_rate": 3.404461433422339e-06,
+      "loss": 0.3602,
+      "step": 5754
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4013417572299446e-06,
+      "loss": 0.2389,
+      "step": 5755
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3982232181230723e-06,
+      "loss": 0.3563,
+      "step": 5756
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.0,
+      "learning_rate": 3.395105816639106e-06,
+      "loss": 0.2552,
+      "step": 5757
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.0,
+      "learning_rate": 3.391989553315236e-06,
+      "loss": 0.36,
+      "step": 5758
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3888744286884657e-06,
+      "loss": 0.3418,
+      "step": 5759
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.0,
+      "learning_rate": 3.385760443295589e-06,
+      "loss": 0.2462,
+      "step": 5760
+    },
+    {
+      "epoch": 6.64,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3826475976732144e-06,
+      "loss": 0.2651,
+      "step": 5761
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3795358923577513e-06,
+      "loss": 0.2504,
+      "step": 5762
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 0.0,
+      "learning_rate": 3.37642532788541e-06,
+      "loss": 0.2756,
+      "step": 5763
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 0.0,
+      "learning_rate": 3.373315904792198e-06,
+      "loss": 0.284,
+      "step": 5764
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3702076236139402e-06,
+      "loss": 0.2745,
+      "step": 5765
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3671004848862565e-06,
+      "loss": 0.2485,
+      "step": 5766
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 0.0,
+      "learning_rate": 3.363994489144575e-06,
+      "loss": 0.3024,
+      "step": 5767
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3608896369241197e-06,
+      "loss": 0.3125,
+      "step": 5768
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 0.0,
+      "learning_rate": 3.357785928759917e-06,
+      "loss": 0.2457,
+      "step": 5769
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 0.0,
+      "learning_rate": 3.354683365186806e-06,
+      "loss": 0.2989,
+      "step": 5770
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3515819467394184e-06,
+      "loss": 0.2458,
+      "step": 5771
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 0.0,
+      "learning_rate": 3.348481673952195e-06,
+      "loss": 0.2645,
+      "step": 5772
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3453825473593805e-06,
+      "loss": 0.2847,
+      "step": 5773
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3422845674950156e-06,
+      "loss": 0.2649,
+      "step": 5774
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 0.0,
+      "learning_rate": 3.339187734892941e-06,
+      "loss": 0.2901,
+      "step": 5775
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3360920500868144e-06,
+      "loss": 0.2293,
+      "step": 5776
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3329975136100787e-06,
+      "loss": 0.2588,
+      "step": 5777
+    },
+    {
+      "epoch": 6.66,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3299041259959906e-06,
+      "loss": 0.2632,
+      "step": 5778
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 0.0,
+      "learning_rate": 3.326811887777607e-06,
+      "loss": 0.2534,
+      "step": 5779
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3237207994877785e-06,
+      "loss": 0.2433,
+      "step": 5780
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 0.0,
+      "learning_rate": 3.320630861659171e-06,
+      "loss": 0.2519,
+      "step": 5781
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3175420748242405e-06,
+      "loss": 0.3026,
+      "step": 5782
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3144544395152457e-06,
+      "loss": 0.4192,
+      "step": 5783
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3113679562642544e-06,
+      "loss": 0.2968,
+      "step": 5784
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 0.0,
+      "learning_rate": 3.308282625603134e-06,
+      "loss": 0.2835,
+      "step": 5785
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3051984480635444e-06,
+      "loss": 0.2747,
+      "step": 5786
+    },
+    {
+      "epoch": 6.67,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3021154241769606e-06,
+      "loss": 0.2892,
+      "step": 5787
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2990335544746434e-06,
+      "loss": 0.2749,
+      "step": 5788
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2959528394876703e-06,
+      "loss": 0.2935,
+      "step": 5789
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2928732797469065e-06,
+      "loss": 0.267,
+      "step": 5790
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2897948757830256e-06,
+      "loss": 0.3014,
+      "step": 5791
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2867176281265054e-06,
+      "loss": 0.2916,
+      "step": 5792
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 0.0,
+      "learning_rate": 3.283641537307615e-06,
+      "loss": 0.2579,
+      "step": 5793
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2805666038564253e-06,
+      "loss": 0.266,
+      "step": 5794
+    },
+    {
+      "epoch": 6.68,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2774928283028153e-06,
+      "loss": 0.3343,
+      "step": 5795
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 0.0,
+      "learning_rate": 3.274420211176462e-06,
+      "loss": 0.3035,
+      "step": 5796
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2713487530068356e-06,
+      "loss": 0.256,
+      "step": 5797
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 0.0,
+      "learning_rate": 3.268278454323218e-06,
+      "loss": 0.3095,
+      "step": 5798
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2652093156546795e-06,
+      "loss": 0.3367,
+      "step": 5799
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2621413375301016e-06,
+      "loss": 0.2469,
+      "step": 5800
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2590745204781537e-06,
+      "loss": 0.3363,
+      "step": 5801
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 0.0,
+      "learning_rate": 3.256008865027317e-06,
+      "loss": 0.2553,
+      "step": 5802
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2529443717058693e-06,
+      "loss": 0.3184,
+      "step": 5803
+    },
+    {
+      "epoch": 6.69,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2498810410418847e-06,
+      "loss": 0.2431,
+      "step": 5804
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.0,
+      "learning_rate": 3.246818873563232e-06,
+      "loss": 0.288,
+      "step": 5805
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2437578697975956e-06,
+      "loss": 0.2743,
+      "step": 5806
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2406980302724422e-06,
+      "loss": 0.3052,
+      "step": 5807
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2376393555150497e-06,
+      "loss": 0.3084,
+      "step": 5808
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2345818460524927e-06,
+      "loss": 0.262,
+      "step": 5809
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.0,
+      "learning_rate": 3.231525502411642e-06,
+      "loss": 0.2431,
+      "step": 5810
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.0,
+      "learning_rate": 3.228470325119164e-06,
+      "loss": 0.245,
+      "step": 5811
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.0,
+      "learning_rate": 3.225416314701537e-06,
+      "loss": 0.2419,
+      "step": 5812
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2223634716850227e-06,
+      "loss": 0.2617,
+      "step": 5813
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2193117965956923e-06,
+      "loss": 0.2943,
+      "step": 5814
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 0.0,
+      "learning_rate": 3.216261289959417e-06,
+      "loss": 0.3375,
+      "step": 5815
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2132119523018556e-06,
+      "loss": 0.2469,
+      "step": 5816
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 0.0,
+      "learning_rate": 3.210163784148478e-06,
+      "loss": 0.2456,
+      "step": 5817
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2071167860245435e-06,
+      "loss": 0.2694,
+      "step": 5818
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 0.0,
+      "learning_rate": 3.20407095845511e-06,
+      "loss": 0.2788,
+      "step": 5819
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2010263019650402e-06,
+      "loss": 0.3488,
+      "step": 5820
+    },
+    {
+      "epoch": 6.71,
+      "grad_norm": 0.0,
+      "learning_rate": 3.197982817078994e-06,
+      "loss": 0.2636,
+      "step": 5821
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.0,
+      "learning_rate": 3.194940504321421e-06,
+      "loss": 0.2794,
+      "step": 5822
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.0,
+      "learning_rate": 3.191899364216581e-06,
+      "loss": 0.319,
+      "step": 5823
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1888593972885194e-06,
+      "loss": 0.3397,
+      "step": 5824
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1858206040610883e-06,
+      "loss": 0.2507,
+      "step": 5825
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1827829850579373e-06,
+      "loss": 0.2951,
+      "step": 5826
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.0,
+      "learning_rate": 3.179746540802506e-06,
+      "loss": 0.2519,
+      "step": 5827
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1767112718180427e-06,
+      "loss": 0.303,
+      "step": 5828
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1736771786275834e-06,
+      "loss": 0.2978,
+      "step": 5829
+    },
+    {
+      "epoch": 6.72,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1706442617539624e-06,
+      "loss": 0.2678,
+      "step": 5830
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1676125217198175e-06,
+      "loss": 0.3343,
+      "step": 5831
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1645819590475825e-06,
+      "loss": 0.3048,
+      "step": 5832
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 0.0,
+      "learning_rate": 3.16155257425948e-06,
+      "loss": 0.2983,
+      "step": 5833
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1585243678775434e-06,
+      "loss": 0.2604,
+      "step": 5834
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 0.0,
+      "learning_rate": 3.155497340423588e-06,
+      "loss": 0.2703,
+      "step": 5835
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 0.0,
+      "learning_rate": 3.152471492419239e-06,
+      "loss": 0.3291,
+      "step": 5836
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 0.0,
+      "learning_rate": 3.149446824385909e-06,
+      "loss": 0.3042,
+      "step": 5837
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1464233368448103e-06,
+      "loss": 0.3456,
+      "step": 5838
+    },
+    {
+      "epoch": 6.73,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1434010303169593e-06,
+      "loss": 0.2414,
+      "step": 5839
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1403799053231565e-06,
+      "loss": 0.2375,
+      "step": 5840
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1373599623840024e-06,
+      "loss": 0.2611,
+      "step": 5841
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1343412020199014e-06,
+      "loss": 0.2342,
+      "step": 5842
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1313236247510414e-06,
+      "loss": 0.2533,
+      "step": 5843
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1283072310974184e-06,
+      "loss": 0.2822,
+      "step": 5844
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 0.0,
+      "learning_rate": 3.125292021578822e-06,
+      "loss": 0.2676,
+      "step": 5845
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 0.0,
+      "learning_rate": 3.122277996714829e-06,
+      "loss": 0.2503,
+      "step": 5846
+    },
+    {
+      "epoch": 6.74,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1192651570248232e-06,
+      "loss": 0.2889,
+      "step": 5847
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.116253503027977e-06,
+      "loss": 0.358,
+      "step": 5848
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.11324303524326e-06,
+      "loss": 0.255,
+      "step": 5849
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.110233754189438e-06,
+      "loss": 0.2414,
+      "step": 5850
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.107225660385077e-06,
+      "loss": 0.3239,
+      "step": 5851
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.104218754348527e-06,
+      "loss": 0.3174,
+      "step": 5852
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1012130365979476e-06,
+      "loss": 0.2576,
+      "step": 5853
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0982085076512793e-06,
+      "loss": 0.357,
+      "step": 5854
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0952051680262675e-06,
+      "loss": 0.2849,
+      "step": 5855
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0922030182404527e-06,
+      "loss": 0.2833,
+      "step": 5856
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.089202058811166e-06,
+      "loss": 0.2727,
+      "step": 5857
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.08620229025553e-06,
+      "loss": 0.2784,
+      "step": 5858
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0832037130904748e-06,
+      "loss": 0.2525,
+      "step": 5859
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.08020632783271e-06,
+      "loss": 0.2933,
+      "step": 5860
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0772101349987506e-06,
+      "loss": 0.3358,
+      "step": 5861
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0742151351049055e-06,
+      "loss": 0.3043,
+      "step": 5862
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0712213286672686e-06,
+      "loss": 0.247,
+      "step": 5863
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.068228716201743e-06,
+      "loss": 0.3136,
+      "step": 5864
+    },
+    {
+      "epoch": 6.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.065237298224013e-06,
+      "loss": 0.256,
+      "step": 5865
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0622470752495582e-06,
+      "loss": 0.2814,
+      "step": 5866
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0592580477936606e-06,
+      "loss": 0.3134,
+      "step": 5867
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0562702163713954e-06,
+      "loss": 0.2713,
+      "step": 5868
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0532835814976193e-06,
+      "loss": 0.2967,
+      "step": 5869
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0502981436869983e-06,
+      "loss": 0.2466,
+      "step": 5870
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.047313903453979e-06,
+      "loss": 0.2967,
+      "step": 5871
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0443308613128155e-06,
+      "loss": 0.2571,
+      "step": 5872
+    },
+    {
+      "epoch": 6.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0413490177775397e-06,
+      "loss": 0.2989,
+      "step": 5873
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0383683733619884e-06,
+      "loss": 0.2534,
+      "step": 5874
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.035388928579792e-06,
+      "loss": 0.3126,
+      "step": 5875
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.032410683944368e-06,
+      "loss": 0.2771,
+      "step": 5876
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.029433639968925e-06,
+      "loss": 0.3807,
+      "step": 5877
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0264577971664766e-06,
+      "loss": 0.3762,
+      "step": 5878
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.023483156049817e-06,
+      "loss": 0.2925,
+      "step": 5879
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.020509717131539e-06,
+      "loss": 0.2475,
+      "step": 5880
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.017537480924032e-06,
+      "loss": 0.3328,
+      "step": 5881
+    },
+    {
+      "epoch": 6.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0145664479394686e-06,
+      "loss": 0.2956,
+      "step": 5882
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 0.0,
+      "learning_rate": 3.011596618689825e-06,
+      "loss": 0.3274,
+      "step": 5883
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 0.0,
+      "learning_rate": 3.008627993686858e-06,
+      "loss": 0.3227,
+      "step": 5884
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0056605734421253e-06,
+      "loss": 0.2348,
+      "step": 5885
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 0.0,
+      "learning_rate": 3.00269435846698e-06,
+      "loss": 0.242,
+      "step": 5886
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9997293492725577e-06,
+      "loss": 0.2342,
+      "step": 5887
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9967655463697875e-06,
+      "loss": 0.3617,
+      "step": 5888
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.993802950269402e-06,
+      "loss": 0.3265,
+      "step": 5889
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.990841561481911e-06,
+      "loss": 0.2678,
+      "step": 5890
+    },
+    {
+      "epoch": 6.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9878813805176252e-06,
+      "loss": 0.2925,
+      "step": 5891
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.984922407886649e-06,
+      "loss": 0.3667,
+      "step": 5892
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9819646440988713e-06,
+      "loss": 0.2828,
+      "step": 5893
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.979008089663974e-06,
+      "loss": 0.3252,
+      "step": 5894
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.976052745091438e-06,
+      "loss": 0.2657,
+      "step": 5895
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9730986108905236e-06,
+      "loss": 0.2623,
+      "step": 5896
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9701456875702937e-06,
+      "loss": 0.3028,
+      "step": 5897
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9671939756396005e-06,
+      "loss": 0.277,
+      "step": 5898
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9642434756070793e-06,
+      "loss": 0.2784,
+      "step": 5899
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9612941879811684e-06,
+      "loss": 0.2603,
+      "step": 5900
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9583461132700875e-06,
+      "loss": 0.2995,
+      "step": 5901
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9553992519818484e-06,
+      "loss": 0.2598,
+      "step": 5902
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9524536046242603e-06,
+      "loss": 0.223,
+      "step": 5903
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9495091717049205e-06,
+      "loss": 0.3236,
+      "step": 5904
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.946565953731211e-06,
+      "loss": 0.2907,
+      "step": 5905
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9436239512103147e-06,
+      "loss": 0.3306,
+      "step": 5906
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.940683164649194e-06,
+      "loss": 0.2489,
+      "step": 5907
+    },
+    {
+      "epoch": 6.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.937743594554614e-06,
+      "loss": 0.33,
+      "step": 5908
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.934805241433115e-06,
+      "loss": 0.361,
+      "step": 5909
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9318681057910416e-06,
+      "loss": 0.3011,
+      "step": 5910
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.3748,
+      "step": 5911
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9259974889694832e-06,
+      "loss": 0.2468,
+      "step": 5912
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9230640088016204e-06,
+      "loss": 0.2203,
+      "step": 5913
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.92013174813644e-06,
+      "loss": 0.2735,
+      "step": 5914
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9172007074792342e-06,
+      "loss": 0.2877,
+      "step": 5915
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9142708873350766e-06,
+      "loss": 0.4099,
+      "step": 5916
+    },
+    {
+      "epoch": 6.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9113422882088403e-06,
+      "loss": 0.3596,
+      "step": 5917
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9084149106051797e-06,
+      "loss": 0.2738,
+      "step": 5918
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9054887550285483e-06,
+      "loss": 0.3367,
+      "step": 5919
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 0.0,
+      "learning_rate": 2.902563821983175e-06,
+      "loss": 0.2493,
+      "step": 5920
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8996401119730923e-06,
+      "loss": 0.2734,
+      "step": 5921
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 0.0,
+      "learning_rate": 2.896717625502117e-06,
+      "loss": 0.2788,
+      "step": 5922
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8937963630738517e-06,
+      "loss": 0.2779,
+      "step": 5923
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 0.0,
+      "learning_rate": 2.890876325191687e-06,
+      "loss": 0.3377,
+      "step": 5924
+    },
+    {
+      "epoch": 6.83,
+      "grad_norm": 0.0,
+      "learning_rate": 2.887957512358813e-06,
+      "loss": 0.2909,
+      "step": 5925
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8850399250781957e-06,
+      "loss": 0.2197,
+      "step": 5926
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 0.0,
+      "learning_rate": 2.882123563852598e-06,
+      "loss": 0.2978,
+      "step": 5927
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8792084291845713e-06,
+      "loss": 0.2924,
+      "step": 5928
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 0.0,
+      "learning_rate": 2.87629452157645e-06,
+      "loss": 0.2637,
+      "step": 5929
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 0.0,
+      "learning_rate": 2.873381841530365e-06,
+      "loss": 0.2679,
+      "step": 5930
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 0.0,
+      "learning_rate": 2.87047038954823e-06,
+      "loss": 0.2991,
+      "step": 5931
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 0.0,
+      "learning_rate": 2.867560166131744e-06,
+      "loss": 0.3612,
+      "step": 5932
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 0.0,
+      "learning_rate": 2.864651171782402e-06,
+      "loss": 0.2905,
+      "step": 5933
+    },
+    {
+      "epoch": 6.84,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8617434070014883e-06,
+      "loss": 0.3945,
+      "step": 5934
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.0,
+      "learning_rate": 2.858836872290064e-06,
+      "loss": 0.2698,
+      "step": 5935
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8559315681489908e-06,
+      "loss": 0.3281,
+      "step": 5936
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8530274950789095e-06,
+      "loss": 0.2807,
+      "step": 5937
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8501246535802506e-06,
+      "loss": 0.3165,
+      "step": 5938
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8472230441532365e-06,
+      "loss": 0.2753,
+      "step": 5939
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8443226672978775e-06,
+      "loss": 0.2438,
+      "step": 5940
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.0,
+      "learning_rate": 2.841423523513961e-06,
+      "loss": 0.2746,
+      "step": 5941
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8385256133010775e-06,
+      "loss": 0.3074,
+      "step": 5942
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 0.0,
+      "learning_rate": 2.835628937158591e-06,
+      "loss": 0.3548,
+      "step": 5943
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 0.0,
+      "learning_rate": 2.83273349558566e-06,
+      "loss": 0.2133,
+      "step": 5944
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8298392890812344e-06,
+      "loss": 0.3257,
+      "step": 5945
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 0.0,
+      "learning_rate": 2.826946318144037e-06,
+      "loss": 0.2796,
+      "step": 5946
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8240545832725963e-06,
+      "loss": 0.238,
+      "step": 5947
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 0.0,
+      "learning_rate": 2.821164084965212e-06,
+      "loss": 0.2538,
+      "step": 5948
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8182748237199755e-06,
+      "loss": 0.2322,
+      "step": 5949
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8153868000347683e-06,
+      "loss": 0.2778,
+      "step": 5950
+    },
+    {
+      "epoch": 6.86,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8125000144072602e-06,
+      "loss": 0.3002,
+      "step": 5951
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 0.0,
+      "learning_rate": 2.809614467334899e-06,
+      "loss": 0.2655,
+      "step": 5952
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 0.0,
+      "learning_rate": 2.806730159314929e-06,
+      "loss": 0.3319,
+      "step": 5953
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8038470908443717e-06,
+      "loss": 0.2775,
+      "step": 5954
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8009652624200436e-06,
+      "loss": 0.269,
+      "step": 5955
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 0.0,
+      "learning_rate": 2.798084674538538e-06,
+      "loss": 0.2729,
+      "step": 5956
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7952053276962444e-06,
+      "loss": 0.2611,
+      "step": 5957
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 0.0,
+      "learning_rate": 2.792327222389335e-06,
+      "loss": 0.2815,
+      "step": 5958
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7894503591137656e-06,
+      "loss": 0.2923,
+      "step": 5959
+    },
+    {
+      "epoch": 6.87,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7865747383652743e-06,
+      "loss": 0.2355,
+      "step": 5960
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7837003606393987e-06,
+      "loss": 0.3157,
+      "step": 5961
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.0,
+      "learning_rate": 2.780827226431445e-06,
+      "loss": 0.2665,
+      "step": 5962
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7779553362365184e-06,
+      "loss": 0.4274,
+      "step": 5963
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7750846905495076e-06,
+      "loss": 0.2503,
+      "step": 5964
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7722152898650788e-06,
+      "loss": 0.2536,
+      "step": 5965
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7693471346776945e-06,
+      "loss": 0.2685,
+      "step": 5966
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.0,
+      "learning_rate": 2.766480225481595e-06,
+      "loss": 0.3109,
+      "step": 5967
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.0,
+      "learning_rate": 2.763614562770803e-06,
+      "loss": 0.3027,
+      "step": 5968
+    },
+    {
+      "epoch": 6.88,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7607501470391384e-06,
+      "loss": 0.2448,
+      "step": 5969
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7578869787801985e-06,
+      "loss": 0.3503,
+      "step": 5970
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 0.0,
+      "learning_rate": 2.755025058487364e-06,
+      "loss": 0.275,
+      "step": 5971
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 0.0,
+      "learning_rate": 2.752164386653806e-06,
+      "loss": 0.3177,
+      "step": 5972
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7493049637724724e-06,
+      "loss": 0.2675,
+      "step": 5973
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 0.0,
+      "learning_rate": 2.746446790336106e-06,
+      "loss": 0.3188,
+      "step": 5974
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7435898668372296e-06,
+      "loss": 0.2912,
+      "step": 5975
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7407341937681485e-06,
+      "loss": 0.2916,
+      "step": 5976
+    },
+    {
+      "epoch": 6.89,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7378797716209506e-06,
+      "loss": 0.253,
+      "step": 5977
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7350266008875194e-06,
+      "loss": 0.2739,
+      "step": 5978
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7321746820595084e-06,
+      "loss": 0.2891,
+      "step": 5979
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7293240156283653e-06,
+      "loss": 0.3515,
+      "step": 5980
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7264746020853217e-06,
+      "loss": 0.2924,
+      "step": 5981
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7236264419213852e-06,
+      "loss": 0.3246,
+      "step": 5982
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 0.0,
+      "learning_rate": 2.720779535627359e-06,
+      "loss": 0.2777,
+      "step": 5983
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 0.0,
+      "learning_rate": 2.717933883693821e-06,
+      "loss": 0.2675,
+      "step": 5984
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 0.0,
+      "learning_rate": 2.715089486611132e-06,
+      "loss": 0.2555,
+      "step": 5985
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7122463448694446e-06,
+      "loss": 0.3274,
+      "step": 5986
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 0.0,
+      "learning_rate": 2.709404458958693e-06,
+      "loss": 0.3014,
+      "step": 5987
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 0.0,
+      "learning_rate": 2.706563829368587e-06,
+      "loss": 0.3595,
+      "step": 5988
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7037244565886335e-06,
+      "loss": 0.2937,
+      "step": 5989
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 0.0,
+      "learning_rate": 2.700886341108108e-06,
+      "loss": 0.2882,
+      "step": 5990
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 0.0,
+      "learning_rate": 2.698049483416083e-06,
+      "loss": 0.3382,
+      "step": 5991
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 0.0,
+      "learning_rate": 2.695213884001402e-06,
+      "loss": 0.2746,
+      "step": 5992
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 0.0,
+      "learning_rate": 2.692379543352699e-06,
+      "loss": 0.4118,
+      "step": 5993
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6895464619583946e-06,
+      "loss": 0.2722,
+      "step": 5994
+    },
+    {
+      "epoch": 6.91,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6867146403066833e-06,
+      "loss": 0.2909,
+      "step": 5995
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 0.0,
+      "learning_rate": 2.683884078885545e-06,
+      "loss": 0.2784,
+      "step": 5996
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 0.0,
+      "learning_rate": 2.681054778182748e-06,
+      "loss": 0.2597,
+      "step": 5997
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6782267386858352e-06,
+      "loss": 0.3103,
+      "step": 5998
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 0.0,
+      "learning_rate": 2.675399960882138e-06,
+      "loss": 0.3215,
+      "step": 5999
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 0.0,
+      "learning_rate": 2.672574445258773e-06,
+      "loss": 0.2382,
+      "step": 6000
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 0.0,
+      "learning_rate": 2.669750192302628e-06,
+      "loss": 0.2838,
+      "step": 6001
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6669272025003877e-06,
+      "loss": 0.2394,
+      "step": 6002
+    },
+    {
+      "epoch": 6.92,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6641054763385044e-06,
+      "loss": 0.2399,
+      "step": 6003
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6612850143032233e-06,
+      "loss": 0.2887,
+      "step": 6004
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.658465816880572e-06,
+      "loss": 0.2531,
+      "step": 6005
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6556478845563525e-06,
+      "loss": 0.3412,
+      "step": 6006
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.652831217816151e-06,
+      "loss": 0.2778,
+      "step": 6007
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6500158171453425e-06,
+      "loss": 0.3018,
+      "step": 6008
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.647201683029075e-06,
+      "loss": 0.2643,
+      "step": 6009
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6443888159522825e-06,
+      "loss": 0.315,
+      "step": 6010
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6415772163996845e-06,
+      "loss": 0.2793,
+      "step": 6011
+    },
+    {
+      "epoch": 6.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6387668848557724e-06,
+      "loss": 0.2845,
+      "step": 6012
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6359578218048287e-06,
+      "loss": 0.2806,
+      "step": 6013
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.633150027730912e-06,
+      "loss": 0.3108,
+      "step": 6014
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.630343503117859e-06,
+      "loss": 0.2652,
+      "step": 6015
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6275382484492973e-06,
+      "loss": 0.2894,
+      "step": 6016
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6247342642086304e-06,
+      "loss": 0.3366,
+      "step": 6017
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6219315508790387e-06,
+      "loss": 0.3316,
+      "step": 6018
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.619130108943494e-06,
+      "loss": 0.3094,
+      "step": 6019
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6163299388847384e-06,
+      "loss": 0.3125,
+      "step": 6020
+    },
+    {
+      "epoch": 6.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6135310411852977e-06,
+      "loss": 0.2927,
+      "step": 6021
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6107334163274823e-06,
+      "loss": 0.3314,
+      "step": 6022
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6079370647933834e-06,
+      "loss": 0.2434,
+      "step": 6023
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6051419870648643e-06,
+      "loss": 0.3399,
+      "step": 6024
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 0.0,
+      "learning_rate": 2.602348183623582e-06,
+      "loss": 0.2812,
+      "step": 6025
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5995556549509604e-06,
+      "loss": 0.3213,
+      "step": 6026
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5967644015282146e-06,
+      "loss": 0.3232,
+      "step": 6027
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 0.0,
+      "learning_rate": 2.593974423836331e-06,
+      "loss": 0.2659,
+      "step": 6028
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 0.0,
+      "learning_rate": 2.591185722356082e-06,
+      "loss": 0.2887,
+      "step": 6029
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5883982975680242e-06,
+      "loss": 0.3133,
+      "step": 6030
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5856121499524832e-06,
+      "loss": 0.267,
+      "step": 6031
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5828272799895683e-06,
+      "loss": 0.2515,
+      "step": 6032
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.0,
+      "learning_rate": 2.580043688159172e-06,
+      "loss": 0.2323,
+      "step": 6033
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.0,
+      "learning_rate": 2.577261374940968e-06,
+      "loss": 0.3626,
+      "step": 6034
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5744803408144026e-06,
+      "loss": 0.2719,
+      "step": 6035
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.0,
+      "learning_rate": 2.57170058625871e-06,
+      "loss": 0.2488,
+      "step": 6036
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.0,
+      "learning_rate": 2.568922111752893e-06,
+      "loss": 0.3017,
+      "step": 6037
+    },
+    {
+      "epoch": 6.96,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5661449177757467e-06,
+      "loss": 0.2779,
+      "step": 6038
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5633690048058346e-06,
+      "loss": 0.2409,
+      "step": 6039
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5605943733215044e-06,
+      "loss": 0.2803,
+      "step": 6040
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5578210238008883e-06,
+      "loss": 0.2654,
+      "step": 6041
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 0.0,
+      "learning_rate": 2.555048956721886e-06,
+      "loss": 0.3635,
+      "step": 6042
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5522781725621814e-06,
+      "loss": 0.3164,
+      "step": 6043
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5495086717992447e-06,
+      "loss": 0.2681,
+      "step": 6044
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5467404549103094e-06,
+      "loss": 0.2371,
+      "step": 6045
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5439735223724006e-06,
+      "loss": 0.3093,
+      "step": 6046
+    },
+    {
+      "epoch": 6.97,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5412078746623225e-06,
+      "loss": 0.3207,
+      "step": 6047
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5384435122566463e-06,
+      "loss": 0.3067,
+      "step": 6048
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5356804356317345e-06,
+      "loss": 0.3307,
+      "step": 6049
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5329186452637213e-06,
+      "loss": 0.3062,
+      "step": 6050
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 0.0,
+      "learning_rate": 2.530158141628515e-06,
+      "loss": 0.2756,
+      "step": 6051
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5273989252018128e-06,
+      "loss": 0.2929,
+      "step": 6052
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5246409964590855e-06,
+      "loss": 0.2567,
+      "step": 6053
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 0.0,
+      "learning_rate": 2.521884355875578e-06,
+      "loss": 0.3264,
+      "step": 6054
+    },
+    {
+      "epoch": 6.98,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5191290039263203e-06,
+      "loss": 0.3386,
+      "step": 6055
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 0.0,
+      "learning_rate": 2.516374941086115e-06,
+      "loss": 0.3366,
+      "step": 6056
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5136221678295413e-06,
+      "loss": 0.3862,
+      "step": 6057
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5108706846309607e-06,
+      "loss": 0.2487,
+      "step": 6058
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 0.0,
+      "learning_rate": 2.508120491964512e-06,
+      "loss": 0.3117,
+      "step": 6059
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 0.0,
+      "learning_rate": 2.505371590304113e-06,
+      "loss": 0.3269,
+      "step": 6060
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 0.0,
+      "learning_rate": 2.502623980123453e-06,
+      "loss": 0.2716,
+      "step": 6061
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4998776618959997e-06,
+      "loss": 0.2785,
+      "step": 6062
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4971326360950032e-06,
+      "loss": 0.2453,
+      "step": 6063
+    },
+    {
+      "epoch": 6.99,
+      "grad_norm": 0.0,
+      "learning_rate": 2.494388903193492e-06,
+      "loss": 0.3141,
+      "step": 6064
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2.491646463664261e-06,
+      "loss": 0.2753,
+      "step": 6065
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2.488905317979896e-06,
+      "loss": 0.2598,
+      "step": 6066
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2.486165466612751e-06,
+      "loss": 0.3195,
+      "step": 6067
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2.483426910034955e-06,
+      "loss": 0.2755,
+      "step": 6068
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4806896487184207e-06,
+      "loss": 0.2625,
+      "step": 6069
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4779536831348392e-06,
+      "loss": 0.2815,
+      "step": 6070
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4752190137556676e-06,
+      "loss": 0.2227,
+      "step": 6071
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4724856410521513e-06,
+      "loss": 0.2455,
+      "step": 6072
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0,
+      "learning_rate": 2.469753565495303e-06,
+      "loss": 0.2947,
+      "step": 6073
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4670227875559196e-06,
+      "loss": 0.2307,
+      "step": 6074
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.464293307704566e-06,
+      "loss": 0.2614,
+      "step": 6075
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4615651264115904e-06,
+      "loss": 0.2685,
+      "step": 6076
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4588382441471193e-06,
+      "loss": 0.2304,
+      "step": 6077
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.456112661381046e-06,
+      "loss": 0.252,
+      "step": 6078
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4533883785830438e-06,
+      "loss": 0.3573,
+      "step": 6079
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.450665396222568e-06,
+      "loss": 0.2445,
+      "step": 6080
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.447943714768839e-06,
+      "loss": 0.2582,
+      "step": 6081
+    },
+    {
+      "epoch": 7.01,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4452233346908626e-06,
+      "loss": 0.2539,
+      "step": 6082
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.2673,
+      "step": 6083
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 0.0,
+      "learning_rate": 2.439786480537054e-06,
+      "loss": 0.225,
+      "step": 6084
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 0.0,
+      "learning_rate": 2.437070007398107e-06,
+      "loss": 0.2524,
+      "step": 6085
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 0.0,
+      "learning_rate": 2.434354837508676e-06,
+      "loss": 0.2502,
+      "step": 6086
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4316409713366353e-06,
+      "loss": 0.3203,
+      "step": 6087
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4289284093496533e-06,
+      "loss": 0.3295,
+      "step": 6088
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4262171520151557e-06,
+      "loss": 0.3513,
+      "step": 6089
+    },
+    {
+      "epoch": 7.02,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4235071998003434e-06,
+      "loss": 0.2436,
+      "step": 6090
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4207985531722034e-06,
+      "loss": 0.3017,
+      "step": 6091
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4180912125974865e-06,
+      "loss": 0.2786,
+      "step": 6092
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.415385178542725e-06,
+      "loss": 0.2741,
+      "step": 6093
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.412680451474231e-06,
+      "loss": 0.3051,
+      "step": 6094
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.409977031858075e-06,
+      "loss": 0.2685,
+      "step": 6095
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.407274920160121e-06,
+      "loss": 0.2945,
+      "step": 6096
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4045741168459944e-06,
+      "loss": 0.2517,
+      "step": 6097
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4018746223810974e-06,
+      "loss": 0.3324,
+      "step": 6098
+    },
+    {
+      "epoch": 7.03,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3991764372306113e-06,
+      "loss": 0.2873,
+      "step": 6099
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3964795618594917e-06,
+      "loss": 0.2763,
+      "step": 6100
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.393783996732462e-06,
+      "loss": 0.2298,
+      "step": 6101
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.391089742314028e-06,
+      "loss": 0.2946,
+      "step": 6102
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.388396799068463e-06,
+      "loss": 0.252,
+      "step": 6103
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3857051674598153e-06,
+      "loss": 0.2622,
+      "step": 6104
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.38301484795191e-06,
+      "loss": 0.2905,
+      "step": 6105
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3803258410083496e-06,
+      "loss": 0.2627,
+      "step": 6106
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.377638147092497e-06,
+      "loss": 0.3457,
+      "step": 6107
+    },
+    {
+      "epoch": 7.04,
+      "grad_norm": 0.0,
+      "learning_rate": 2.374951766667506e-06,
+      "loss": 0.2293,
+      "step": 6108
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3722667001962898e-06,
+      "loss": 0.2438,
+      "step": 6109
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3695829481415444e-06,
+      "loss": 0.2482,
+      "step": 6110
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 0.0,
+      "learning_rate": 2.366900510965733e-06,
+      "loss": 0.2327,
+      "step": 6111
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 0.0,
+      "learning_rate": 2.364219389131095e-06,
+      "loss": 0.2562,
+      "step": 6112
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 0.0,
+      "learning_rate": 2.361539583099649e-06,
+      "loss": 0.3198,
+      "step": 6113
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 0.0,
+      "learning_rate": 2.358861093333177e-06,
+      "loss": 0.2314,
+      "step": 6114
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3561839202932344e-06,
+      "loss": 0.26,
+      "step": 6115
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 0.0,
+      "learning_rate": 2.35350806444116e-06,
+      "loss": 0.2814,
+      "step": 6116
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3508335262380542e-06,
+      "loss": 0.2543,
+      "step": 6117
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 0.0,
+      "learning_rate": 2.348160306144799e-06,
+      "loss": 0.2348,
+      "step": 6118
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3454884046220463e-06,
+      "loss": 0.2436,
+      "step": 6119
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3428178221302143e-06,
+      "loss": 0.2389,
+      "step": 6120
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 0.0,
+      "learning_rate": 2.340148559129507e-06,
+      "loss": 0.3206,
+      "step": 6121
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 0.0,
+      "learning_rate": 2.337480616079888e-06,
+      "loss": 0.236,
+      "step": 6122
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3348139934411008e-06,
+      "loss": 0.2922,
+      "step": 6123
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 0.0,
+      "learning_rate": 2.332148691672662e-06,
+      "loss": 0.2383,
+      "step": 6124
+    },
+    {
+      "epoch": 7.06,
+      "grad_norm": 0.0,
+      "learning_rate": 2.329484711233858e-06,
+      "loss": 0.2328,
+      "step": 6125
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3268220525837436e-06,
+      "loss": 0.2449,
+      "step": 6126
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3241607161811553e-06,
+      "loss": 0.275,
+      "step": 6127
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3215007024846912e-06,
+      "loss": 0.2966,
+      "step": 6128
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.0,
+      "learning_rate": 2.31884201195273e-06,
+      "loss": 0.2826,
+      "step": 6129
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.0,
+      "learning_rate": 2.316184645043421e-06,
+      "loss": 0.2726,
+      "step": 6130
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3135286022146785e-06,
+      "loss": 0.2423,
+      "step": 6131
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3108738839241993e-06,
+      "loss": 0.2631,
+      "step": 6132
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3082204906294448e-06,
+      "loss": 0.2776,
+      "step": 6133
+    },
+    {
+      "epoch": 7.07,
+      "grad_norm": 0.0,
+      "learning_rate": 2.305568422787645e-06,
+      "loss": 0.2764,
+      "step": 6134
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3029176808558097e-06,
+      "loss": 0.2474,
+      "step": 6135
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3002682652907184e-06,
+      "loss": 0.277,
+      "step": 6136
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.0,
+      "learning_rate": 2.297620176548916e-06,
+      "loss": 0.2671,
+      "step": 6137
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.0,
+      "learning_rate": 2.294973415086729e-06,
+      "loss": 0.2419,
+      "step": 6138
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.0,
+      "learning_rate": 2.292327981360245e-06,
+      "loss": 0.2296,
+      "step": 6139
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.0,
+      "learning_rate": 2.289683875825326e-06,
+      "loss": 0.2863,
+      "step": 6140
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2870410989376067e-06,
+      "loss": 0.2741,
+      "step": 6141
+    },
+    {
+      "epoch": 7.08,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2843996511524937e-06,
+      "loss": 0.2459,
+      "step": 6142
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2817595329251663e-06,
+      "loss": 0.2442,
+      "step": 6143
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2791207447105677e-06,
+      "loss": 0.2567,
+      "step": 6144
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2764832869634134e-06,
+      "loss": 0.2535,
+      "step": 6145
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 0.0,
+      "learning_rate": 2.273847160138198e-06,
+      "loss": 0.3242,
+      "step": 6146
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 0.0,
+      "learning_rate": 2.271212364689176e-06,
+      "loss": 0.2239,
+      "step": 6147
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 0.0,
+      "learning_rate": 2.268578901070377e-06,
+      "loss": 0.2945,
+      "step": 6148
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2659467697356074e-06,
+      "loss": 0.2518,
+      "step": 6149
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2633159711384336e-06,
+      "loss": 0.2491,
+      "step": 6150
+    },
+    {
+      "epoch": 7.09,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2606865057321938e-06,
+      "loss": 0.2638,
+      "step": 6151
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2580583739700023e-06,
+      "loss": 0.288,
+      "step": 6152
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 0.0,
+      "learning_rate": 2.255431576304744e-06,
+      "loss": 0.2523,
+      "step": 6153
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2528061131890657e-06,
+      "loss": 0.296,
+      "step": 6154
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2501819850753925e-06,
+      "loss": 0.2542,
+      "step": 6155
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 0.0,
+      "learning_rate": 2.247559192415911e-06,
+      "loss": 0.3192,
+      "step": 6156
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 0.0,
+      "learning_rate": 2.244937735662589e-06,
+      "loss": 0.2915,
+      "step": 6157
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2423176152671535e-06,
+      "loss": 0.267,
+      "step": 6158
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 0.0,
+      "learning_rate": 2.239698831681105e-06,
+      "loss": 0.2892,
+      "step": 6159
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2370813853557194e-06,
+      "loss": 0.249,
+      "step": 6160
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2344652767420337e-06,
+      "loss": 0.2618,
+      "step": 6161
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 0.0,
+      "learning_rate": 2.231850506290856e-06,
+      "loss": 0.2814,
+      "step": 6162
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 0.0,
+      "learning_rate": 2.229237074452768e-06,
+      "loss": 0.2539,
+      "step": 6163
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 0.0,
+      "learning_rate": 2.226624981678115e-06,
+      "loss": 0.2585,
+      "step": 6164
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2240142284170165e-06,
+      "loss": 0.251,
+      "step": 6165
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 0.0,
+      "learning_rate": 2.221404815119361e-06,
+      "loss": 0.2701,
+      "step": 6166
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2187967422348023e-06,
+      "loss": 0.2354,
+      "step": 6167
+    },
+    {
+      "epoch": 7.11,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2161900102127688e-06,
+      "loss": 0.2723,
+      "step": 6168
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.0,
+      "learning_rate": 2.213584619502451e-06,
+      "loss": 0.3249,
+      "step": 6169
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.0,
+      "learning_rate": 2.21098057055281e-06,
+      "loss": 0.2275,
+      "step": 6170
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2083778638125796e-06,
+      "loss": 0.3117,
+      "step": 6171
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.0,
+      "learning_rate": 2.205776499730263e-06,
+      "loss": 0.2287,
+      "step": 6172
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2031764787541244e-06,
+      "loss": 0.2809,
+      "step": 6173
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2005778013322067e-06,
+      "loss": 0.233,
+      "step": 6174
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1979804679123108e-06,
+      "loss": 0.2729,
+      "step": 6175
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1953844789420108e-06,
+      "loss": 0.2515,
+      "step": 6176
+    },
+    {
+      "epoch": 7.12,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1927898348686515e-06,
+      "loss": 0.2713,
+      "step": 6177
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 0.0,
+      "learning_rate": 2.190196536139344e-06,
+      "loss": 0.2536,
+      "step": 6178
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1876045832009694e-06,
+      "loss": 0.2541,
+      "step": 6179
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 0.0,
+      "learning_rate": 2.185013976500173e-06,
+      "loss": 0.3306,
+      "step": 6180
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1824247164833656e-06,
+      "loss": 0.302,
+      "step": 6181
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1798368035967356e-06,
+      "loss": 0.273,
+      "step": 6182
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 0.0,
+      "learning_rate": 2.177250238286235e-06,
+      "loss": 0.2764,
+      "step": 6183
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1746650209975785e-06,
+      "loss": 0.2311,
+      "step": 6184
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1720811521762554e-06,
+      "loss": 0.2414,
+      "step": 6185
+    },
+    {
+      "epoch": 7.13,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1694986322675204e-06,
+      "loss": 0.2319,
+      "step": 6186
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 0.0,
+      "learning_rate": 2.16691746171639e-06,
+      "loss": 0.2764,
+      "step": 6187
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1643376409676563e-06,
+      "loss": 0.3182,
+      "step": 6188
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 0.0,
+      "learning_rate": 2.161759170465878e-06,
+      "loss": 0.2668,
+      "step": 6189
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1591820506553794e-06,
+      "loss": 0.2414,
+      "step": 6190
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 0.0,
+      "learning_rate": 2.15660628198025e-06,
+      "loss": 0.2644,
+      "step": 6191
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 0.0,
+      "learning_rate": 2.154031864884345e-06,
+      "loss": 0.2396,
+      "step": 6192
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1514587998112956e-06,
+      "loss": 0.2851,
+      "step": 6193
+    },
+    {
+      "epoch": 7.14,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1488870872044887e-06,
+      "loss": 0.2272,
+      "step": 6194
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1463167275070863e-06,
+      "loss": 0.2538,
+      "step": 6195
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 0.0,
+      "learning_rate": 2.143747721162016e-06,
+      "loss": 0.2223,
+      "step": 6196
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 0.0,
+      "learning_rate": 2.141180068611971e-06,
+      "loss": 0.2571,
+      "step": 6197
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1386137702994057e-06,
+      "loss": 0.2715,
+      "step": 6198
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1360488266665534e-06,
+      "loss": 0.235,
+      "step": 6199
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1334852381554005e-06,
+      "loss": 0.2875,
+      "step": 6200
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 0.0,
+      "learning_rate": 2.13092300520771e-06,
+      "loss": 0.2366,
+      "step": 6201
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 0.0,
+      "learning_rate": 2.12836212826501e-06,
+      "loss": 0.2255,
+      "step": 6202
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 0.0,
+      "learning_rate": 2.125802607768588e-06,
+      "loss": 0.2527,
+      "step": 6203
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.0,
+      "learning_rate": 2.123244444159507e-06,
+      "loss": 0.2676,
+      "step": 6204
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1206876378785877e-06,
+      "loss": 0.2963,
+      "step": 6205
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1181321893664187e-06,
+      "loss": 0.3533,
+      "step": 6206
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1155780990633644e-06,
+      "loss": 0.2622,
+      "step": 6207
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1130253674095435e-06,
+      "loss": 0.2606,
+      "step": 6208
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1104739948448415e-06,
+      "loss": 0.2313,
+      "step": 6209
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.0,
+      "learning_rate": 2.107923981808918e-06,
+      "loss": 0.2857,
+      "step": 6210
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1053753287411895e-06,
+      "loss": 0.2828,
+      "step": 6211
+    },
+    {
+      "epoch": 7.16,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1028280360808405e-06,
+      "loss": 0.2215,
+      "step": 6212
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 0.0,
+      "learning_rate": 2.100282104266829e-06,
+      "loss": 0.2802,
+      "step": 6213
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0977375337378635e-06,
+      "loss": 0.2447,
+      "step": 6214
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0951943249324334e-06,
+      "loss": 0.2596,
+      "step": 6215
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0926524782887837e-06,
+      "loss": 0.2378,
+      "step": 6216
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 0.0,
+      "learning_rate": 2.090111994244923e-06,
+      "loss": 0.247,
+      "step": 6217
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0875728732386336e-06,
+      "loss": 0.2645,
+      "step": 6218
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 0.0,
+      "learning_rate": 2.08503511570746e-06,
+      "loss": 0.2944,
+      "step": 6219
+    },
+    {
+      "epoch": 7.17,
+      "grad_norm": 0.0,
+      "learning_rate": 2.082498722088706e-06,
+      "loss": 0.3044,
+      "step": 6220
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 0.0,
+      "learning_rate": 2.07996369281945e-06,
+      "loss": 0.2812,
+      "step": 6221
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 0.0,
+      "learning_rate": 2.077430028336527e-06,
+      "loss": 0.2607,
+      "step": 6222
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 0.0,
+      "learning_rate": 2.074897729076536e-06,
+      "loss": 0.2214,
+      "step": 6223
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0723667954758498e-06,
+      "loss": 0.2785,
+      "step": 6224
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 0.0,
+      "learning_rate": 2.069837227970599e-06,
+      "loss": 0.2354,
+      "step": 6225
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0673090269966846e-06,
+      "loss": 0.2381,
+      "step": 6226
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 0.0,
+      "learning_rate": 2.064782192989765e-06,
+      "loss": 0.2451,
+      "step": 6227
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 0.0,
+      "learning_rate": 2.062256726385261e-06,
+      "loss": 0.2454,
+      "step": 6228
+    },
+    {
+      "epoch": 7.18,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0597326276183716e-06,
+      "loss": 0.2975,
+      "step": 6229
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0572098971240427e-06,
+      "loss": 0.2688,
+      "step": 6230
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0546885353369984e-06,
+      "loss": 0.2227,
+      "step": 6231
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0521685426917227e-06,
+      "loss": 0.2399,
+      "step": 6232
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0496499196224605e-06,
+      "loss": 0.2215,
+      "step": 6233
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0471326665632206e-06,
+      "loss": 0.274,
+      "step": 6234
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0446167839477815e-06,
+      "loss": 0.231,
+      "step": 6235
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0421022722096784e-06,
+      "loss": 0.2928,
+      "step": 6236
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0395891317822146e-06,
+      "loss": 0.288,
+      "step": 6237
+    },
+    {
+      "epoch": 7.19,
+      "grad_norm": 0.0,
+      "learning_rate": 2.037077363098461e-06,
+      "loss": 0.2228,
+      "step": 6238
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0345669665912402e-06,
+      "loss": 0.2454,
+      "step": 6239
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0320579426931517e-06,
+      "loss": 0.2665,
+      "step": 6240
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0295502918365473e-06,
+      "loss": 0.2746,
+      "step": 6241
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0270440144535497e-06,
+      "loss": 0.3015,
+      "step": 6242
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0245391109760437e-06,
+      "loss": 0.3376,
+      "step": 6243
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.0,
+      "learning_rate": 2.022035581835675e-06,
+      "loss": 0.2547,
+      "step": 6244
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.0,
+      "learning_rate": 2.019533427463851e-06,
+      "loss": 0.378,
+      "step": 6245
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0170326482917492e-06,
+      "loss": 0.216,
+      "step": 6246
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0145332447502995e-06,
+      "loss": 0.2497,
+      "step": 6247
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 0.0,
+      "learning_rate": 2.012035217270206e-06,
+      "loss": 0.2794,
+      "step": 6248
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 0.0,
+      "learning_rate": 2.009538566281931e-06,
+      "loss": 0.3187,
+      "step": 6249
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0070432922156956e-06,
+      "loss": 0.2771,
+      "step": 6250
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0045493955014915e-06,
+      "loss": 0.2589,
+      "step": 6251
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 0.0,
+      "learning_rate": 2.002056876569066e-06,
+      "loss": 0.3071,
+      "step": 6252
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9995657358479293e-06,
+      "loss": 0.3152,
+      "step": 6253
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9970759737673596e-06,
+      "loss": 0.2951,
+      "step": 6254
+    },
+    {
+      "epoch": 7.21,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994587590756397e-06,
+      "loss": 0.2187,
+      "step": 6255
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.992100587243836e-06,
+      "loss": 0.3196,
+      "step": 6256
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.989614963658244e-06,
+      "loss": 0.2881,
+      "step": 6257
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.987130720427943e-06,
+      "loss": 0.2347,
+      "step": 6258
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.984647857981017e-06,
+      "loss": 0.2927,
+      "step": 6259
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.982166376745317e-06,
+      "loss": 0.261,
+      "step": 6260
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9796862771484526e-06,
+      "loss": 0.2868,
+      "step": 6261
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.977207559617802e-06,
+      "loss": 0.2739,
+      "step": 6262
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9747302245804944e-06,
+      "loss": 0.2576,
+      "step": 6263
+    },
+    {
+      "epoch": 7.22,
+      "grad_norm": 0.0,
+      "learning_rate": 1.972254272463423e-06,
+      "loss": 0.3057,
+      "step": 6264
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.969779703693253e-06,
+      "loss": 0.2619,
+      "step": 6265
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.967306518696397e-06,
+      "loss": 0.2515,
+      "step": 6266
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.96483471789904e-06,
+      "loss": 0.2492,
+      "step": 6267
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9623643017271253e-06,
+      "loss": 0.2946,
+      "step": 6268
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.959895270606356e-06,
+      "loss": 0.3081,
+      "step": 6269
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9574276249621925e-06,
+      "loss": 0.2781,
+      "step": 6270
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9549613652198663e-06,
+      "loss": 0.246,
+      "step": 6271
+    },
+    {
+      "epoch": 7.23,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9524964918043633e-06,
+      "loss": 0.2637,
+      "step": 6272
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.950033005140436e-06,
+      "loss": 0.2418,
+      "step": 6273
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9475709056525906e-06,
+      "loss": 0.2446,
+      "step": 6274
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9451101937650963e-06,
+      "loss": 0.2863,
+      "step": 6275
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9426508699019885e-06,
+      "loss": 0.3019,
+      "step": 6276
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9401929344870563e-06,
+      "loss": 0.2915,
+      "step": 6277
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.937736387943854e-06,
+      "loss": 0.2369,
+      "step": 6278
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9352812306956993e-06,
+      "loss": 0.2519,
+      "step": 6279
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.932827463165662e-06,
+      "loss": 0.2357,
+      "step": 6280
+    },
+    {
+      "epoch": 7.24,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9303750857765767e-06,
+      "loss": 0.2455,
+      "step": 6281
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.927924098951043e-06,
+      "loss": 0.3026,
+      "step": 6282
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.925474503111412e-06,
+      "loss": 0.242,
+      "step": 6283
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.923026298679801e-06,
+      "loss": 0.2663,
+      "step": 6284
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.920579486078091e-06,
+      "loss": 0.2714,
+      "step": 6285
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9181340657279114e-06,
+      "loss": 0.2253,
+      "step": 6286
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.915690038050666e-06,
+      "loss": 0.3439,
+      "step": 6287
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9132474034675066e-06,
+      "loss": 0.2302,
+      "step": 6288
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.910806162399349e-06,
+      "loss": 0.2999,
+      "step": 6289
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 0.0,
+      "learning_rate": 1.908366315266871e-06,
+      "loss": 0.2327,
+      "step": 6290
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.905927862490512e-06,
+      "loss": 0.2605,
+      "step": 6291
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9034908044904632e-06,
+      "loss": 0.2562,
+      "step": 6292
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.901055141686684e-06,
+      "loss": 0.2361,
+      "step": 6293
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8986208744988876e-06,
+      "loss": 0.2729,
+      "step": 6294
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.896188003346544e-06,
+      "loss": 0.3559,
+      "step": 6295
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8937565286488967e-06,
+      "loss": 0.2621,
+      "step": 6296
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.891326450824933e-06,
+      "loss": 0.2619,
+      "step": 6297
+    },
+    {
+      "epoch": 7.26,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8888977702934086e-06,
+      "loss": 0.2351,
+      "step": 6298
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8864704874728346e-06,
+      "loss": 0.243,
+      "step": 6299
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8840446027814797e-06,
+      "loss": 0.2723,
+      "step": 6300
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.881620116637375e-06,
+      "loss": 0.2402,
+      "step": 6301
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8791970294583129e-06,
+      "loss": 0.2616,
+      "step": 6302
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8767753416618372e-06,
+      "loss": 0.2837,
+      "step": 6303
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.87435505366526e-06,
+      "loss": 0.2787,
+      "step": 6304
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.871936165885644e-06,
+      "loss": 0.2823,
+      "step": 6305
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8695186787398111e-06,
+      "loss": 0.2351,
+      "step": 6306
+    },
+    {
+      "epoch": 7.27,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8671025926443464e-06,
+      "loss": 0.2607,
+      "step": 6307
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8646879080155923e-06,
+      "loss": 0.2375,
+      "step": 6308
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8622746252696522e-06,
+      "loss": 0.3547,
+      "step": 6309
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8598627448223804e-06,
+      "loss": 0.2892,
+      "step": 6310
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8574522670893935e-06,
+      "loss": 0.217,
+      "step": 6311
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.85504319248607e-06,
+      "loss": 0.2466,
+      "step": 6312
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8526355214275382e-06,
+      "loss": 0.305,
+      "step": 6313
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8502292543286938e-06,
+      "loss": 0.2557,
+      "step": 6314
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8478243916041882e-06,
+      "loss": 0.2335,
+      "step": 6315
+    },
+    {
+      "epoch": 7.28,
+      "grad_norm": 0.0,
+      "learning_rate": 1.845420933668427e-06,
+      "loss": 0.2789,
+      "step": 6316
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.843018880935572e-06,
+      "loss": 0.2462,
+      "step": 6317
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8406182338195522e-06,
+      "loss": 0.2417,
+      "step": 6318
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8382189927340454e-06,
+      "loss": 0.2724,
+      "step": 6319
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.83582115809249e-06,
+      "loss": 0.3226,
+      "step": 6320
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8334247303080888e-06,
+      "loss": 0.2564,
+      "step": 6321
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8310297097937868e-06,
+      "loss": 0.2329,
+      "step": 6322
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.828636096962304e-06,
+      "loss": 0.2373,
+      "step": 6323
+    },
+    {
+      "epoch": 7.29,
+      "grad_norm": 0.0,
+      "learning_rate": 1.826243892226106e-06,
+      "loss": 0.2645,
+      "step": 6324
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8238530959974143e-06,
+      "loss": 0.2407,
+      "step": 6325
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8214637086882214e-06,
+      "loss": 0.3061,
+      "step": 6326
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8190757307102646e-06,
+      "loss": 0.2806,
+      "step": 6327
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8166891624750393e-06,
+      "loss": 0.409,
+      "step": 6328
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8143040043938054e-06,
+      "loss": 0.2811,
+      "step": 6329
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8119202568775706e-06,
+      "loss": 0.234,
+      "step": 6330
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8095379203371044e-06,
+      "loss": 0.3393,
+      "step": 6331
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8071569951829381e-06,
+      "loss": 0.2845,
+      "step": 6332
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8047774818253483e-06,
+      "loss": 0.267,
+      "step": 6333
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8023993806743777e-06,
+      "loss": 0.2438,
+      "step": 6334
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8000226921398234e-06,
+      "loss": 0.244,
+      "step": 6335
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7976474166312318e-06,
+      "loss": 0.3129,
+      "step": 6336
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7952735545579181e-06,
+      "loss": 0.2356,
+      "step": 6337
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.792901106328948e-06,
+      "loss": 0.2505,
+      "step": 6338
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7905300723531393e-06,
+      "loss": 0.2745,
+      "step": 6339
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.788160453039075e-06,
+      "loss": 0.236,
+      "step": 6340
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7857922487950873e-06,
+      "loss": 0.2708,
+      "step": 6341
+    },
+    {
+      "epoch": 7.31,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7834254600292643e-06,
+      "loss": 0.2449,
+      "step": 6342
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7810600871494555e-06,
+      "loss": 0.2141,
+      "step": 6343
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7786961305632633e-06,
+      "loss": 0.2505,
+      "step": 6344
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7763335906780487e-06,
+      "loss": 0.2947,
+      "step": 6345
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.773972467900924e-06,
+      "loss": 0.2487,
+      "step": 6346
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.771612762638758e-06,
+      "loss": 0.2813,
+      "step": 6347
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.76925447529818e-06,
+      "loss": 0.2618,
+      "step": 6348
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.766897606285568e-06,
+      "loss": 0.2807,
+      "step": 6349
+    },
+    {
+      "epoch": 7.32,
+      "grad_norm": 0.0,
+      "learning_rate": 1.764542156007062e-06,
+      "loss": 0.275,
+      "step": 6350
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7621881248685569e-06,
+      "loss": 0.2483,
+      "step": 6351
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7598355132756983e-06,
+      "loss": 0.2505,
+      "step": 6352
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7574843216338878e-06,
+      "loss": 0.2838,
+      "step": 6353
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7551345503482897e-06,
+      "loss": 0.236,
+      "step": 6354
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7527861998238094e-06,
+      "loss": 0.2463,
+      "step": 6355
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7504392704651286e-06,
+      "loss": 0.2926,
+      "step": 6356
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7480937626766648e-06,
+      "loss": 0.2428,
+      "step": 6357
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.745749676862596e-06,
+      "loss": 0.2595,
+      "step": 6358
+    },
+    {
+      "epoch": 7.33,
+      "grad_norm": 0.0,
+      "learning_rate": 1.743407013426861e-06,
+      "loss": 0.3125,
+      "step": 6359
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7410657727731461e-06,
+      "loss": 0.3185,
+      "step": 6360
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7387259553048963e-06,
+      "loss": 0.2691,
+      "step": 6361
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7363875614253133e-06,
+      "loss": 0.2667,
+      "step": 6362
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7340505915373495e-06,
+      "loss": 0.232,
+      "step": 6363
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7317150460437093e-06,
+      "loss": 0.2403,
+      "step": 6364
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7293809253468607e-06,
+      "loss": 0.3168,
+      "step": 6365
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7270482298490176e-06,
+      "loss": 0.2977,
+      "step": 6366
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7247169599521529e-06,
+      "loss": 0.2509,
+      "step": 6367
+    },
+    {
+      "epoch": 7.34,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7223871160579952e-06,
+      "loss": 0.3372,
+      "step": 6368
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.72005869856802e-06,
+      "loss": 0.2386,
+      "step": 6369
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7177317078834677e-06,
+      "loss": 0.3412,
+      "step": 6370
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7154061444053239e-06,
+      "loss": 0.2487,
+      "step": 6371
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.713082008534328e-06,
+      "loss": 0.2898,
+      "step": 6372
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7107593006709799e-06,
+      "loss": 0.2528,
+      "step": 6373
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7084380212155339e-06,
+      "loss": 0.2307,
+      "step": 6374
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7061181705679874e-06,
+      "loss": 0.2957,
+      "step": 6375
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7037997491281056e-06,
+      "loss": 0.2696,
+      "step": 6376
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7014827572953974e-06,
+      "loss": 0.2408,
+      "step": 6377
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6991671954691257e-06,
+      "loss": 0.2225,
+      "step": 6378
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6968530640483126e-06,
+      "loss": 0.2525,
+      "step": 6379
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6945403634317314e-06,
+      "loss": 0.265,
+      "step": 6380
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6922290940179098e-06,
+      "loss": 0.2561,
+      "step": 6381
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6899192562051248e-06,
+      "loss": 0.3203,
+      "step": 6382
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.687610850391408e-06,
+      "loss": 0.244,
+      "step": 6383
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6853038769745466e-06,
+      "loss": 0.2503,
+      "step": 6384
+    },
+    {
+      "epoch": 7.36,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6829983363520842e-06,
+      "loss": 0.2708,
+      "step": 6385
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6806942289213068e-06,
+      "loss": 0.2209,
+      "step": 6386
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6783915550792652e-06,
+      "loss": 0.2961,
+      "step": 6387
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6760903152227547e-06,
+      "loss": 0.2484,
+      "step": 6388
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6737905097483254e-06,
+      "loss": 0.3063,
+      "step": 6389
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6714921390522844e-06,
+      "loss": 0.262,
+      "step": 6390
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6691952035306857e-06,
+      "loss": 0.3965,
+      "step": 6391
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6668997035793432e-06,
+      "loss": 0.2472,
+      "step": 6392
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.664605639593818e-06,
+      "loss": 0.2862,
+      "step": 6393
+    },
+    {
+      "epoch": 7.37,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6623130119694197e-06,
+      "loss": 0.2514,
+      "step": 6394
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.660021821101222e-06,
+      "loss": 0.2678,
+      "step": 6395
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6577320673840392e-06,
+      "loss": 0.2419,
+      "step": 6396
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6554437512124465e-06,
+      "loss": 0.2952,
+      "step": 6397
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.653156872980769e-06,
+      "loss": 0.2442,
+      "step": 6398
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6508714330830823e-06,
+      "loss": 0.2401,
+      "step": 6399
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6485874319132123e-06,
+      "loss": 0.2722,
+      "step": 6400
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6463048698647444e-06,
+      "loss": 0.2245,
+      "step": 6401
+    },
+    {
+      "epoch": 7.38,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6440237473310062e-06,
+      "loss": 0.2635,
+      "step": 6402
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6417440647050853e-06,
+      "loss": 0.3026,
+      "step": 6403
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6394658223798198e-06,
+      "loss": 0.2569,
+      "step": 6404
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6371890207477937e-06,
+      "loss": 0.3115,
+      "step": 6405
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6349136602013527e-06,
+      "loss": 0.258,
+      "step": 6406
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6326397411325845e-06,
+      "loss": 0.2257,
+      "step": 6407
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6303672639333323e-06,
+      "loss": 0.2927,
+      "step": 6408
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6280962289951907e-06,
+      "loss": 0.254,
+      "step": 6409
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.625826636709511e-06,
+      "loss": 0.2628,
+      "step": 6410
+    },
+    {
+      "epoch": 7.39,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6235584874673848e-06,
+      "loss": 0.2852,
+      "step": 6411
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6212917816596663e-06,
+      "loss": 0.2507,
+      "step": 6412
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6190265196769539e-06,
+      "loss": 0.2967,
+      "step": 6413
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6167627019095934e-06,
+      "loss": 0.3079,
+      "step": 6414
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6145003287476968e-06,
+      "loss": 0.2597,
+      "step": 6415
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6122394005811127e-06,
+      "loss": 0.3285,
+      "step": 6416
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6099799177994491e-06,
+      "loss": 0.28,
+      "step": 6417
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.607721880792059e-06,
+      "loss": 0.2512,
+      "step": 6418
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6054652899480472e-06,
+      "loss": 0.2438,
+      "step": 6419
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6032101456562744e-06,
+      "loss": 0.2494,
+      "step": 6420
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6009564483053496e-06,
+      "loss": 0.242,
+      "step": 6421
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.598704198283627e-06,
+      "loss": 0.2523,
+      "step": 6422
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5964533959792217e-06,
+      "loss": 0.2595,
+      "step": 6423
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5942040417799897e-06,
+      "loss": 0.3254,
+      "step": 6424
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5919561360735403e-06,
+      "loss": 0.238,
+      "step": 6425
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5897096792472354e-06,
+      "loss": 0.3069,
+      "step": 6426
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.244,
+      "step": 6427
+    },
+    {
+      "epoch": 7.41,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5852211137832585e-06,
+      "loss": 0.2635,
+      "step": 6428
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5829790059190586e-06,
+      "loss": 0.2749,
+      "step": 6429
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5807383484819483e-06,
+      "loss": 0.2685,
+      "step": 6430
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.578499141858041e-06,
+      "loss": 0.2416,
+      "step": 6431
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5762613864331965e-06,
+      "loss": 0.2743,
+      "step": 6432
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.574025082593027e-06,
+      "loss": 0.2395,
+      "step": 6433
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5717902307228972e-06,
+      "loss": 0.216,
+      "step": 6434
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5695568312079156e-06,
+      "loss": 0.2501,
+      "step": 6435
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5673248844329402e-06,
+      "loss": 0.289,
+      "step": 6436
+    },
+    {
+      "epoch": 7.42,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5650943907825888e-06,
+      "loss": 0.2589,
+      "step": 6437
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5628653506412139e-06,
+      "loss": 0.2822,
+      "step": 6438
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5606377643929305e-06,
+      "loss": 0.2391,
+      "step": 6439
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5584116324215982e-06,
+      "loss": 0.2744,
+      "step": 6440
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5561869551108212e-06,
+      "loss": 0.2554,
+      "step": 6441
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5539637328439627e-06,
+      "loss": 0.2342,
+      "step": 6442
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5517419660041277e-06,
+      "loss": 0.2268,
+      "step": 6443
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5495216549741677e-06,
+      "loss": 0.2405,
+      "step": 6444
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5473028001366973e-06,
+      "loss": 0.2504,
+      "step": 6445
+    },
+    {
+      "epoch": 7.43,
+      "grad_norm": 0.0,
+      "learning_rate": 1.545085401874068e-06,
+      "loss": 0.2391,
+      "step": 6446
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.542869460568379e-06,
+      "loss": 0.2724,
+      "step": 6447
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.540654976601489e-06,
+      "loss": 0.2797,
+      "step": 6448
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5384419503549941e-06,
+      "loss": 0.2464,
+      "step": 6449
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5362303822102464e-06,
+      "loss": 0.2601,
+      "step": 6450
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.534020272548349e-06,
+      "loss": 0.2689,
+      "step": 6451
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5318116217501445e-06,
+      "loss": 0.2665,
+      "step": 6452
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.529604430196232e-06,
+      "loss": 0.2371,
+      "step": 6453
+    },
+    {
+      "epoch": 7.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5273986982669553e-06,
+      "loss": 0.3032,
+      "step": 6454
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5251944263424056e-06,
+      "loss": 0.2198,
+      "step": 6455
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.522991614802427e-06,
+      "loss": 0.2422,
+      "step": 6456
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5207902640266114e-06,
+      "loss": 0.2617,
+      "step": 6457
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5185903743942932e-06,
+      "loss": 0.2974,
+      "step": 6458
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5163919462845622e-06,
+      "loss": 0.2402,
+      "step": 6459
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5141949800762523e-06,
+      "loss": 0.3342,
+      "step": 6460
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5119994761479429e-06,
+      "loss": 0.266,
+      "step": 6461
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5098054348779678e-06,
+      "loss": 0.2491,
+      "step": 6462
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.507612856644405e-06,
+      "loss": 0.3408,
+      "step": 6463
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5054217418250848e-06,
+      "loss": 0.2534,
+      "step": 6464
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5032320907975772e-06,
+      "loss": 0.2816,
+      "step": 6465
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5010439039392043e-06,
+      "loss": 0.3472,
+      "step": 6466
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4988571816270402e-06,
+      "loss": 0.24,
+      "step": 6467
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.496671924237898e-06,
+      "loss": 0.2662,
+      "step": 6468
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4944881321483428e-06,
+      "loss": 0.2764,
+      "step": 6469
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.492305805734693e-06,
+      "loss": 0.2779,
+      "step": 6470
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4901249453730038e-06,
+      "loss": 0.282,
+      "step": 6471
+    },
+    {
+      "epoch": 7.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4879455514390817e-06,
+      "loss": 0.3083,
+      "step": 6472
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.485767624308485e-06,
+      "loss": 0.2616,
+      "step": 6473
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4835911643565094e-06,
+      "loss": 0.2571,
+      "step": 6474
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4814161719582132e-06,
+      "loss": 0.2818,
+      "step": 6475
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4792426474883869e-06,
+      "loss": 0.4101,
+      "step": 6476
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4770705913215743e-06,
+      "loss": 0.2353,
+      "step": 6477
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4749000038320671e-06,
+      "loss": 0.2475,
+      "step": 6478
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4727308853939004e-06,
+      "loss": 0.2403,
+      "step": 6479
+    },
+    {
+      "epoch": 7.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4705632363808586e-06,
+      "loss": 0.2397,
+      "step": 6480
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4683970571664763e-06,
+      "loss": 0.3382,
+      "step": 6481
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4662323481240281e-06,
+      "loss": 0.2549,
+      "step": 6482
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4640691096265358e-06,
+      "loss": 0.2897,
+      "step": 6483
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4619073420467756e-06,
+      "loss": 0.2791,
+      "step": 6484
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4597470457572583e-06,
+      "loss": 0.2926,
+      "step": 6485
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4575882211302527e-06,
+      "loss": 0.2864,
+      "step": 6486
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4554308685377694e-06,
+      "loss": 0.2503,
+      "step": 6487
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4532749883515606e-06,
+      "loss": 0.3363,
+      "step": 6488
+    },
+    {
+      "epoch": 7.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.451120580943134e-06,
+      "loss": 0.2361,
+      "step": 6489
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4489676466837366e-06,
+      "loss": 0.2579,
+      "step": 6490
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4468161859443609e-06,
+      "loss": 0.2532,
+      "step": 6491
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.44466619909575e-06,
+      "loss": 0.3106,
+      "step": 6492
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4425176865083946e-06,
+      "loss": 0.2515,
+      "step": 6493
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4403706485525226e-06,
+      "loss": 0.2961,
+      "step": 6494
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4382250855981172e-06,
+      "loss": 0.2498,
+      "step": 6495
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4360809980149027e-06,
+      "loss": 0.2624,
+      "step": 6496
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4339383861723465e-06,
+      "loss": 0.2276,
+      "step": 6497
+    },
+    {
+      "epoch": 7.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.431797250439667e-06,
+      "loss": 0.2141,
+      "step": 6498
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4296575911858268e-06,
+      "loss": 0.2344,
+      "step": 6499
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4275194087795352e-06,
+      "loss": 0.2367,
+      "step": 6500
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4253827035892442e-06,
+      "loss": 0.2782,
+      "step": 6501
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4232474759831483e-06,
+      "loss": 0.2749,
+      "step": 6502
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4211137263291952e-06,
+      "loss": 0.302,
+      "step": 6503
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4189814549950765e-06,
+      "loss": 0.2867,
+      "step": 6504
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4168506623482202e-06,
+      "loss": 0.2904,
+      "step": 6505
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4147213487558131e-06,
+      "loss": 0.2227,
+      "step": 6506
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.412593514584777e-06,
+      "loss": 0.2828,
+      "step": 6507
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4104671602017795e-06,
+      "loss": 0.2327,
+      "step": 6508
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4083422859732365e-06,
+      "loss": 0.2603,
+      "step": 6509
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4062188922653086e-06,
+      "loss": 0.3697,
+      "step": 6510
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4040969794439041e-06,
+      "loss": 0.2204,
+      "step": 6511
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4019765478746683e-06,
+      "loss": 0.2472,
+      "step": 6512
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3998575979229944e-06,
+      "loss": 0.3166,
+      "step": 6513
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.397740129954026e-06,
+      "loss": 0.2547,
+      "step": 6514
+    },
+    {
+      "epoch": 7.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3956241443326423e-06,
+      "loss": 0.3025,
+      "step": 6515
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3935096414234729e-06,
+      "loss": 0.2658,
+      "step": 6516
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3913966215908926e-06,
+      "loss": 0.28,
+      "step": 6517
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.389285085199017e-06,
+      "loss": 0.2304,
+      "step": 6518
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.387175032611705e-06,
+      "loss": 0.2681,
+      "step": 6519
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.385066464192567e-06,
+      "loss": 0.2119,
+      "step": 6520
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.382959380304948e-06,
+      "loss": 0.2622,
+      "step": 6521
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.380853781311945e-06,
+      "loss": 0.2281,
+      "step": 6522
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.378749667576399e-06,
+      "loss": 0.2692,
+      "step": 6523
+    },
+    {
+      "epoch": 7.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3766470394608867e-06,
+      "loss": 0.3046,
+      "step": 6524
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3745458973277392e-06,
+      "loss": 0.3048,
+      "step": 6525
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.372446241539026e-06,
+      "loss": 0.3723,
+      "step": 6526
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3703480724565577e-06,
+      "loss": 0.2198,
+      "step": 6527
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3682513904418947e-06,
+      "loss": 0.2279,
+      "step": 6528
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3661561958563407e-06,
+      "loss": 0.2763,
+      "step": 6529
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3640624890609376e-06,
+      "loss": 0.2753,
+      "step": 6530
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3619702704164783e-06,
+      "loss": 0.2331,
+      "step": 6531
+    },
+    {
+      "epoch": 7.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.359879540283492e-06,
+      "loss": 0.3003,
+      "step": 6532
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3577902990222547e-06,
+      "loss": 0.2728,
+      "step": 6533
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.355702546992791e-06,
+      "loss": 0.2733,
+      "step": 6534
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3536162845548572e-06,
+      "loss": 0.3443,
+      "step": 6535
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3515315120679651e-06,
+      "loss": 0.3139,
+      "step": 6536
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3494482298913614e-06,
+      "loss": 0.2898,
+      "step": 6537
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3473664383840367e-06,
+      "loss": 0.2963,
+      "step": 6538
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3452861379047289e-06,
+      "loss": 0.2629,
+      "step": 6539
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3432073288119196e-06,
+      "loss": 0.3028,
+      "step": 6540
+    },
+    {
+      "epoch": 7.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.341130011463826e-06,
+      "loss": 0.2811,
+      "step": 6541
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.339054186218418e-06,
+      "loss": 0.3102,
+      "step": 6542
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.336979853433399e-06,
+      "loss": 0.2333,
+      "step": 6543
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3349070134662201e-06,
+      "loss": 0.2613,
+      "step": 6544
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3328356666740739e-06,
+      "loss": 0.2732,
+      "step": 6545
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.330765813413899e-06,
+      "loss": 0.2718,
+      "step": 6546
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3286974540423747e-06,
+      "loss": 0.2223,
+      "step": 6547
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3266305889159204e-06,
+      "loss": 0.2291,
+      "step": 6548
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3245652183906965e-06,
+      "loss": 0.223,
+      "step": 6549
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3225013428226151e-06,
+      "loss": 0.2899,
+      "step": 6550
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3204389625673208e-06,
+      "loss": 0.308,
+      "step": 6551
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3183780779802047e-06,
+      "loss": 0.2362,
+      "step": 6552
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3163186894164026e-06,
+      "loss": 0.4049,
+      "step": 6553
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3142607972307886e-06,
+      "loss": 0.2747,
+      "step": 6554
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3122044017779768e-06,
+      "loss": 0.237,
+      "step": 6555
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3101495034123313e-06,
+      "loss": 0.2648,
+      "step": 6556
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3080961024879501e-06,
+      "loss": 0.2671,
+      "step": 6557
+    },
+    {
+      "epoch": 7.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3060441993586793e-06,
+      "loss": 0.2883,
+      "step": 6558
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3039937943781046e-06,
+      "loss": 0.302,
+      "step": 6559
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.30194488789955e-06,
+      "loss": 0.2516,
+      "step": 6560
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.29989748027609e-06,
+      "loss": 0.2394,
+      "step": 6561
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2978515718605311e-06,
+      "loss": 0.2516,
+      "step": 6562
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2958071630054214e-06,
+      "loss": 0.2728,
+      "step": 6563
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2937642540630658e-06,
+      "loss": 0.2318,
+      "step": 6564
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2917228453854936e-06,
+      "loss": 0.2665,
+      "step": 6565
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.28968293732448e-06,
+      "loss": 0.2411,
+      "step": 6566
+    },
+    {
+      "epoch": 7.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2876445302315489e-06,
+      "loss": 0.2668,
+      "step": 6567
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.285607624457954e-06,
+      "loss": 0.2458,
+      "step": 6568
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2835722203546996e-06,
+      "loss": 0.2642,
+      "step": 6569
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2815383182725295e-06,
+      "loss": 0.2856,
+      "step": 6570
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.279505918561923e-06,
+      "loss": 0.2599,
+      "step": 6571
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2774750215731103e-06,
+      "loss": 0.2444,
+      "step": 6572
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2754456276560534e-06,
+      "loss": 0.2827,
+      "step": 6573
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2734177371604562e-06,
+      "loss": 0.2439,
+      "step": 6574
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.27139135043577e-06,
+      "loss": 0.2572,
+      "step": 6575
+    },
+    {
+      "epoch": 7.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2693664678311846e-06,
+      "loss": 0.2757,
+      "step": 6576
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2673430896956252e-06,
+      "loss": 0.3018,
+      "step": 6577
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2653212163777662e-06,
+      "loss": 0.2774,
+      "step": 6578
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2633008482260146e-06,
+      "loss": 0.279,
+      "step": 6579
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2612819855885218e-06,
+      "loss": 0.2982,
+      "step": 6580
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2592646288131804e-06,
+      "loss": 0.2188,
+      "step": 6581
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2572487782476228e-06,
+      "loss": 0.2862,
+      "step": 6582
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2552344342392242e-06,
+      "loss": 0.3095,
+      "step": 6583
+    },
+    {
+      "epoch": 7.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.253221597135097e-06,
+      "loss": 0.2528,
+      "step": 6584
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2512102672820902e-06,
+      "loss": 0.3087,
+      "step": 6585
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2492004450268025e-06,
+      "loss": 0.2813,
+      "step": 6586
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2471921307155655e-06,
+      "loss": 0.2951,
+      "step": 6587
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2451853246944534e-06,
+      "loss": 0.3032,
+      "step": 6588
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2431800273092832e-06,
+      "loss": 0.2422,
+      "step": 6589
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.241176238905607e-06,
+      "loss": 0.2462,
+      "step": 6590
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2391739598287167e-06,
+      "loss": 0.2732,
+      "step": 6591
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2371731904236507e-06,
+      "loss": 0.3063,
+      "step": 6592
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2351739310351796e-06,
+      "loss": 0.2852,
+      "step": 6593
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2331761820078214e-06,
+      "loss": 0.2459,
+      "step": 6594
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2311799436858275e-06,
+      "loss": 0.2399,
+      "step": 6595
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.229185216413189e-06,
+      "loss": 0.26,
+      "step": 6596
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2271920005336425e-06,
+      "loss": 0.2548,
+      "step": 6597
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2252002963906551e-06,
+      "loss": 0.225,
+      "step": 6598
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2232101043274437e-06,
+      "loss": 0.2292,
+      "step": 6599
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.221221424686958e-06,
+      "loss": 0.2806,
+      "step": 6600
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.21923425781189e-06,
+      "loss": 0.2731,
+      "step": 6601
+    },
+    {
+      "epoch": 7.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2172486040446663e-06,
+      "loss": 0.2718,
+      "step": 6602
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2152644637274603e-06,
+      "loss": 0.3664,
+      "step": 6603
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.213281837202176e-06,
+      "loss": 0.2476,
+      "step": 6604
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2113007248104625e-06,
+      "loss": 0.267,
+      "step": 6605
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.209321126893711e-06,
+      "loss": 0.2812,
+      "step": 6606
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2073430437930412e-06,
+      "loss": 0.3068,
+      "step": 6607
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2053664758493212e-06,
+      "loss": 0.2666,
+      "step": 6608
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2033914234031552e-06,
+      "loss": 0.3089,
+      "step": 6609
+    },
+    {
+      "epoch": 7.62,
+      "grad_norm": 0.0,
+      "learning_rate": 1.201417886794881e-06,
+      "loss": 0.2543,
+      "step": 6610
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1994458663645836e-06,
+      "loss": 0.2885,
+      "step": 6611
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.197475362452084e-06,
+      "loss": 0.2611,
+      "step": 6612
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1955063753969365e-06,
+      "loss": 0.2368,
+      "step": 6613
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1935389055384439e-06,
+      "loss": 0.2841,
+      "step": 6614
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1915729532156372e-06,
+      "loss": 0.2428,
+      "step": 6615
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1896085187672945e-06,
+      "loss": 0.23,
+      "step": 6616
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.187645602531925e-06,
+      "loss": 0.2342,
+      "step": 6617
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1856842048477813e-06,
+      "loss": 0.2514,
+      "step": 6618
+    },
+    {
+      "epoch": 7.63,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1837243260528542e-06,
+      "loss": 0.2187,
+      "step": 6619
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1817659664848714e-06,
+      "loss": 0.2795,
+      "step": 6620
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.179809126481295e-06,
+      "loss": 0.2716,
+      "step": 6621
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1778538063793322e-06,
+      "loss": 0.2536,
+      "step": 6622
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1759000065159276e-06,
+      "loss": 0.3192,
+      "step": 6623
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1739477272277554e-06,
+      "loss": 0.3135,
+      "step": 6624
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1719969688512389e-06,
+      "loss": 0.2876,
+      "step": 6625
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1700477317225333e-06,
+      "loss": 0.2833,
+      "step": 6626
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.168100016177528e-06,
+      "loss": 0.2421,
+      "step": 6627
+    },
+    {
+      "epoch": 7.64,
+      "grad_norm": 0.0,
+      "learning_rate": 1.16615382255186e-06,
+      "loss": 0.2716,
+      "step": 6628
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.164209151180895e-06,
+      "loss": 0.2879,
+      "step": 6629
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1622660023997455e-06,
+      "loss": 0.2699,
+      "step": 6630
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1603243765432527e-06,
+      "loss": 0.259,
+      "step": 6631
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1583842739459971e-06,
+      "loss": 0.2381,
+      "step": 6632
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.156445694942302e-06,
+      "loss": 0.24,
+      "step": 6633
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.154508639866222e-06,
+      "loss": 0.2589,
+      "step": 6634
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1525731090515536e-06,
+      "loss": 0.2696,
+      "step": 6635
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1506391028318298e-06,
+      "loss": 0.3457,
+      "step": 6636
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1487066215403186e-06,
+      "loss": 0.2456,
+      "step": 6637
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.146775665510025e-06,
+      "loss": 0.3241,
+      "step": 6638
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1448462350736956e-06,
+      "loss": 0.2554,
+      "step": 6639
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.142918330563808e-06,
+      "loss": 0.2788,
+      "step": 6640
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1409919523125823e-06,
+      "loss": 0.3075,
+      "step": 6641
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.139067100651976e-06,
+      "loss": 0.2966,
+      "step": 6642
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.137143775913675e-06,
+      "loss": 0.2759,
+      "step": 6643
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1352219784291129e-06,
+      "loss": 0.2834,
+      "step": 6644
+    },
+    {
+      "epoch": 7.66,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1333017085294528e-06,
+      "loss": 0.3358,
+      "step": 6645
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1313829665455967e-06,
+      "loss": 0.3106,
+      "step": 6646
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1294657528081842e-06,
+      "loss": 0.231,
+      "step": 6647
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1275500676475925e-06,
+      "loss": 0.2594,
+      "step": 6648
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1256359113939309e-06,
+      "loss": 0.2796,
+      "step": 6649
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1237232843770517e-06,
+      "loss": 0.3049,
+      "step": 6650
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1218121869265365e-06,
+      "loss": 0.2532,
+      "step": 6651
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.119902619371709e-06,
+      "loss": 0.2769,
+      "step": 6652
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1179945820416282e-06,
+      "loss": 0.2421,
+      "step": 6653
+    },
+    {
+      "epoch": 7.67,
+      "grad_norm": 0.0,
+      "learning_rate": 1.116088075265085e-06,
+      "loss": 0.2817,
+      "step": 6654
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1141830993706149e-06,
+      "loss": 0.2388,
+      "step": 6655
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1122796546864823e-06,
+      "loss": 0.3053,
+      "step": 6656
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.110377741540687e-06,
+      "loss": 0.2114,
+      "step": 6657
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1084773602609712e-06,
+      "loss": 0.2783,
+      "step": 6658
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1065785111748117e-06,
+      "loss": 0.2587,
+      "step": 6659
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.104681194609416e-06,
+      "loss": 0.2702,
+      "step": 6660
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1027854108917335e-06,
+      "loss": 0.2429,
+      "step": 6661
+    },
+    {
+      "epoch": 7.68,
+      "grad_norm": 0.0,
+      "learning_rate": 1.100891160348445e-06,
+      "loss": 0.251,
+      "step": 6662
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.098998443305972e-06,
+      "loss": 0.2754,
+      "step": 6663
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0971072600904654e-06,
+      "loss": 0.2753,
+      "step": 6664
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0952176110278167e-06,
+      "loss": 0.2265,
+      "step": 6665
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0933294964436537e-06,
+      "loss": 0.2707,
+      "step": 6666
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0914429166633355e-06,
+      "loss": 0.2728,
+      "step": 6667
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0895578720119571e-06,
+      "loss": 0.2289,
+      "step": 6668
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.087674362814355e-06,
+      "loss": 0.2597,
+      "step": 6669
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0857923893950929e-06,
+      "loss": 0.2245,
+      "step": 6670
+    },
+    {
+      "epoch": 7.69,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0839119520784758e-06,
+      "loss": 0.3287,
+      "step": 6671
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0820330511885425e-06,
+      "loss": 0.2333,
+      "step": 6672
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.080155687049067e-06,
+      "loss": 0.2329,
+      "step": 6673
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0782798599835541e-06,
+      "loss": 0.3032,
+      "step": 6674
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.076405570315252e-06,
+      "loss": 0.2589,
+      "step": 6675
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0745328183671366e-06,
+      "loss": 0.2257,
+      "step": 6676
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0726616044619243e-06,
+      "loss": 0.2755,
+      "step": 6677
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0707919289220637e-06,
+      "loss": 0.2582,
+      "step": 6678
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.068923792069736e-06,
+      "loss": 0.285,
+      "step": 6679
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0670571942268637e-06,
+      "loss": 0.3027,
+      "step": 6680
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0651921357150997e-06,
+      "loss": 0.2619,
+      "step": 6681
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0633286168558255e-06,
+      "loss": 0.2366,
+      "step": 6682
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0614666379701732e-06,
+      "loss": 0.3285,
+      "step": 6683
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0596061993789974e-06,
+      "loss": 0.2188,
+      "step": 6684
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0577473014028872e-06,
+      "loss": 0.2658,
+      "step": 6685
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.055889944362174e-06,
+      "loss": 0.3293,
+      "step": 6686
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.054034128576913e-06,
+      "loss": 0.2553,
+      "step": 6687
+    },
+    {
+      "epoch": 7.71,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0521798543669038e-06,
+      "loss": 0.2457,
+      "step": 6688
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0503271220516775e-06,
+      "loss": 0.2689,
+      "step": 6689
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0484759319504945e-06,
+      "loss": 0.2688,
+      "step": 6690
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.046626284382356e-06,
+      "loss": 0.2752,
+      "step": 6691
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.044778179665994e-06,
+      "loss": 0.2586,
+      "step": 6692
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0429316181198735e-06,
+      "loss": 0.3088,
+      "step": 6693
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0410866000621955e-06,
+      "loss": 0.2969,
+      "step": 6694
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.039243125810898e-06,
+      "loss": 0.2779,
+      "step": 6695
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.037401195683646e-06,
+      "loss": 0.2903,
+      "step": 6696
+    },
+    {
+      "epoch": 7.72,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0355608099978454e-06,
+      "loss": 0.3427,
+      "step": 6697
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0337219690706302e-06,
+      "loss": 0.2532,
+      "step": 6698
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0318846732188737e-06,
+      "loss": 0.3143,
+      "step": 6699
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0300489227591758e-06,
+      "loss": 0.2308,
+      "step": 6700
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0282147180078761e-06,
+      "loss": 0.2258,
+      "step": 6701
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0263820592810503e-06,
+      "loss": 0.2788,
+      "step": 6702
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0245509468944992e-06,
+      "loss": 0.2928,
+      "step": 6703
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0227213811637593e-06,
+      "loss": 0.2983,
+      "step": 6704
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0208933624041085e-06,
+      "loss": 0.2444,
+      "step": 6705
+    },
+    {
+      "epoch": 7.73,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0190668909305469e-06,
+      "loss": 0.2615,
+      "step": 6706
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.017241967057816e-06,
+      "loss": 0.2872,
+      "step": 6707
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.01541859110039e-06,
+      "loss": 0.2594,
+      "step": 6708
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0135967633724708e-06,
+      "loss": 0.2859,
+      "step": 6709
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0117764841879973e-06,
+      "loss": 0.2306,
+      "step": 6710
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.009957753860642e-06,
+      "loss": 0.2696,
+      "step": 6711
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0081405727038106e-06,
+      "loss": 0.2524,
+      "step": 6712
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.006324941030643e-06,
+      "loss": 0.2485,
+      "step": 6713
+    },
+    {
+      "epoch": 7.74,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0045108591540075e-06,
+      "loss": 0.2674,
+      "step": 6714
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0026983273865055e-06,
+      "loss": 0.2769,
+      "step": 6715
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 0.0,
+      "learning_rate": 1.000887346040479e-06,
+      "loss": 0.2966,
+      "step": 6716
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.99077915427994e-07,
+      "loss": 0.2691,
+      "step": 6717
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.972700358608534e-07,
+      "loss": 0.2425,
+      "step": 6718
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.954637076505946e-07,
+      "loss": 0.2243,
+      "step": 6719
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.936589311084843e-07,
+      "loss": 0.2246,
+      "step": 6720
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.918557065455193e-07,
+      "loss": 0.2522,
+      "step": 6721
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.900540342724386e-07,
+      "loss": 0.2572,
+      "step": 6722
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 0.0,
+      "learning_rate": 9.882539145997027e-07,
+      "loss": 0.2604,
+      "step": 6723
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.864553478375105e-07,
+      "loss": 0.3069,
+      "step": 6724
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.84658334295796e-07,
+      "loss": 0.2857,
+      "step": 6725
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.828628742842161e-07,
+      "loss": 0.2692,
+      "step": 6726
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.81068968112172e-07,
+      "loss": 0.2564,
+      "step": 6727
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.792766160887868e-07,
+      "loss": 0.2365,
+      "step": 6728
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.774858185229198e-07,
+      "loss": 0.2588,
+      "step": 6729
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.756965757231618e-07,
+      "loss": 0.2276,
+      "step": 6730
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.739088879978409e-07,
+      "loss": 0.2647,
+      "step": 6731
+    },
+    {
+      "epoch": 7.76,
+      "grad_norm": 0.0,
+      "learning_rate": 9.721227556550083e-07,
+      "loss": 0.2538,
+      "step": 6732
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.703381790024547e-07,
+      "loss": 0.36,
+      "step": 6733
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.68555158347696e-07,
+      "loss": 0.2798,
+      "step": 6734
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.667736939979888e-07,
+      "loss": 0.2407,
+      "step": 6735
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.649937862603097e-07,
+      "loss": 0.2624,
+      "step": 6736
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.63215435441378e-07,
+      "loss": 0.2138,
+      "step": 6737
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.614386418476419e-07,
+      "loss": 0.2763,
+      "step": 6738
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.59663405785277e-07,
+      "loss": 0.2753,
+      "step": 6739
+    },
+    {
+      "epoch": 7.77,
+      "grad_norm": 0.0,
+      "learning_rate": 9.578897275601917e-07,
+      "loss": 0.2251,
+      "step": 6740
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.561176074780299e-07,
+      "loss": 0.2344,
+      "step": 6741
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.54347045844165e-07,
+      "loss": 0.2513,
+      "step": 6742
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.525780429636999e-07,
+      "loss": 0.2512,
+      "step": 6743
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.508105991414729e-07,
+      "loss": 0.2199,
+      "step": 6744
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.49044714682047e-07,
+      "loss": 0.3179,
+      "step": 6745
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.472803898897253e-07,
+      "loss": 0.2922,
+      "step": 6746
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.455176250685338e-07,
+      "loss": 0.2279,
+      "step": 6747
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.437564205222349e-07,
+      "loss": 0.2595,
+      "step": 6748
+    },
+    {
+      "epoch": 7.78,
+      "grad_norm": 0.0,
+      "learning_rate": 9.419967765543225e-07,
+      "loss": 0.2648,
+      "step": 6749
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.402386934680174e-07,
+      "loss": 0.2855,
+      "step": 6750
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.384821715662729e-07,
+      "loss": 0.3111,
+      "step": 6751
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.367272111517778e-07,
+      "loss": 0.2161,
+      "step": 6752
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.349738125269425e-07,
+      "loss": 0.2408,
+      "step": 6753
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.332219759939187e-07,
+      "loss": 0.304,
+      "step": 6754
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.314717018545838e-07,
+      "loss": 0.2572,
+      "step": 6755
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.297229904105454e-07,
+      "loss": 0.2746,
+      "step": 6756
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.279758419631402e-07,
+      "loss": 0.2598,
+      "step": 6757
+    },
+    {
+      "epoch": 7.79,
+      "grad_norm": 0.0,
+      "learning_rate": 9.262302568134418e-07,
+      "loss": 0.2711,
+      "step": 6758
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.244862352622485e-07,
+      "loss": 0.3308,
+      "step": 6759
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.227437776100911e-07,
+      "loss": 0.3003,
+      "step": 6760
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.210028841572338e-07,
+      "loss": 0.2953,
+      "step": 6761
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.192635552036655e-07,
+      "loss": 0.2576,
+      "step": 6762
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.17525791049112e-07,
+      "loss": 0.2583,
+      "step": 6763
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.157895919930238e-07,
+      "loss": 0.3011,
+      "step": 6764
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.140549583345826e-07,
+      "loss": 0.2801,
+      "step": 6765
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 0.0,
+      "learning_rate": 9.123218903727027e-07,
+      "loss": 0.3147,
+      "step": 6766
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.105903884060307e-07,
+      "loss": 0.25,
+      "step": 6767
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.088604527329359e-07,
+      "loss": 0.294,
+      "step": 6768
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.071320836515263e-07,
+      "loss": 0.2825,
+      "step": 6769
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.054052814596304e-07,
+      "loss": 0.2516,
+      "step": 6770
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.2413,
+      "step": 6771
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.019563789343766e-07,
+      "loss": 0.2631,
+      "step": 6772
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 0.0,
+      "learning_rate": 9.002342791953345e-07,
+      "loss": 0.2292,
+      "step": 6773
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 0.0,
+      "learning_rate": 8.985137475344441e-07,
+      "loss": 0.2514,
+      "step": 6774
+    },
+    {
+      "epoch": 7.81,
+      "grad_norm": 0.0,
+      "learning_rate": 8.967947842481894e-07,
+      "loss": 0.2903,
+      "step": 6775
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 0.0,
+      "learning_rate": 8.950773896327802e-07,
+      "loss": 0.2752,
+      "step": 6776
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 0.0,
+      "learning_rate": 8.933615639841608e-07,
+      "loss": 0.2531,
+      "step": 6777
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 0.0,
+      "learning_rate": 8.916473075980058e-07,
+      "loss": 0.2547,
+      "step": 6778
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 0.0,
+      "learning_rate": 8.899346207697135e-07,
+      "loss": 0.2264,
+      "step": 6779
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 0.0,
+      "learning_rate": 8.882235037944187e-07,
+      "loss": 0.2703,
+      "step": 6780
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 0.0,
+      "learning_rate": 8.865139569669778e-07,
+      "loss": 0.3171,
+      "step": 6781
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 0.0,
+      "learning_rate": 8.848059805819875e-07,
+      "loss": 0.3066,
+      "step": 6782
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 0.0,
+      "learning_rate": 8.830995749337612e-07,
+      "loss": 0.3207,
+      "step": 6783
+    },
+    {
+      "epoch": 7.82,
+      "grad_norm": 0.0,
+      "learning_rate": 8.813947403163514e-07,
+      "loss": 0.2531,
+      "step": 6784
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 0.0,
+      "learning_rate": 8.796914770235365e-07,
+      "loss": 0.2991,
+      "step": 6785
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 0.0,
+      "learning_rate": 8.779897853488239e-07,
+      "loss": 0.2803,
+      "step": 6786
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 0.0,
+      "learning_rate": 8.762896655854481e-07,
+      "loss": 0.219,
+      "step": 6787
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 0.0,
+      "learning_rate": 8.745911180263778e-07,
+      "loss": 0.2574,
+      "step": 6788
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 0.0,
+      "learning_rate": 8.728941429643045e-07,
+      "loss": 0.2869,
+      "step": 6789
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 0.0,
+      "learning_rate": 8.711987406916556e-07,
+      "loss": 0.3129,
+      "step": 6790
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 0.0,
+      "learning_rate": 8.695049115005838e-07,
+      "loss": 0.2755,
+      "step": 6791
+    },
+    {
+      "epoch": 7.83,
+      "grad_norm": 0.0,
+      "learning_rate": 8.67812655682968e-07,
+      "loss": 0.2368,
+      "step": 6792
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.0,
+      "learning_rate": 8.661219735304238e-07,
+      "loss": 0.2316,
+      "step": 6793
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.0,
+      "learning_rate": 8.644328653342871e-07,
+      "loss": 0.2578,
+      "step": 6794
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.0,
+      "learning_rate": 8.627453313856249e-07,
+      "loss": 0.2952,
+      "step": 6795
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.0,
+      "learning_rate": 8.610593719752369e-07,
+      "loss": 0.2299,
+      "step": 6796
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.0,
+      "learning_rate": 8.593749873936497e-07,
+      "loss": 0.2358,
+      "step": 6797
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.0,
+      "learning_rate": 8.576921779311143e-07,
+      "loss": 0.2202,
+      "step": 6798
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.0,
+      "learning_rate": 8.560109438776176e-07,
+      "loss": 0.2628,
+      "step": 6799
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.0,
+      "learning_rate": 8.543312855228669e-07,
+      "loss": 0.2432,
+      "step": 6800
+    },
+    {
+      "epoch": 7.84,
+      "grad_norm": 0.0,
+      "learning_rate": 8.526532031563051e-07,
+      "loss": 0.2588,
+      "step": 6801
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.0,
+      "learning_rate": 8.509766970671007e-07,
+      "loss": 0.2671,
+      "step": 6802
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.0,
+      "learning_rate": 8.493017675441495e-07,
+      "loss": 0.2472,
+      "step": 6803
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.0,
+      "learning_rate": 8.476284148760749e-07,
+      "loss": 0.2789,
+      "step": 6804
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.0,
+      "learning_rate": 8.459566393512331e-07,
+      "loss": 0.2744,
+      "step": 6805
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.0,
+      "learning_rate": 8.442864412577034e-07,
+      "loss": 0.2939,
+      "step": 6806
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.0,
+      "learning_rate": 8.426178208832958e-07,
+      "loss": 0.243,
+      "step": 6807
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.0,
+      "learning_rate": 8.409507785155513e-07,
+      "loss": 0.252,
+      "step": 6808
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.0,
+      "learning_rate": 8.392853144417312e-07,
+      "loss": 0.2496,
+      "step": 6809
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 0.0,
+      "learning_rate": 8.376214289488338e-07,
+      "loss": 0.2456,
+      "step": 6810
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 0.0,
+      "learning_rate": 8.359591223235785e-07,
+      "loss": 0.2983,
+      "step": 6811
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 0.0,
+      "learning_rate": 8.342983948524141e-07,
+      "loss": 0.264,
+      "step": 6812
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 0.0,
+      "learning_rate": 8.326392468215206e-07,
+      "loss": 0.2968,
+      "step": 6813
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 0.0,
+      "learning_rate": 8.309816785168035e-07,
+      "loss": 0.2328,
+      "step": 6814
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 0.0,
+      "learning_rate": 8.293256902238933e-07,
+      "loss": 0.2689,
+      "step": 6815
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 0.0,
+      "learning_rate": 8.276712822281563e-07,
+      "loss": 0.2426,
+      "step": 6816
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 0.0,
+      "learning_rate": 8.260184548146755e-07,
+      "loss": 0.2849,
+      "step": 6817
+    },
+    {
+      "epoch": 7.86,
+      "grad_norm": 0.0,
+      "learning_rate": 8.243672082682719e-07,
+      "loss": 0.2871,
+      "step": 6818
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 0.0,
+      "learning_rate": 8.227175428734868e-07,
+      "loss": 0.262,
+      "step": 6819
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 0.0,
+      "learning_rate": 8.210694589145918e-07,
+      "loss": 0.2172,
+      "step": 6820
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 0.0,
+      "learning_rate": 8.194229566755885e-07,
+      "loss": 0.3006,
+      "step": 6821
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 0.0,
+      "learning_rate": 8.17778036440201e-07,
+      "loss": 0.2313,
+      "step": 6822
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 0.0,
+      "learning_rate": 8.161346984918827e-07,
+      "loss": 0.2655,
+      "step": 6823
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 0.0,
+      "learning_rate": 8.144929431138171e-07,
+      "loss": 0.2541,
+      "step": 6824
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 0.0,
+      "learning_rate": 8.128527705889088e-07,
+      "loss": 0.2478,
+      "step": 6825
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 0.0,
+      "learning_rate": 8.112141811997975e-07,
+      "loss": 0.2913,
+      "step": 6826
+    },
+    {
+      "epoch": 7.87,
+      "grad_norm": 0.0,
+      "learning_rate": 8.095771752288451e-07,
+      "loss": 0.2187,
+      "step": 6827
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.0,
+      "learning_rate": 8.079417529581401e-07,
+      "loss": 0.2869,
+      "step": 6828
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.0,
+      "learning_rate": 8.063079146695019e-07,
+      "loss": 0.2871,
+      "step": 6829
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.0,
+      "learning_rate": 8.046756606444717e-07,
+      "loss": 0.2889,
+      "step": 6830
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.0,
+      "learning_rate": 8.030449911643223e-07,
+      "loss": 0.2601,
+      "step": 6831
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.0,
+      "learning_rate": 8.014159065100546e-07,
+      "loss": 0.2417,
+      "step": 6832
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.0,
+      "learning_rate": 7.997884069623907e-07,
+      "loss": 0.3726,
+      "step": 6833
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.0,
+      "learning_rate": 7.981624928017806e-07,
+      "loss": 0.2561,
+      "step": 6834
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.0,
+      "learning_rate": 7.965381643084069e-07,
+      "loss": 0.3209,
+      "step": 6835
+    },
+    {
+      "epoch": 7.88,
+      "grad_norm": 0.0,
+      "learning_rate": 7.949154217621713e-07,
+      "loss": 0.2491,
+      "step": 6836
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.932942654427089e-07,
+      "loss": 0.2862,
+      "step": 6837
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.916746956293786e-07,
+      "loss": 0.281,
+      "step": 6838
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.900567126012648e-07,
+      "loss": 0.3424,
+      "step": 6839
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.884403166371779e-07,
+      "loss": 0.2378,
+      "step": 6840
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.868255080156606e-07,
+      "loss": 0.2831,
+      "step": 6841
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.852122870149726e-07,
+      "loss": 0.3128,
+      "step": 6842
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.83600653913108e-07,
+      "loss": 0.2629,
+      "step": 6843
+    },
+    {
+      "epoch": 7.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.819906089877871e-07,
+      "loss": 0.3106,
+      "step": 6844
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 0.0,
+      "learning_rate": 7.80382152516449e-07,
+      "loss": 0.2626,
+      "step": 6845
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 0.0,
+      "learning_rate": 7.787752847762686e-07,
+      "loss": 0.2452,
+      "step": 6846
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 0.0,
+      "learning_rate": 7.7717000604414e-07,
+      "loss": 0.2608,
+      "step": 6847
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 0.0,
+      "learning_rate": 7.755663165966854e-07,
+      "loss": 0.2956,
+      "step": 6848
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 0.0,
+      "learning_rate": 7.739642167102546e-07,
+      "loss": 0.2518,
+      "step": 6849
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 0.0,
+      "learning_rate": 7.723637066609246e-07,
+      "loss": 0.2745,
+      "step": 6850
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 0.0,
+      "learning_rate": 7.707647867244927e-07,
+      "loss": 0.2694,
+      "step": 6851
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 0.0,
+      "learning_rate": 7.691674571764896e-07,
+      "loss": 0.2274,
+      "step": 6852
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 0.0,
+      "learning_rate": 7.675717182921649e-07,
+      "loss": 0.2709,
+      "step": 6853
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 0.0,
+      "learning_rate": 7.659775703465e-07,
+      "loss": 0.3478,
+      "step": 6854
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 0.0,
+      "learning_rate": 7.643850136141972e-07,
+      "loss": 0.2618,
+      "step": 6855
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 0.0,
+      "learning_rate": 7.627940483696894e-07,
+      "loss": 0.2423,
+      "step": 6856
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 0.0,
+      "learning_rate": 7.612046748871327e-07,
+      "loss": 0.2678,
+      "step": 6857
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 0.0,
+      "learning_rate": 7.596168934404091e-07,
+      "loss": 0.2882,
+      "step": 6858
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 0.0,
+      "learning_rate": 7.580307043031232e-07,
+      "loss": 0.2401,
+      "step": 6859
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 0.0,
+      "learning_rate": 7.564461077486107e-07,
+      "loss": 0.2298,
+      "step": 6860
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 0.0,
+      "learning_rate": 7.548631040499321e-07,
+      "loss": 0.2647,
+      "step": 6861
+    },
+    {
+      "epoch": 7.91,
+      "grad_norm": 0.0,
+      "learning_rate": 7.532816934798681e-07,
+      "loss": 0.2731,
+      "step": 6862
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 0.0,
+      "learning_rate": 7.517018763109318e-07,
+      "loss": 0.2854,
+      "step": 6863
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 0.0,
+      "learning_rate": 7.501236528153555e-07,
+      "loss": 0.2503,
+      "step": 6864
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 0.0,
+      "learning_rate": 7.485470232651027e-07,
+      "loss": 0.3096,
+      "step": 6865
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 0.0,
+      "learning_rate": 7.46971987931856e-07,
+      "loss": 0.2377,
+      "step": 6866
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 0.0,
+      "learning_rate": 7.453985470870284e-07,
+      "loss": 0.3243,
+      "step": 6867
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 0.0,
+      "learning_rate": 7.438267010017586e-07,
+      "loss": 0.2815,
+      "step": 6868
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 0.0,
+      "learning_rate": 7.422564499469065e-07,
+      "loss": 0.2978,
+      "step": 6869
+    },
+    {
+      "epoch": 7.92,
+      "grad_norm": 0.0,
+      "learning_rate": 7.406877941930557e-07,
+      "loss": 0.2548,
+      "step": 6870
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.0,
+      "learning_rate": 7.391207340105233e-07,
+      "loss": 0.2695,
+      "step": 6871
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.0,
+      "learning_rate": 7.375552696693411e-07,
+      "loss": 0.243,
+      "step": 6872
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.0,
+      "learning_rate": 7.359914014392744e-07,
+      "loss": 0.2531,
+      "step": 6873
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.0,
+      "learning_rate": 7.34429129589811e-07,
+      "loss": 0.2297,
+      "step": 6874
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.0,
+      "learning_rate": 7.328684543901598e-07,
+      "loss": 0.2751,
+      "step": 6875
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.0,
+      "learning_rate": 7.313093761092593e-07,
+      "loss": 0.2707,
+      "step": 6876
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.0,
+      "learning_rate": 7.297518950157712e-07,
+      "loss": 0.2461,
+      "step": 6877
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.0,
+      "learning_rate": 7.281960113780795e-07,
+      "loss": 0.3059,
+      "step": 6878
+    },
+    {
+      "epoch": 7.93,
+      "grad_norm": 0.0,
+      "learning_rate": 7.266417254642966e-07,
+      "loss": 0.2268,
+      "step": 6879
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 0.0,
+      "learning_rate": 7.250890375422603e-07,
+      "loss": 0.2808,
+      "step": 6880
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 0.0,
+      "learning_rate": 7.235379478795268e-07,
+      "loss": 0.2416,
+      "step": 6881
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 0.0,
+      "learning_rate": 7.219884567433844e-07,
+      "loss": 0.2233,
+      "step": 6882
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 0.0,
+      "learning_rate": 7.204405644008416e-07,
+      "loss": 0.2293,
+      "step": 6883
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 0.0,
+      "learning_rate": 7.188942711186297e-07,
+      "loss": 0.2557,
+      "step": 6884
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 0.0,
+      "learning_rate": 7.1734957716321e-07,
+      "loss": 0.2876,
+      "step": 6885
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 0.0,
+      "learning_rate": 7.15806482800766e-07,
+      "loss": 0.2962,
+      "step": 6886
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 0.0,
+      "learning_rate": 7.142649882972008e-07,
+      "loss": 0.2687,
+      "step": 6887
+    },
+    {
+      "epoch": 7.94,
+      "grad_norm": 0.0,
+      "learning_rate": 7.127250939181507e-07,
+      "loss": 0.238,
+      "step": 6888
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 0.0,
+      "learning_rate": 7.111867999289679e-07,
+      "loss": 0.2524,
+      "step": 6889
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 0.0,
+      "learning_rate": 7.096501065947326e-07,
+      "loss": 0.2733,
+      "step": 6890
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 0.0,
+      "learning_rate": 7.081150141802518e-07,
+      "loss": 0.2988,
+      "step": 6891
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 0.0,
+      "learning_rate": 7.065815229500506e-07,
+      "loss": 0.2452,
+      "step": 6892
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 0.0,
+      "learning_rate": 7.050496331683843e-07,
+      "loss": 0.2507,
+      "step": 6893
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 0.0,
+      "learning_rate": 7.035193450992262e-07,
+      "loss": 0.3273,
+      "step": 6894
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 0.0,
+      "learning_rate": 7.019906590062764e-07,
+      "loss": 0.3123,
+      "step": 6895
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 0.0,
+      "learning_rate": 7.004635751529609e-07,
+      "loss": 0.2657,
+      "step": 6896
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.0,
+      "learning_rate": 6.989380938024293e-07,
+      "loss": 0.243,
+      "step": 6897
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.0,
+      "learning_rate": 6.974142152175512e-07,
+      "loss": 0.2819,
+      "step": 6898
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.0,
+      "learning_rate": 6.958919396609231e-07,
+      "loss": 0.3032,
+      "step": 6899
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.0,
+      "learning_rate": 6.943712673948643e-07,
+      "loss": 0.2422,
+      "step": 6900
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.0,
+      "learning_rate": 6.928521986814196e-07,
+      "loss": 0.2302,
+      "step": 6901
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.0,
+      "learning_rate": 6.913347337823528e-07,
+      "loss": 0.2604,
+      "step": 6902
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.0,
+      "learning_rate": 6.898188729591582e-07,
+      "loss": 0.268,
+      "step": 6903
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.0,
+      "learning_rate": 6.883046164730489e-07,
+      "loss": 0.2366,
+      "step": 6904
+    },
+    {
+      "epoch": 7.96,
+      "grad_norm": 0.0,
+      "learning_rate": 6.867919645849619e-07,
+      "loss": 0.2701,
+      "step": 6905
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 0.0,
+      "learning_rate": 6.852809175555586e-07,
+      "loss": 0.2477,
+      "step": 6906
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 0.0,
+      "learning_rate": 6.837714756452241e-07,
+      "loss": 0.253,
+      "step": 6907
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 0.0,
+      "learning_rate": 6.822636391140658e-07,
+      "loss": 0.2666,
+      "step": 6908
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 0.0,
+      "learning_rate": 6.807574082219148e-07,
+      "loss": 0.2707,
+      "step": 6909
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 0.0,
+      "learning_rate": 6.792527832283291e-07,
+      "loss": 0.2945,
+      "step": 6910
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 0.0,
+      "learning_rate": 6.777497643925834e-07,
+      "loss": 0.2675,
+      "step": 6911
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 0.0,
+      "learning_rate": 6.762483519736806e-07,
+      "loss": 0.2603,
+      "step": 6912
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 0.0,
+      "learning_rate": 6.747485462303449e-07,
+      "loss": 0.2932,
+      "step": 6913
+    },
+    {
+      "epoch": 7.97,
+      "grad_norm": 0.0,
+      "learning_rate": 6.732503474210217e-07,
+      "loss": 0.2205,
+      "step": 6914
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 0.0,
+      "learning_rate": 6.717537558038845e-07,
+      "loss": 0.2742,
+      "step": 6915
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 0.0,
+      "learning_rate": 6.702587716368269e-07,
+      "loss": 0.2643,
+      "step": 6916
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 0.0,
+      "learning_rate": 6.687653951774642e-07,
+      "loss": 0.2563,
+      "step": 6917
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 0.0,
+      "learning_rate": 6.67273626683137e-07,
+      "loss": 0.2744,
+      "step": 6918
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 0.0,
+      "learning_rate": 6.657834664109075e-07,
+      "loss": 0.2291,
+      "step": 6919
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 0.0,
+      "learning_rate": 6.642949146175603e-07,
+      "loss": 0.2671,
+      "step": 6920
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 0.0,
+      "learning_rate": 6.62807971559607e-07,
+      "loss": 0.2726,
+      "step": 6921
+    },
+    {
+      "epoch": 7.98,
+      "grad_norm": 0.0,
+      "learning_rate": 6.613226374932758e-07,
+      "loss": 0.278,
+      "step": 6922
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 0.0,
+      "learning_rate": 6.598389126745209e-07,
+      "loss": 0.2533,
+      "step": 6923
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 0.0,
+      "learning_rate": 6.583567973590199e-07,
+      "loss": 0.2688,
+      "step": 6924
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 0.0,
+      "learning_rate": 6.568762918021688e-07,
+      "loss": 0.2933,
+      "step": 6925
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 0.0,
+      "learning_rate": 6.553973962590932e-07,
+      "loss": 0.3719,
+      "step": 6926
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 0.0,
+      "learning_rate": 6.539201109846372e-07,
+      "loss": 0.3013,
+      "step": 6927
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 0.0,
+      "learning_rate": 6.524444362333648e-07,
+      "loss": 0.2522,
+      "step": 6928
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 0.0,
+      "learning_rate": 6.509703722595684e-07,
+      "loss": 0.2817,
+      "step": 6929
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 0.0,
+      "learning_rate": 6.49497919317259e-07,
+      "loss": 0.2326,
+      "step": 6930
+    },
+    {
+      "epoch": 7.99,
+      "grad_norm": 0.0,
+      "learning_rate": 6.480270776601682e-07,
+      "loss": 0.2861,
+      "step": 6931
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6.465578475417544e-07,
+      "loss": 0.2596,
+      "step": 6932
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6.450902292151973e-07,
+      "loss": 0.2441,
+      "step": 6933
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6.436242229333967e-07,
+      "loss": 0.2746,
+      "step": 6934
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6.421598289489772e-07,
+      "loss": 0.273,
+      "step": 6935
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6.406970475142815e-07,
+      "loss": 0.2321,
+      "step": 6936
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6.392358788813802e-07,
+      "loss": 0.2481,
+      "step": 6937
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6.377763233020607e-07,
+      "loss": 0.2445,
+      "step": 6938
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6.36318381027835e-07,
+      "loss": 0.2541,
+      "step": 6939
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.0,
+      "learning_rate": 6.348620523099391e-07,
+      "loss": 0.3047,
+      "step": 6940
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 0.0,
+      "learning_rate": 6.334073373993266e-07,
+      "loss": 0.261,
+      "step": 6941
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 0.0,
+      "learning_rate": 6.319542365466747e-07,
+      "loss": 0.2693,
+      "step": 6942
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 0.0,
+      "learning_rate": 6.305027500023841e-07,
+      "loss": 0.2841,
+      "step": 6943
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 0.0,
+      "learning_rate": 6.290528780165739e-07,
+      "loss": 0.2683,
+      "step": 6944
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 0.0,
+      "learning_rate": 6.276046208390873e-07,
+      "loss": 0.2561,
+      "step": 6945
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 0.0,
+      "learning_rate": 6.261579787194927e-07,
+      "loss": 0.2489,
+      "step": 6946
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 0.0,
+      "learning_rate": 6.247129519070728e-07,
+      "loss": 0.2796,
+      "step": 6947
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 0.0,
+      "learning_rate": 6.232695406508371e-07,
+      "loss": 0.2322,
+      "step": 6948
+    },
+    {
+      "epoch": 8.01,
+      "grad_norm": 0.0,
+      "learning_rate": 6.218277451995147e-07,
+      "loss": 0.2795,
+      "step": 6949
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 0.0,
+      "learning_rate": 6.203875658015568e-07,
+      "loss": 0.2586,
+      "step": 6950
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 0.0,
+      "learning_rate": 6.18949002705137e-07,
+      "loss": 0.257,
+      "step": 6951
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 0.0,
+      "learning_rate": 6.175120561581493e-07,
+      "loss": 0.2761,
+      "step": 6952
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 0.0,
+      "learning_rate": 6.160767264082079e-07,
+      "loss": 0.2224,
+      "step": 6953
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 0.0,
+      "learning_rate": 6.146430137026527e-07,
+      "loss": 0.2163,
+      "step": 6954
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 0.0,
+      "learning_rate": 6.132109182885382e-07,
+      "loss": 0.2517,
+      "step": 6955
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 0.0,
+      "learning_rate": 6.11780440412646e-07,
+      "loss": 0.2404,
+      "step": 6956
+    },
+    {
+      "epoch": 8.02,
+      "grad_norm": 0.0,
+      "learning_rate": 6.103515803214799e-07,
+      "loss": 0.2481,
+      "step": 6957
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 0.0,
+      "learning_rate": 6.089243382612576e-07,
+      "loss": 0.2755,
+      "step": 6958
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 0.0,
+      "learning_rate": 6.074987144779254e-07,
+      "loss": 0.2587,
+      "step": 6959
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 0.0,
+      "learning_rate": 6.06074709217147e-07,
+      "loss": 0.2601,
+      "step": 6960
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 0.0,
+      "learning_rate": 6.046523227243062e-07,
+      "loss": 0.2381,
+      "step": 6961
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 0.0,
+      "learning_rate": 6.032315552445123e-07,
+      "loss": 0.261,
+      "step": 6962
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 0.0,
+      "learning_rate": 6.018124070225928e-07,
+      "loss": 0.2448,
+      "step": 6963
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 0.0,
+      "learning_rate": 6.003948783030933e-07,
+      "loss": 0.3215,
+      "step": 6964
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.989789693302872e-07,
+      "loss": 0.2688,
+      "step": 6965
+    },
+    {
+      "epoch": 8.03,
+      "grad_norm": 0.0,
+      "learning_rate": 5.975646803481639e-07,
+      "loss": 0.2584,
+      "step": 6966
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.961520116004326e-07,
+      "loss": 0.2481,
+      "step": 6967
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.947409633305268e-07,
+      "loss": 0.259,
+      "step": 6968
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.933315357816005e-07,
+      "loss": 0.2612,
+      "step": 6969
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.919237291965252e-07,
+      "loss": 0.2401,
+      "step": 6970
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.905175438178979e-07,
+      "loss": 0.2794,
+      "step": 6971
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.891129798880291e-07,
+      "loss": 0.3099,
+      "step": 6972
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.877100376489597e-07,
+      "loss": 0.2472,
+      "step": 6973
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.863087173424409e-07,
+      "loss": 0.2207,
+      "step": 6974
+    },
+    {
+      "epoch": 8.04,
+      "grad_norm": 0.0,
+      "learning_rate": 5.849090192099505e-07,
+      "loss": 0.2344,
+      "step": 6975
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.835109434926889e-07,
+      "loss": 0.3101,
+      "step": 6976
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.821144904315701e-07,
+      "loss": 0.293,
+      "step": 6977
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.807196602672305e-07,
+      "loss": 0.2375,
+      "step": 6978
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.793264532400311e-07,
+      "loss": 0.2762,
+      "step": 6979
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.779348695900511e-07,
+      "loss": 0.2296,
+      "step": 6980
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.765449095570863e-07,
+      "loss": 0.2283,
+      "step": 6981
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.751565733806575e-07,
+      "loss": 0.2809,
+      "step": 6982
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 0.0,
+      "learning_rate": 5.737698613000031e-07,
+      "loss": 0.305,
+      "step": 6983
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.723847735540833e-07,
+      "loss": 0.268,
+      "step": 6984
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.710013103815748e-07,
+      "loss": 0.2143,
+      "step": 6985
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.696194720208792e-07,
+      "loss": 0.2735,
+      "step": 6986
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.68239258710116e-07,
+      "loss": 0.2659,
+      "step": 6987
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.66860670687125e-07,
+      "loss": 0.2545,
+      "step": 6988
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.654837081894626e-07,
+      "loss": 0.2699,
+      "step": 6989
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.641083714544116e-07,
+      "loss": 0.2365,
+      "step": 6990
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.627346607189665e-07,
+      "loss": 0.2722,
+      "step": 6991
+    },
+    {
+      "epoch": 8.06,
+      "grad_norm": 0.0,
+      "learning_rate": 5.613625762198505e-07,
+      "loss": 0.2571,
+      "step": 6992
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.599921181935009e-07,
+      "loss": 0.2625,
+      "step": 6993
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.586232868760743e-07,
+      "loss": 0.2671,
+      "step": 6994
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.572560825034523e-07,
+      "loss": 0.2666,
+      "step": 6995
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.558905053112296e-07,
+      "loss": 0.2593,
+      "step": 6996
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.545265555347235e-07,
+      "loss": 0.2284,
+      "step": 6997
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.531642334089704e-07,
+      "loss": 0.291,
+      "step": 6998
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.518035391687293e-07,
+      "loss": 0.2289,
+      "step": 6999
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.504444730484725e-07,
+      "loss": 0.237,
+      "step": 7000
+    },
+    {
+      "epoch": 8.07,
+      "grad_norm": 0.0,
+      "learning_rate": 5.490870352823996e-07,
+      "loss": 0.2215,
+      "step": 7001
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.477312261044221e-07,
+      "loss": 0.2299,
+      "step": 7002
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.463770457481732e-07,
+      "loss": 0.2663,
+      "step": 7003
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.450244944470073e-07,
+      "loss": 0.2394,
+      "step": 7004
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.436735724339981e-07,
+      "loss": 0.2885,
+      "step": 7005
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.423242799419382e-07,
+      "loss": 0.2613,
+      "step": 7006
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.40976617203336e-07,
+      "loss": 0.2608,
+      "step": 7007
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.396305844504223e-07,
+      "loss": 0.2555,
+      "step": 7008
+    },
+    {
+      "epoch": 8.08,
+      "grad_norm": 0.0,
+      "learning_rate": 5.382861819151475e-07,
+      "loss": 0.2607,
+      "step": 7009
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.369434098291804e-07,
+      "loss": 0.2432,
+      "step": 7010
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.35602268423906e-07,
+      "loss": 0.2117,
+      "step": 7011
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.342627579304338e-07,
+      "loss": 0.233,
+      "step": 7012
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.329248785795882e-07,
+      "loss": 0.2685,
+      "step": 7013
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.315886306019113e-07,
+      "loss": 0.2381,
+      "step": 7014
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.302540142276679e-07,
+      "loss": 0.2689,
+      "step": 7015
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.289210296868418e-07,
+      "loss": 0.2479,
+      "step": 7016
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.275896772091316e-07,
+      "loss": 0.278,
+      "step": 7017
+    },
+    {
+      "epoch": 8.09,
+      "grad_norm": 0.0,
+      "learning_rate": 5.26259957023958e-07,
+      "loss": 0.2556,
+      "step": 7018
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 0.0,
+      "learning_rate": 5.249318693604577e-07,
+      "loss": 0.2344,
+      "step": 7019
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 0.0,
+      "learning_rate": 5.236054144474911e-07,
+      "loss": 0.2474,
+      "step": 7020
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 0.0,
+      "learning_rate": 5.222805925136298e-07,
+      "loss": 0.2611,
+      "step": 7021
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 0.0,
+      "learning_rate": 5.209574037871701e-07,
+      "loss": 0.2709,
+      "step": 7022
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 0.0,
+      "learning_rate": 5.196358484961251e-07,
+      "loss": 0.247,
+      "step": 7023
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 0.0,
+      "learning_rate": 5.183159268682269e-07,
+      "loss": 0.2575,
+      "step": 7024
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 0.0,
+      "learning_rate": 5.169976391309228e-07,
+      "loss": 0.2813,
+      "step": 7025
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 0.0,
+      "learning_rate": 5.156809855113832e-07,
+      "loss": 0.2584,
+      "step": 7026
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 0.0,
+      "learning_rate": 5.143659662364931e-07,
+      "loss": 0.2519,
+      "step": 7027
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 0.0,
+      "learning_rate": 5.13052581532858e-07,
+      "loss": 0.2308,
+      "step": 7028
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 0.0,
+      "learning_rate": 5.117408316268047e-07,
+      "loss": 0.2406,
+      "step": 7029
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 0.0,
+      "learning_rate": 5.104307167443689e-07,
+      "loss": 0.264,
+      "step": 7030
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 0.0,
+      "learning_rate": 5.091222371113158e-07,
+      "loss": 0.2575,
+      "step": 7031
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 0.0,
+      "learning_rate": 5.078153929531216e-07,
+      "loss": 0.2822,
+      "step": 7032
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 0.0,
+      "learning_rate": 5.065101844949793e-07,
+      "loss": 0.2774,
+      "step": 7033
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 0.0,
+      "learning_rate": 5.052066119618093e-07,
+      "loss": 0.2261,
+      "step": 7034
+    },
+    {
+      "epoch": 8.11,
+      "grad_norm": 0.0,
+      "learning_rate": 5.039046755782417e-07,
+      "loss": 0.2436,
+      "step": 7035
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 0.0,
+      "learning_rate": 5.026043755686249e-07,
+      "loss": 0.2179,
+      "step": 7036
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 0.0,
+      "learning_rate": 5.013057121570308e-07,
+      "loss": 0.2761,
+      "step": 7037
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 0.0,
+      "learning_rate": 5.000086855672437e-07,
+      "loss": 0.2501,
+      "step": 7038
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.987132960227681e-07,
+      "loss": 0.2567,
+      "step": 7039
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.974195437468298e-07,
+      "loss": 0.2499,
+      "step": 7040
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.961274289623641e-07,
+      "loss": 0.2449,
+      "step": 7041
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.948369518920349e-07,
+      "loss": 0.244,
+      "step": 7042
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.935481127582131e-07,
+      "loss": 0.2832,
+      "step": 7043
+    },
+    {
+      "epoch": 8.12,
+      "grad_norm": 0.0,
+      "learning_rate": 4.922609117829946e-07,
+      "loss": 0.24,
+      "step": 7044
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.909753491881896e-07,
+      "loss": 0.2578,
+      "step": 7045
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.896914251953288e-07,
+      "loss": 0.234,
+      "step": 7046
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.884091400256574e-07,
+      "loss": 0.2947,
+      "step": 7047
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.871284939001419e-07,
+      "loss": 0.2972,
+      "step": 7048
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.858494870394625e-07,
+      "loss": 0.2998,
+      "step": 7049
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.845721196640174e-07,
+      "loss": 0.298,
+      "step": 7050
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.83296391993926e-07,
+      "loss": 0.2375,
+      "step": 7051
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.820223042490214e-07,
+      "loss": 0.3099,
+      "step": 7052
+    },
+    {
+      "epoch": 8.13,
+      "grad_norm": 0.0,
+      "learning_rate": 4.807498566488545e-07,
+      "loss": 0.2935,
+      "step": 7053
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.794790494126966e-07,
+      "loss": 0.2163,
+      "step": 7054
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.782098827595305e-07,
+      "loss": 0.2429,
+      "step": 7055
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.769423569080645e-07,
+      "loss": 0.2541,
+      "step": 7056
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.75676472076716e-07,
+      "loss": 0.2604,
+      "step": 7057
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.744122284836239e-07,
+      "loss": 0.2602,
+      "step": 7058
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7314962634664616e-07,
+      "loss": 0.256,
+      "step": 7059
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7188866588335213e-07,
+      "loss": 0.2577,
+      "step": 7060
+    },
+    {
+      "epoch": 8.14,
+      "grad_norm": 0.0,
+      "learning_rate": 4.706293473110313e-07,
+      "loss": 0.2614,
+      "step": 7061
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6937167084669334e-07,
+      "loss": 0.2481,
+      "step": 7062
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.681156367070594e-07,
+      "loss": 0.2386,
+      "step": 7063
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6686124510856967e-07,
+      "loss": 0.2437,
+      "step": 7064
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.656084962673857e-07,
+      "loss": 0.253,
+      "step": 7065
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.64357390399377e-07,
+      "loss": 0.2423,
+      "step": 7066
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.631079277201389e-07,
+      "loss": 0.2658,
+      "step": 7067
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6186010844497696e-07,
+      "loss": 0.2372,
+      "step": 7068
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.606139327889181e-07,
+      "loss": 0.2466,
+      "step": 7069
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 0.0,
+      "learning_rate": 4.593694009667049e-07,
+      "loss": 0.2949,
+      "step": 7070
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.581265131927948e-07,
+      "loss": 0.2685,
+      "step": 7071
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5688526968136193e-07,
+      "loss": 0.2384,
+      "step": 7072
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5564567064630085e-07,
+      "loss": 0.2915,
+      "step": 7073
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5440771630121617e-07,
+      "loss": 0.2489,
+      "step": 7074
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5317140685943726e-07,
+      "loss": 0.2582,
+      "step": 7075
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.519367425340049e-07,
+      "loss": 0.2568,
+      "step": 7076
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5070372353767543e-07,
+      "loss": 0.2243,
+      "step": 7077
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4947235008292565e-07,
+      "loss": 0.2433,
+      "step": 7078
+    },
+    {
+      "epoch": 8.16,
+      "grad_norm": 0.0,
+      "learning_rate": 4.482426223819458e-07,
+      "loss": 0.2095,
+      "step": 7079
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.470145406466431e-07,
+      "loss": 0.2635,
+      "step": 7080
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4578810508864145e-07,
+      "loss": 0.2445,
+      "step": 7081
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.445633159192819e-07,
+      "loss": 0.2279,
+      "step": 7082
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.433401733496201e-07,
+      "loss": 0.253,
+      "step": 7083
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.421186775904307e-07,
+      "loss": 0.2614,
+      "step": 7084
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4089882885220204e-07,
+      "loss": 0.2309,
+      "step": 7085
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3968062734513705e-07,
+      "loss": 0.2671,
+      "step": 7086
+    },
+    {
+      "epoch": 8.17,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3846407327915784e-07,
+      "loss": 0.2455,
+      "step": 7087
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.372491668639034e-07,
+      "loss": 0.2641,
+      "step": 7088
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3603590830872843e-07,
+      "loss": 0.2565,
+      "step": 7089
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.348242978227002e-07,
+      "loss": 0.2525,
+      "step": 7090
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3361433561460274e-07,
+      "loss": 0.2163,
+      "step": 7091
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3240602189294267e-07,
+      "loss": 0.2267,
+      "step": 7092
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.311993568659323e-07,
+      "loss": 0.2654,
+      "step": 7093
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.299943407415064e-07,
+      "loss": 0.2622,
+      "step": 7094
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.287909737273177e-07,
+      "loss": 0.2638,
+      "step": 7095
+    },
+    {
+      "epoch": 8.18,
+      "grad_norm": 0.0,
+      "learning_rate": 4.275892560307282e-07,
+      "loss": 0.3182,
+      "step": 7096
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.263891878588178e-07,
+      "loss": 0.2157,
+      "step": 7097
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2519076941838435e-07,
+      "loss": 0.2569,
+      "step": 7098
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2399400091594154e-07,
+      "loss": 0.2422,
+      "step": 7099
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.227988825577145e-07,
+      "loss": 0.2524,
+      "step": 7100
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2160541454964954e-07,
+      "loss": 0.247,
+      "step": 7101
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2041359709740436e-07,
+      "loss": 0.2414,
+      "step": 7102
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.192234304063558e-07,
+      "loss": 0.2577,
+      "step": 7103
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.18034914681591e-07,
+      "loss": 0.2505,
+      "step": 7104
+    },
+    {
+      "epoch": 8.19,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1684805012791706e-07,
+      "loss": 0.294,
+      "step": 7105
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.156628369498583e-07,
+      "loss": 0.2946,
+      "step": 7106
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.14479275351648e-07,
+      "loss": 0.2255,
+      "step": 7107
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1329736553723853e-07,
+      "loss": 0.2411,
+      "step": 7108
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1211710771030034e-07,
+      "loss": 0.2813,
+      "step": 7109
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.109385020742118e-07,
+      "loss": 0.2224,
+      "step": 7110
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0976154883207497e-07,
+      "loss": 0.2391,
+      "step": 7111
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.08586248186702e-07,
+      "loss": 0.2446,
+      "step": 7112
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 0.0,
+      "learning_rate": 4.074126003406198e-07,
+      "loss": 0.2601,
+      "step": 7113
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.062406054960766e-07,
+      "loss": 0.2463,
+      "step": 7114
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.2853,
+      "step": 7115
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.039015756191467e-07,
+      "loss": 0.24,
+      "step": 7116
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0273454098982424e-07,
+      "loss": 0.2319,
+      "step": 7117
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0156916016816594e-07,
+      "loss": 0.277,
+      "step": 7118
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0040543335498783e-07,
+      "loss": 0.2422,
+      "step": 7119
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9924336075082835e-07,
+      "loss": 0.2397,
+      "step": 7120
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9808294255593293e-07,
+      "loss": 0.2822,
+      "step": 7121
+    },
+    {
+      "epoch": 8.21,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9692417897026713e-07,
+      "loss": 0.2487,
+      "step": 7122
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9576707019350903e-07,
+      "loss": 0.3258,
+      "step": 7123
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 0.0,
+      "learning_rate": 3.946116164250535e-07,
+      "loss": 0.2629,
+      "step": 7124
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 0.0,
+      "learning_rate": 3.934578178640103e-07,
+      "loss": 0.275,
+      "step": 7125
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9230567470920134e-07,
+      "loss": 0.2147,
+      "step": 7126
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9115518715916345e-07,
+      "loss": 0.2698,
+      "step": 7127
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9000635541215135e-07,
+      "loss": 0.2641,
+      "step": 7128
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8885917966613227e-07,
+      "loss": 0.2651,
+      "step": 7129
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8771366011878807e-07,
+      "loss": 0.313,
+      "step": 7130
+    },
+    {
+      "epoch": 8.22,
+      "grad_norm": 0.0,
+      "learning_rate": 3.865697969675164e-07,
+      "loss": 0.2585,
+      "step": 7131
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 0.0,
+      "learning_rate": 3.854275904094273e-07,
+      "loss": 0.2453,
+      "step": 7132
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 0.0,
+      "learning_rate": 3.842870406413457e-07,
+      "loss": 0.2533,
+      "step": 7133
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8314814785981424e-07,
+      "loss": 0.236,
+      "step": 7134
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 0.0,
+      "learning_rate": 3.82010912261086e-07,
+      "loss": 0.265,
+      "step": 7135
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 0.0,
+      "learning_rate": 3.808753340411331e-07,
+      "loss": 0.2393,
+      "step": 7136
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7974141339563673e-07,
+      "loss": 0.2592,
+      "step": 7137
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 0.0,
+      "learning_rate": 3.78609150519994e-07,
+      "loss": 0.2602,
+      "step": 7138
+    },
+    {
+      "epoch": 8.23,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7747854560931996e-07,
+      "loss": 0.278,
+      "step": 7139
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 0.0,
+      "learning_rate": 3.763495988584387e-07,
+      "loss": 0.2625,
+      "step": 7140
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7522231046189237e-07,
+      "loss": 0.2352,
+      "step": 7141
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7409668061393787e-07,
+      "loss": 0.27,
+      "step": 7142
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7297270950854224e-07,
+      "loss": 0.2553,
+      "step": 7143
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 0.0,
+      "learning_rate": 3.718503973393883e-07,
+      "loss": 0.2417,
+      "step": 7144
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 0.0,
+      "learning_rate": 3.707297442998758e-07,
+      "loss": 0.2663,
+      "step": 7145
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 0.0,
+      "learning_rate": 3.696107505831148e-07,
+      "loss": 0.2384,
+      "step": 7146
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 0.0,
+      "learning_rate": 3.684934163819309e-07,
+      "loss": 0.2651,
+      "step": 7147
+    },
+    {
+      "epoch": 8.24,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6737774188886686e-07,
+      "loss": 0.308,
+      "step": 7148
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6626372729617335e-07,
+      "loss": 0.2757,
+      "step": 7149
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6515137279582e-07,
+      "loss": 0.2401,
+      "step": 7150
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.0,
+      "learning_rate": 3.640406785794892e-07,
+      "loss": 0.2112,
+      "step": 7151
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.0,
+      "learning_rate": 3.629316448385711e-07,
+      "loss": 0.2601,
+      "step": 7152
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.0,
+      "learning_rate": 3.618242717641829e-07,
+      "loss": 0.2195,
+      "step": 7153
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.0,
+      "learning_rate": 3.607185595471441e-07,
+      "loss": 0.2447,
+      "step": 7154
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.0,
+      "learning_rate": 3.596145083779912e-07,
+      "loss": 0.2296,
+      "step": 7155
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5851211844697643e-07,
+      "loss": 0.2512,
+      "step": 7156
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5741138994406346e-07,
+      "loss": 0.2961,
+      "step": 7157
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5631232305893047e-07,
+      "loss": 0.2458,
+      "step": 7158
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5521491798097163e-07,
+      "loss": 0.2645,
+      "step": 7159
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5411917489929003e-07,
+      "loss": 0.2664,
+      "step": 7160
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5302509400270693e-07,
+      "loss": 0.3013,
+      "step": 7161
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 0.0,
+      "learning_rate": 3.519326754797547e-07,
+      "loss": 0.247,
+      "step": 7162
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 0.0,
+      "learning_rate": 3.508419195186774e-07,
+      "loss": 0.2773,
+      "step": 7163
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 0.0,
+      "learning_rate": 3.497528263074368e-07,
+      "loss": 0.2315,
+      "step": 7164
+    },
+    {
+      "epoch": 8.26,
+      "grad_norm": 0.0,
+      "learning_rate": 3.486653960337061e-07,
+      "loss": 0.2446,
+      "step": 7165
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4757962888487205e-07,
+      "loss": 0.2378,
+      "step": 7166
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 0.0,
+      "learning_rate": 3.464955250480351e-07,
+      "loss": 0.2298,
+      "step": 7167
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 0.0,
+      "learning_rate": 3.45413084710009e-07,
+      "loss": 0.2555,
+      "step": 7168
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 0.0,
+      "learning_rate": 3.443323080573191e-07,
+      "loss": 0.2454,
+      "step": 7169
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 0.0,
+      "learning_rate": 3.432531952762064e-07,
+      "loss": 0.2896,
+      "step": 7170
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 0.0,
+      "learning_rate": 3.421757465526243e-07,
+      "loss": 0.3071,
+      "step": 7171
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4109996207224103e-07,
+      "loss": 0.3338,
+      "step": 7172
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 0.0,
+      "learning_rate": 3.40025842020435e-07,
+      "loss": 0.2619,
+      "step": 7173
+    },
+    {
+      "epoch": 8.27,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3895338658229915e-07,
+      "loss": 0.2471,
+      "step": 7174
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3788259594264126e-07,
+      "loss": 0.2401,
+      "step": 7175
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 0.0,
+      "learning_rate": 3.368134702859782e-07,
+      "loss": 0.2325,
+      "step": 7176
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 0.0,
+      "learning_rate": 3.357460097965448e-07,
+      "loss": 0.2794,
+      "step": 7177
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3468021465828617e-07,
+      "loss": 0.2771,
+      "step": 7178
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 0.0,
+      "learning_rate": 3.33616085054862e-07,
+      "loss": 0.2975,
+      "step": 7179
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3255362116964006e-07,
+      "loss": 0.2687,
+      "step": 7180
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 0.0,
+      "learning_rate": 3.314928231857084e-07,
+      "loss": 0.2788,
+      "step": 7181
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 0.0,
+      "learning_rate": 3.30433691285863e-07,
+      "loss": 0.2449,
+      "step": 7182
+    },
+    {
+      "epoch": 8.28,
+      "grad_norm": 0.0,
+      "learning_rate": 3.293762256526134e-07,
+      "loss": 0.272,
+      "step": 7183
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2832042646818605e-07,
+      "loss": 0.3046,
+      "step": 7184
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2726629391451436e-07,
+      "loss": 0.2682,
+      "step": 7185
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2621382817324743e-07,
+      "loss": 0.2761,
+      "step": 7186
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2516302942574794e-07,
+      "loss": 0.2482,
+      "step": 7187
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2411389785308887e-07,
+      "loss": 0.2845,
+      "step": 7188
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2306643363605893e-07,
+      "loss": 0.2347,
+      "step": 7189
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2202063695515707e-07,
+      "loss": 0.236,
+      "step": 7190
+    },
+    {
+      "epoch": 8.29,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2097650799059464e-07,
+      "loss": 0.237,
+      "step": 7191
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 0.0,
+      "learning_rate": 3.199340469223e-07,
+      "loss": 0.2571,
+      "step": 7192
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1889325392990613e-07,
+      "loss": 0.253,
+      "step": 7193
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1785412919276617e-07,
+      "loss": 0.2645,
+      "step": 7194
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1681667288994353e-07,
+      "loss": 0.2717,
+      "step": 7195
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 0.0,
+      "learning_rate": 3.15780885200212e-07,
+      "loss": 0.2454,
+      "step": 7196
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 0.0,
+      "learning_rate": 3.147467663020598e-07,
+      "loss": 0.2485,
+      "step": 7197
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1371431637368664e-07,
+      "loss": 0.2971,
+      "step": 7198
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1268353559300467e-07,
+      "loss": 0.2285,
+      "step": 7199
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 0.0,
+      "learning_rate": 3.116544241376385e-07,
+      "loss": 0.2273,
+      "step": 7200
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.106269821849273e-07,
+      "loss": 0.2436,
+      "step": 7201
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0960120991191833e-07,
+      "loss": 0.2728,
+      "step": 7202
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0857710749537585e-07,
+      "loss": 0.2249,
+      "step": 7203
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0755467511177197e-07,
+      "loss": 0.2405,
+      "step": 7204
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.065339129372935e-07,
+      "loss": 0.2797,
+      "step": 7205
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.055148211478376e-07,
+      "loss": 0.2702,
+      "step": 7206
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0449739991901703e-07,
+      "loss": 0.2831,
+      "step": 7207
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.03481649426155e-07,
+      "loss": 0.26,
+      "step": 7208
+    },
+    {
+      "epoch": 8.31,
+      "grad_norm": 0.0,
+      "learning_rate": 3.024675698442858e-07,
+      "loss": 0.2417,
+      "step": 7209
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.014551613481531e-07,
+      "loss": 0.2733,
+      "step": 7210
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0044442411222066e-07,
+      "loss": 0.2375,
+      "step": 7211
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 0.0,
+      "learning_rate": 2.994353583106557e-07,
+      "loss": 0.2666,
+      "step": 7212
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9842796411734263e-07,
+      "loss": 0.2475,
+      "step": 7213
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9742224170587697e-07,
+      "loss": 0.2273,
+      "step": 7214
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 0.0,
+      "learning_rate": 2.964181912495656e-07,
+      "loss": 0.2573,
+      "step": 7215
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 0.0,
+      "learning_rate": 2.954158129214257e-07,
+      "loss": 0.2894,
+      "step": 7216
+    },
+    {
+      "epoch": 8.32,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9441510689418806e-07,
+      "loss": 0.2501,
+      "step": 7217
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9341607334029685e-07,
+      "loss": 0.2431,
+      "step": 7218
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9241871243190555e-07,
+      "loss": 0.2393,
+      "step": 7219
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.0,
+      "learning_rate": 2.914230243408789e-07,
+      "loss": 0.2416,
+      "step": 7220
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9042900923879515e-07,
+      "loss": 0.2395,
+      "step": 7221
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.0,
+      "learning_rate": 2.89436667296944e-07,
+      "loss": 0.2436,
+      "step": 7222
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.0,
+      "learning_rate": 2.884459986863264e-07,
+      "loss": 0.2562,
+      "step": 7223
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.0,
+      "learning_rate": 2.874570035776547e-07,
+      "loss": 0.2471,
+      "step": 7224
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.0,
+      "learning_rate": 2.864696821413548e-07,
+      "loss": 0.2495,
+      "step": 7225
+    },
+    {
+      "epoch": 8.33,
+      "grad_norm": 0.0,
+      "learning_rate": 2.854840345475607e-07,
+      "loss": 0.2483,
+      "step": 7226
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 0.0,
+      "learning_rate": 2.845000609661208e-07,
+      "loss": 0.2425,
+      "step": 7227
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8351776156659407e-07,
+      "loss": 0.257,
+      "step": 7228
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 0.0,
+      "learning_rate": 2.825371365182505e-07,
+      "loss": 0.2634,
+      "step": 7229
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 0.0,
+      "learning_rate": 2.815581859900729e-07,
+      "loss": 0.2954,
+      "step": 7230
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 0.0,
+      "learning_rate": 2.805809101507539e-07,
+      "loss": 0.2362,
+      "step": 7231
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 0.0,
+      "learning_rate": 2.79605309168699e-07,
+      "loss": 0.2145,
+      "step": 7232
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 0.0,
+      "learning_rate": 2.786313832120258e-07,
+      "loss": 0.2315,
+      "step": 7233
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7765913244856e-07,
+      "loss": 0.2683,
+      "step": 7234
+    },
+    {
+      "epoch": 8.34,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7668855704583997e-07,
+      "loss": 0.2356,
+      "step": 7235
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 0.0,
+      "learning_rate": 2.757196571711163e-07,
+      "loss": 0.2438,
+      "step": 7236
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7475243299135315e-07,
+      "loss": 0.2684,
+      "step": 7237
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7378688467321945e-07,
+      "loss": 0.2485,
+      "step": 7238
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7282301238310326e-07,
+      "loss": 0.2431,
+      "step": 7239
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7186081628709615e-07,
+      "loss": 0.3008,
+      "step": 7240
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7090029655100434e-07,
+      "loss": 0.2786,
+      "step": 7241
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6994145334034993e-07,
+      "loss": 0.2397,
+      "step": 7242
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 0.0,
+      "learning_rate": 2.689842868203563e-07,
+      "loss": 0.3068,
+      "step": 7243
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6802879715596585e-07,
+      "loss": 0.23,
+      "step": 7244
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 0.0,
+      "learning_rate": 2.670749845118281e-07,
+      "loss": 0.2305,
+      "step": 7245
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 0.0,
+      "learning_rate": 2.661228490523049e-07,
+      "loss": 0.2219,
+      "step": 7246
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 0.0,
+      "learning_rate": 2.651723909414694e-07,
+      "loss": 0.2942,
+      "step": 7247
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 0.0,
+      "learning_rate": 2.642236103431062e-07,
+      "loss": 0.3007,
+      "step": 7248
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6327650742070665e-07,
+      "loss": 0.2823,
+      "step": 7249
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6233108233747916e-07,
+      "loss": 0.2937,
+      "step": 7250
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6138733525633896e-07,
+      "loss": 0.2419,
+      "step": 7251
+    },
+    {
+      "epoch": 8.36,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6044526633991264e-07,
+      "loss": 0.2995,
+      "step": 7252
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.0,
+      "learning_rate": 2.595048757505392e-07,
+      "loss": 0.2536,
+      "step": 7253
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5856616365026674e-07,
+      "loss": 0.2535,
+      "step": 7254
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.0,
+      "learning_rate": 2.576291302008571e-07,
+      "loss": 0.2378,
+      "step": 7255
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.0,
+      "learning_rate": 2.566937755637777e-07,
+      "loss": 0.3048,
+      "step": 7256
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5576009990020856e-07,
+      "loss": 0.2961,
+      "step": 7257
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5482810337104534e-07,
+      "loss": 0.2485,
+      "step": 7258
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5389778613688744e-07,
+      "loss": 0.2514,
+      "step": 7259
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5296914835804766e-07,
+      "loss": 0.2902,
+      "step": 7260
+    },
+    {
+      "epoch": 8.37,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5204219019455243e-07,
+      "loss": 0.3009,
+      "step": 7261
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5111691180613296e-07,
+      "loss": 0.267,
+      "step": 7262
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5019331335223383e-07,
+      "loss": 0.2558,
+      "step": 7263
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4927139499201225e-07,
+      "loss": 0.2392,
+      "step": 7264
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 0.0,
+      "learning_rate": 2.483511568843322e-07,
+      "loss": 0.2601,
+      "step": 7265
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 0.0,
+      "learning_rate": 2.474325991877702e-07,
+      "loss": 0.2264,
+      "step": 7266
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 0.0,
+      "learning_rate": 2.46515722060614e-07,
+      "loss": 0.2421,
+      "step": 7267
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4560052566085736e-07,
+      "loss": 0.244,
+      "step": 7268
+    },
+    {
+      "epoch": 8.38,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4468701014621177e-07,
+      "loss": 0.2709,
+      "step": 7269
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 0.0,
+      "learning_rate": 2.437751756740925e-07,
+      "loss": 0.2427,
+      "step": 7270
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 0.0,
+      "learning_rate": 2.42865022401626e-07,
+      "loss": 0.254,
+      "step": 7271
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 0.0,
+      "learning_rate": 2.419565504856536e-07,
+      "loss": 0.2383,
+      "step": 7272
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4104976008272217e-07,
+      "loss": 0.2531,
+      "step": 7273
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4014465134909125e-07,
+      "loss": 0.2408,
+      "step": 7274
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 0.0,
+      "learning_rate": 2.392412244407294e-07,
+      "loss": 0.2396,
+      "step": 7275
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3833947951331427e-07,
+      "loss": 0.2321,
+      "step": 7276
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3743941672223825e-07,
+      "loss": 0.2342,
+      "step": 7277
+    },
+    {
+      "epoch": 8.39,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3654103622259838e-07,
+      "loss": 0.2796,
+      "step": 7278
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3564433816920528e-07,
+      "loss": 0.2445,
+      "step": 7279
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.0,
+      "learning_rate": 2.347493227165798e-07,
+      "loss": 0.2689,
+      "step": 7280
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3385599001894965e-07,
+      "loss": 0.2179,
+      "step": 7281
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3296434023025504e-07,
+      "loss": 0.25,
+      "step": 7282
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3207437350414418e-07,
+      "loss": 0.2332,
+      "step": 7283
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3118608999398106e-07,
+      "loss": 0.2716,
+      "step": 7284
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.0,
+      "learning_rate": 2.30299489852831e-07,
+      "loss": 0.2832,
+      "step": 7285
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2941457323347627e-07,
+      "loss": 0.2456,
+      "step": 7286
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2853134028840594e-07,
+      "loss": 0.2423,
+      "step": 7287
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2764979116981723e-07,
+      "loss": 0.2716,
+      "step": 7288
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2676992602962078e-07,
+      "loss": 0.2888,
+      "step": 7289
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2589174501943644e-07,
+      "loss": 0.2445,
+      "step": 7290
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2501524829059208e-07,
+      "loss": 0.2666,
+      "step": 7291
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2414043599412682e-07,
+      "loss": 0.2459,
+      "step": 7292
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2326730828078792e-07,
+      "loss": 0.2321,
+      "step": 7293
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2239586530103497e-07,
+      "loss": 0.2321,
+      "step": 7294
+    },
+    {
+      "epoch": 8.41,
+      "grad_norm": 0.0,
+      "learning_rate": 2.215261072050323e-07,
+      "loss": 0.2452,
+      "step": 7295
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 0.0,
+      "learning_rate": 2.206580341426612e-07,
+      "loss": 0.2219,
+      "step": 7296
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1979164626350745e-07,
+      "loss": 0.2553,
+      "step": 7297
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 0.0,
+      "learning_rate": 2.189269437168673e-07,
+      "loss": 0.2149,
+      "step": 7298
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 0.0,
+      "learning_rate": 2.180639266517448e-07,
+      "loss": 0.2268,
+      "step": 7299
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1720259521685995e-07,
+      "loss": 0.2436,
+      "step": 7300
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1634294956063284e-07,
+      "loss": 0.2542,
+      "step": 7301
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1548498983120393e-07,
+      "loss": 0.2298,
+      "step": 7302
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1462871617641379e-07,
+      "loss": 0.2145,
+      "step": 7303
+    },
+    {
+      "epoch": 8.42,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1377412874381664e-07,
+      "loss": 0.2364,
+      "step": 7304
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 0.0,
+      "learning_rate": 2.129212276806769e-07,
+      "loss": 0.2721,
+      "step": 7305
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 0.0,
+      "learning_rate": 2.120700131339648e-07,
+      "loss": 0.3117,
+      "step": 7306
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1122048525036409e-07,
+      "loss": 0.2531,
+      "step": 7307
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1037264417626547e-07,
+      "loss": 0.2219,
+      "step": 7308
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0952649005777093e-07,
+      "loss": 0.2291,
+      "step": 7309
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 0.0,
+      "learning_rate": 2.086820230406872e-07,
+      "loss": 0.2602,
+      "step": 7310
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0783924327053562e-07,
+      "loss": 0.2494,
+      "step": 7311
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 0.0,
+      "learning_rate": 2.069981508925434e-07,
+      "loss": 0.2442,
+      "step": 7312
+    },
+    {
+      "epoch": 8.43,
+      "grad_norm": 0.0,
+      "learning_rate": 2.061587460516501e-07,
+      "loss": 0.2632,
+      "step": 7313
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0532102889250116e-07,
+      "loss": 0.2939,
+      "step": 7314
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0448499955945223e-07,
+      "loss": 0.2875,
+      "step": 7315
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0365065819657025e-07,
+      "loss": 0.2477,
+      "step": 7316
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0281800494762803e-07,
+      "loss": 0.2934,
+      "step": 7317
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 0.0,
+      "learning_rate": 2.019870399561097e-07,
+      "loss": 0.2541,
+      "step": 7318
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0115776336520622e-07,
+      "loss": 0.217,
+      "step": 7319
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0033017531782108e-07,
+      "loss": 0.2549,
+      "step": 7320
+    },
+    {
+      "epoch": 8.44,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9950427595656352e-07,
+      "loss": 0.2753,
+      "step": 7321
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9868006542375417e-07,
+      "loss": 0.2235,
+      "step": 7322
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9785754386142164e-07,
+      "loss": 0.2516,
+      "step": 7323
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9703671141130144e-07,
+      "loss": 0.291,
+      "step": 7324
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.962175682148426e-07,
+      "loss": 0.3185,
+      "step": 7325
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9540011441319784e-07,
+      "loss": 0.2548,
+      "step": 7326
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9458435014723443e-07,
+      "loss": 0.2639,
+      "step": 7327
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9377027555752327e-07,
+      "loss": 0.2883,
+      "step": 7328
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9295789078434658e-07,
+      "loss": 0.2521,
+      "step": 7329
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 0.0,
+      "learning_rate": 1.921471959676957e-07,
+      "loss": 0.2372,
+      "step": 7330
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9133819124727003e-07,
+      "loss": 0.2984,
+      "step": 7331
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9053087676247696e-07,
+      "loss": 0.2687,
+      "step": 7332
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8972525265243512e-07,
+      "loss": 0.2209,
+      "step": 7333
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.88921319055968e-07,
+      "loss": 0.2315,
+      "step": 7334
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.881190761116125e-07,
+      "loss": 0.235,
+      "step": 7335
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8731852395760918e-07,
+      "loss": 0.2186,
+      "step": 7336
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8651966273191103e-07,
+      "loss": 0.2697,
+      "step": 7337
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.857224925721801e-07,
+      "loss": 0.2275,
+      "step": 7338
+    },
+    {
+      "epoch": 8.46,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8492701361578326e-07,
+      "loss": 0.2482,
+      "step": 7339
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8413322599979745e-07,
+      "loss": 0.2299,
+      "step": 7340
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8334112986100994e-07,
+      "loss": 0.264,
+      "step": 7341
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8255072533591488e-07,
+      "loss": 0.249,
+      "step": 7342
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8176201256071556e-07,
+      "loss": 0.2646,
+      "step": 7343
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8097499167132436e-07,
+      "loss": 0.2561,
+      "step": 7344
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8018966280336058e-07,
+      "loss": 0.2519,
+      "step": 7345
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.794060260921504e-07,
+      "loss": 0.2356,
+      "step": 7346
+    },
+    {
+      "epoch": 8.47,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7862408167273472e-07,
+      "loss": 0.261,
+      "step": 7347
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.778438296798557e-07,
+      "loss": 0.2509,
+      "step": 7348
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7706527024796805e-07,
+      "loss": 0.299,
+      "step": 7349
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7628840351123444e-07,
+      "loss": 0.2672,
+      "step": 7350
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7551322960352334e-07,
+      "loss": 0.2501,
+      "step": 7351
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.747397486584157e-07,
+      "loss": 0.2678,
+      "step": 7352
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7396796080919599e-07,
+      "loss": 0.2631,
+      "step": 7353
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7319786618885892e-07,
+      "loss": 0.3035,
+      "step": 7354
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.724294649301095e-07,
+      "loss": 0.2955,
+      "step": 7355
+    },
+    {
+      "epoch": 8.48,
+      "grad_norm": 0.0,
+      "learning_rate": 1.716627571653595e-07,
+      "loss": 0.2371,
+      "step": 7356
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.708977430267278e-07,
+      "loss": 0.2488,
+      "step": 7357
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7013442264604108e-07,
+      "loss": 0.2483,
+      "step": 7358
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6937279615483748e-07,
+      "loss": 0.2327,
+      "step": 7359
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6861286368435758e-07,
+      "loss": 0.253,
+      "step": 7360
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6785462536555775e-07,
+      "loss": 0.2675,
+      "step": 7361
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6709808132909454e-07,
+      "loss": 0.2229,
+      "step": 7362
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6634323170533928e-07,
+      "loss": 0.2582,
+      "step": 7363
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6559007662436567e-07,
+      "loss": 0.265,
+      "step": 7364
+    },
+    {
+      "epoch": 8.49,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6483861621595875e-07,
+      "loss": 0.252,
+      "step": 7365
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6408885060961056e-07,
+      "loss": 0.2572,
+      "step": 7366
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6334077993452324e-07,
+      "loss": 0.2903,
+      "step": 7367
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.625944043196015e-07,
+      "loss": 0.2181,
+      "step": 7368
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6184972389346355e-07,
+      "loss": 0.3192,
+      "step": 7369
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.611067387844334e-07,
+      "loss": 0.246,
+      "step": 7370
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6036544912054087e-07,
+      "loss": 0.2338,
+      "step": 7371
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.596258550295271e-07,
+      "loss": 0.2527,
+      "step": 7372
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5888795663883904e-07,
+      "loss": 0.242,
+      "step": 7373
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5815175407563166e-07,
+      "loss": 0.2213,
+      "step": 7374
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.57417247466769e-07,
+      "loss": 0.2663,
+      "step": 7375
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5668443693882095e-07,
+      "loss": 0.2315,
+      "step": 7376
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.559533226180654e-07,
+      "loss": 0.2453,
+      "step": 7377
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.552239046304893e-07,
+      "loss": 0.2815,
+      "step": 7378
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.544961831017855e-07,
+      "loss": 0.2505,
+      "step": 7379
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5377015815735696e-07,
+      "loss": 0.2833,
+      "step": 7380
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5304582992231255e-07,
+      "loss": 0.3035,
+      "step": 7381
+    },
+    {
+      "epoch": 8.51,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5232319852146682e-07,
+      "loss": 0.2441,
+      "step": 7382
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5160226407934797e-07,
+      "loss": 0.2748,
+      "step": 7383
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.508830267201844e-07,
+      "loss": 0.2477,
+      "step": 7384
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5016548656791697e-07,
+      "loss": 0.2337,
+      "step": 7385
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4944964374619342e-07,
+      "loss": 0.2803,
+      "step": 7386
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.487354983783673e-07,
+      "loss": 0.2291,
+      "step": 7387
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4802305058750134e-07,
+      "loss": 0.3303,
+      "step": 7388
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4731230049636503e-07,
+      "loss": 0.2421,
+      "step": 7389
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4660324822743378e-07,
+      "loss": 0.2097,
+      "step": 7390
+    },
+    {
+      "epoch": 8.52,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4589589390289427e-07,
+      "loss": 0.2418,
+      "step": 7391
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.451902376446368e-07,
+      "loss": 0.2634,
+      "step": 7392
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4448627957426186e-07,
+      "loss": 0.2656,
+      "step": 7393
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4378401981307466e-07,
+      "loss": 0.2334,
+      "step": 7394
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.430834584820895e-07,
+      "loss": 0.2364,
+      "step": 7395
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4238459570202644e-07,
+      "loss": 0.2765,
+      "step": 7396
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.41687431593317e-07,
+      "loss": 0.3027,
+      "step": 7397
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4099196627609503e-07,
+      "loss": 0.2607,
+      "step": 7398
+    },
+    {
+      "epoch": 8.53,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4029819987020466e-07,
+      "loss": 0.2436,
+      "step": 7399
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3960613249519472e-07,
+      "loss": 0.2497,
+      "step": 7400
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3891576427032317e-07,
+      "loss": 0.2331,
+      "step": 7401
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.382270953145548e-07,
+      "loss": 0.2651,
+      "step": 7402
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.375401257465625e-07,
+      "loss": 0.2479,
+      "step": 7403
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.368548556847238e-07,
+      "loss": 0.2896,
+      "step": 7404
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3617128524712643e-07,
+      "loss": 0.3011,
+      "step": 7405
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3548941455156395e-07,
+      "loss": 0.2641,
+      "step": 7406
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.348092437155346e-07,
+      "loss": 0.2464,
+      "step": 7407
+    },
+    {
+      "epoch": 8.54,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3413077285624797e-07,
+      "loss": 0.2357,
+      "step": 7408
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.334540020906172e-07,
+      "loss": 0.2864,
+      "step": 7409
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3277893153526568e-07,
+      "loss": 0.2417,
+      "step": 7410
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3210556130652031e-07,
+      "loss": 0.2699,
+      "step": 7411
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.314338915204172e-07,
+      "loss": 0.2535,
+      "step": 7412
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3076392229270041e-07,
+      "loss": 0.2752,
+      "step": 7413
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.300956537388176e-07,
+      "loss": 0.2815,
+      "step": 7414
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2942908597392555e-07,
+      "loss": 0.2519,
+      "step": 7415
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2876421911288906e-07,
+      "loss": 0.296,
+      "step": 7416
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2810105327027754e-07,
+      "loss": 0.2644,
+      "step": 7417
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2743958856036743e-07,
+      "loss": 0.2785,
+      "step": 7418
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2677982509714415e-07,
+      "loss": 0.2861,
+      "step": 7419
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.261217629942968e-07,
+      "loss": 0.2484,
+      "step": 7420
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2546540236522575e-07,
+      "loss": 0.2687,
+      "step": 7421
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.248107433230339e-07,
+      "loss": 0.2896,
+      "step": 7422
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2415778598053207e-07,
+      "loss": 0.2739,
+      "step": 7423
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.235065304502403e-07,
+      "loss": 0.2342,
+      "step": 7424
+    },
+    {
+      "epoch": 8.56,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2285697684438104e-07,
+      "loss": 0.2736,
+      "step": 7425
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2220912527488804e-07,
+      "loss": 0.2641,
+      "step": 7426
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2156297585339872e-07,
+      "loss": 0.2533,
+      "step": 7427
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.209185286912573e-07,
+      "loss": 0.3595,
+      "step": 7428
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.20275783899515e-07,
+      "loss": 0.2366,
+      "step": 7429
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1963474158893207e-07,
+      "loss": 0.2286,
+      "step": 7430
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1899540186997127e-07,
+      "loss": 0.2143,
+      "step": 7431
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.183577648528056e-07,
+      "loss": 0.2568,
+      "step": 7432
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1772183064731268e-07,
+      "loss": 0.2992,
+      "step": 7433
+    },
+    {
+      "epoch": 8.57,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1708759936307601e-07,
+      "loss": 0.2235,
+      "step": 7434
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1645507110938925e-07,
+      "loss": 0.2383,
+      "step": 7435
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1582424599524855e-07,
+      "loss": 0.2484,
+      "step": 7436
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1519512412935807e-07,
+      "loss": 0.2508,
+      "step": 7437
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1456770562012776e-07,
+      "loss": 0.2461,
+      "step": 7438
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1394199057567667e-07,
+      "loss": 0.2515,
+      "step": 7439
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1331797910382747e-07,
+      "loss": 0.2765,
+      "step": 7440
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1269567131211079e-07,
+      "loss": 0.2974,
+      "step": 7441
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1207506730776308e-07,
+      "loss": 0.2686,
+      "step": 7442
+    },
+    {
+      "epoch": 8.58,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1145616719772545e-07,
+      "loss": 0.2642,
+      "step": 7443
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1083897108864927e-07,
+      "loss": 0.2447,
+      "step": 7444
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1022347908689057e-07,
+      "loss": 0.2588,
+      "step": 7445
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0960969129851007e-07,
+      "loss": 0.2297,
+      "step": 7446
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0899760782927649e-07,
+      "loss": 0.2567,
+      "step": 7447
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0838722878466435e-07,
+      "loss": 0.2474,
+      "step": 7448
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0777855426985395e-07,
+      "loss": 0.2713,
+      "step": 7449
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0717158438973474e-07,
+      "loss": 0.2267,
+      "step": 7450
+    },
+    {
+      "epoch": 8.59,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0656631924889749e-07,
+      "loss": 0.2378,
+      "step": 7451
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0596275895164431e-07,
+      "loss": 0.2987,
+      "step": 7452
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0536090360198092e-07,
+      "loss": 0.2164,
+      "step": 7453
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0476075330361656e-07,
+      "loss": 0.2775,
+      "step": 7454
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0416230815997297e-07,
+      "loss": 0.2401,
+      "step": 7455
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0356556827417208e-07,
+      "loss": 0.2877,
+      "step": 7456
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.029705337490472e-07,
+      "loss": 0.2863,
+      "step": 7457
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0237720468713297e-07,
+      "loss": 0.2518,
+      "step": 7458
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.2726,
+      "step": 7459
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0119566336161624e-07,
+      "loss": 0.2802,
+      "step": 7460
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0060745130161753e-07,
+      "loss": 0.2848,
+      "step": 7461
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0002094511203819e-07,
+      "loss": 0.2352,
+      "step": 7462
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 0.0,
+      "learning_rate": 9.943614489394626e-08,
+      "loss": 0.2337,
+      "step": 7463
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 0.0,
+      "learning_rate": 9.885305074811336e-08,
+      "loss": 0.2809,
+      "step": 7464
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 0.0,
+      "learning_rate": 9.827166277501909e-08,
+      "loss": 0.2332,
+      "step": 7465
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 0.0,
+      "learning_rate": 9.769198107485001e-08,
+      "loss": 0.3003,
+      "step": 7466
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 0.0,
+      "learning_rate": 9.711400574749507e-08,
+      "loss": 0.2908,
+      "step": 7467
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 0.0,
+      "learning_rate": 9.653773689255353e-08,
+      "loss": 0.2607,
+      "step": 7468
+    },
+    {
+      "epoch": 8.61,
+      "grad_norm": 0.0,
+      "learning_rate": 9.596317460932703e-08,
+      "loss": 0.2768,
+      "step": 7469
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 0.0,
+      "learning_rate": 9.539031899682527e-08,
+      "loss": 0.2469,
+      "step": 7470
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 0.0,
+      "learning_rate": 9.481917015376262e-08,
+      "loss": 0.2361,
+      "step": 7471
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 0.0,
+      "learning_rate": 9.424972817856037e-08,
+      "loss": 0.2554,
+      "step": 7472
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 0.0,
+      "learning_rate": 9.368199316934446e-08,
+      "loss": 0.2717,
+      "step": 7473
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 0.0,
+      "learning_rate": 9.311596522394772e-08,
+      "loss": 0.3023,
+      "step": 7474
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 0.0,
+      "learning_rate": 9.255164443990994e-08,
+      "loss": 0.3011,
+      "step": 7475
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 0.0,
+      "learning_rate": 9.19890309144722e-08,
+      "loss": 0.264,
+      "step": 7476
+    },
+    {
+      "epoch": 8.62,
+      "grad_norm": 0.0,
+      "learning_rate": 9.142812474458806e-08,
+      "loss": 0.2219,
+      "step": 7477
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 0.0,
+      "learning_rate": 9.086892602691022e-08,
+      "loss": 0.2331,
+      "step": 7478
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 0.0,
+      "learning_rate": 9.031143485780159e-08,
+      "loss": 0.2359,
+      "step": 7479
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 0.0,
+      "learning_rate": 8.975565133332976e-08,
+      "loss": 0.2421,
+      "step": 7480
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 0.0,
+      "learning_rate": 8.92015755492659e-08,
+      "loss": 0.2498,
+      "step": 7481
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 0.0,
+      "learning_rate": 8.864920760109252e-08,
+      "loss": 0.2292,
+      "step": 7482
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 0.0,
+      "learning_rate": 8.809854758399017e-08,
+      "loss": 0.2386,
+      "step": 7483
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 0.0,
+      "learning_rate": 8.75495955928507e-08,
+      "loss": 0.2684,
+      "step": 7484
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 0.0,
+      "learning_rate": 8.700235172226845e-08,
+      "loss": 0.2264,
+      "step": 7485
+    },
+    {
+      "epoch": 8.63,
+      "grad_norm": 0.0,
+      "learning_rate": 8.645681606654799e-08,
+      "loss": 0.2681,
+      "step": 7486
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.0,
+      "learning_rate": 8.591298871969411e-08,
+      "loss": 0.2674,
+      "step": 7487
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.0,
+      "learning_rate": 8.537086977541964e-08,
+      "loss": 0.2508,
+      "step": 7488
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.0,
+      "learning_rate": 8.483045932714318e-08,
+      "loss": 0.2442,
+      "step": 7489
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.0,
+      "learning_rate": 8.429175746798912e-08,
+      "loss": 0.2358,
+      "step": 7490
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.0,
+      "learning_rate": 8.375476429078543e-08,
+      "loss": 0.2529,
+      "step": 7491
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.0,
+      "learning_rate": 8.321947988806811e-08,
+      "loss": 0.2656,
+      "step": 7492
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.0,
+      "learning_rate": 8.268590435208002e-08,
+      "loss": 0.2572,
+      "step": 7493
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.0,
+      "learning_rate": 8.215403777476317e-08,
+      "loss": 0.2683,
+      "step": 7494
+    },
+    {
+      "epoch": 8.64,
+      "grad_norm": 0.0,
+      "learning_rate": 8.162388024777202e-08,
+      "loss": 0.2596,
+      "step": 7495
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 0.0,
+      "learning_rate": 8.109543186246349e-08,
+      "loss": 0.2389,
+      "step": 7496
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 0.0,
+      "learning_rate": 8.05686927098981e-08,
+      "loss": 0.2243,
+      "step": 7497
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 0.0,
+      "learning_rate": 8.004366288084542e-08,
+      "loss": 0.2892,
+      "step": 7498
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 0.0,
+      "learning_rate": 7.952034246577977e-08,
+      "loss": 0.2625,
+      "step": 7499
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 0.0,
+      "learning_rate": 7.899873155487902e-08,
+      "loss": 0.2371,
+      "step": 7500
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 0.0,
+      "learning_rate": 7.847883023802793e-08,
+      "loss": 0.2834,
+      "step": 7501
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 0.0,
+      "learning_rate": 7.796063860481595e-08,
+      "loss": 0.2367,
+      "step": 7502
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 0.0,
+      "learning_rate": 7.744415674453721e-08,
+      "loss": 0.248,
+      "step": 7503
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 0.0,
+      "learning_rate": 7.692938474619494e-08,
+      "loss": 0.2731,
+      "step": 7504
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 0.0,
+      "learning_rate": 7.641632269849263e-08,
+      "loss": 0.2618,
+      "step": 7505
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 0.0,
+      "learning_rate": 7.590497068984293e-08,
+      "loss": 0.2684,
+      "step": 7506
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 0.0,
+      "learning_rate": 7.539532880836087e-08,
+      "loss": 0.2599,
+      "step": 7507
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 0.0,
+      "learning_rate": 7.488739714187066e-08,
+      "loss": 0.2822,
+      "step": 7508
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 0.0,
+      "learning_rate": 7.438117577789561e-08,
+      "loss": 0.2796,
+      "step": 7509
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 0.0,
+      "learning_rate": 7.387666480367262e-08,
+      "loss": 0.2685,
+      "step": 7510
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 0.0,
+      "learning_rate": 7.33738643061377e-08,
+      "loss": 0.2693,
+      "step": 7511
+    },
+    {
+      "epoch": 8.66,
+      "grad_norm": 0.0,
+      "learning_rate": 7.287277437193153e-08,
+      "loss": 0.2309,
+      "step": 7512
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.0,
+      "learning_rate": 7.237339508740615e-08,
+      "loss": 0.2691,
+      "step": 7513
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.0,
+      "learning_rate": 7.187572653861163e-08,
+      "loss": 0.2642,
+      "step": 7514
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.0,
+      "learning_rate": 7.137976881130826e-08,
+      "loss": 0.2164,
+      "step": 7515
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.0,
+      "learning_rate": 7.088552199095877e-08,
+      "loss": 0.2695,
+      "step": 7516
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.0,
+      "learning_rate": 7.039298616273393e-08,
+      "loss": 0.244,
+      "step": 7517
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.990216141150696e-08,
+      "loss": 0.2695,
+      "step": 7518
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.941304782185577e-08,
+      "loss": 0.3221,
+      "step": 7519
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.892564547806624e-08,
+      "loss": 0.2375,
+      "step": 7520
+    },
+    {
+      "epoch": 8.67,
+      "grad_norm": 0.0,
+      "learning_rate": 6.843995446412788e-08,
+      "loss": 0.2168,
+      "step": 7521
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.795597486373595e-08,
+      "loss": 0.2316,
+      "step": 7522
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.747370676028819e-08,
+      "loss": 0.2285,
+      "step": 7523
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.699315023689035e-08,
+      "loss": 0.243,
+      "step": 7524
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.651430537635284e-08,
+      "loss": 0.2281,
+      "step": 7525
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.603717226118966e-08,
+      "loss": 0.2372,
+      "step": 7526
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.556175097362061e-08,
+      "loss": 0.2619,
+      "step": 7527
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.508804159557235e-08,
+      "loss": 0.2832,
+      "step": 7528
+    },
+    {
+      "epoch": 8.68,
+      "grad_norm": 0.0,
+      "learning_rate": 6.461604420867296e-08,
+      "loss": 0.2614,
+      "step": 7529
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.414575889425844e-08,
+      "loss": 0.2264,
+      "step": 7530
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.367718573336845e-08,
+      "loss": 0.2893,
+      "step": 7531
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.321032480674727e-08,
+      "loss": 0.2339,
+      "step": 7532
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.274517619484499e-08,
+      "loss": 0.2372,
+      "step": 7533
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.228173997781639e-08,
+      "loss": 0.2669,
+      "step": 7534
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.182001623552202e-08,
+      "loss": 0.2591,
+      "step": 7535
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.136000504752493e-08,
+      "loss": 0.2245,
+      "step": 7536
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.090170649309502e-08,
+      "loss": 0.2643,
+      "step": 7537
+    },
+    {
+      "epoch": 8.69,
+      "grad_norm": 0.0,
+      "learning_rate": 6.044512065120689e-08,
+      "loss": 0.2757,
+      "step": 7538
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.0,
+      "learning_rate": 5.999024760054095e-08,
+      "loss": 0.2731,
+      "step": 7539
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.0,
+      "learning_rate": 5.953708741947894e-08,
+      "loss": 0.258,
+      "step": 7540
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.0,
+      "learning_rate": 5.9085640186111735e-08,
+      "loss": 0.251,
+      "step": 7541
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8635905978231544e-08,
+      "loss": 0.2882,
+      "step": 7542
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8187884873338595e-08,
+      "loss": 0.2475,
+      "step": 7543
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.0,
+      "learning_rate": 5.774157694863336e-08,
+      "loss": 0.2449,
+      "step": 7544
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7296982281026534e-08,
+      "loss": 0.2982,
+      "step": 7545
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.0,
+      "learning_rate": 5.685410094713017e-08,
+      "loss": 0.2315,
+      "step": 7546
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 0.0,
+      "learning_rate": 5.641293302326323e-08,
+      "loss": 0.2687,
+      "step": 7547
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 0.0,
+      "learning_rate": 5.597347858544488e-08,
+      "loss": 0.2562,
+      "step": 7548
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 0.0,
+      "learning_rate": 5.5535737709404566e-08,
+      "loss": 0.3346,
+      "step": 7549
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 0.0,
+      "learning_rate": 5.509971047057416e-08,
+      "loss": 0.2672,
+      "step": 7550
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 0.0,
+      "learning_rate": 5.466539694408912e-08,
+      "loss": 0.2765,
+      "step": 7551
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 0.0,
+      "learning_rate": 5.42327972047918e-08,
+      "loss": 0.2268,
+      "step": 7552
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 0.0,
+      "learning_rate": 5.380191132722701e-08,
+      "loss": 0.2281,
+      "step": 7553
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 0.0,
+      "learning_rate": 5.3372739385645356e-08,
+      "loss": 0.296,
+      "step": 7554
+    },
+    {
+      "epoch": 8.71,
+      "grad_norm": 0.0,
+      "learning_rate": 5.2945281454003236e-08,
+      "loss": 0.2407,
+      "step": 7555
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.0,
+      "learning_rate": 5.2519537605959513e-08,
+      "loss": 0.2791,
+      "step": 7556
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.0,
+      "learning_rate": 5.209550791487883e-08,
+      "loss": 0.2595,
+      "step": 7557
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.0,
+      "learning_rate": 5.167319245383162e-08,
+      "loss": 0.2658,
+      "step": 7558
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.0,
+      "learning_rate": 5.125259129558857e-08,
+      "loss": 0.2888,
+      "step": 7559
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.0,
+      "learning_rate": 5.083370451263059e-08,
+      "loss": 0.2291,
+      "step": 7560
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.0,
+      "learning_rate": 5.041653217713993e-08,
+      "loss": 0.3176,
+      "step": 7561
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0001074361003546e-08,
+      "loss": 0.2657,
+      "step": 7562
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.0,
+      "learning_rate": 4.958733113581415e-08,
+      "loss": 0.2462,
+      "step": 7563
+    },
+    {
+      "epoch": 8.72,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9175302572866956e-08,
+      "loss": 0.2583,
+      "step": 7564
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 0.0,
+      "learning_rate": 4.876498874316515e-08,
+      "loss": 0.2407,
+      "step": 7565
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 0.0,
+      "learning_rate": 4.835638971741219e-08,
+      "loss": 0.2258,
+      "step": 7566
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7949505566018405e-08,
+      "loss": 0.265,
+      "step": 7567
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 0.0,
+      "learning_rate": 4.754433635909994e-08,
+      "loss": 0.3081,
+      "step": 7568
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 0.0,
+      "learning_rate": 4.714088216647428e-08,
+      "loss": 0.2858,
+      "step": 7569
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6739143057665806e-08,
+      "loss": 0.2165,
+      "step": 7570
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6339119101902475e-08,
+      "loss": 0.273,
+      "step": 7571
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5940810368116925e-08,
+      "loss": 0.2543,
+      "step": 7572
+    },
+    {
+      "epoch": 8.73,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5544216924944265e-08,
+      "loss": 0.248,
+      "step": 7573
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5149338840727586e-08,
+      "loss": 0.2766,
+      "step": 7574
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 0.0,
+      "learning_rate": 4.475617618351136e-08,
+      "loss": 0.2904,
+      "step": 7575
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4364729021048044e-08,
+      "loss": 0.3101,
+      "step": 7576
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3974997420789257e-08,
+      "loss": 0.219,
+      "step": 7577
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 0.0,
+      "learning_rate": 4.35869814498946e-08,
+      "loss": 0.2492,
+      "step": 7578
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 0.0,
+      "learning_rate": 4.320068117522835e-08,
+      "loss": 0.277,
+      "step": 7579
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 0.0,
+      "learning_rate": 4.281609666335618e-08,
+      "loss": 0.2551,
+      "step": 7580
+    },
+    {
+      "epoch": 8.74,
+      "grad_norm": 0.0,
+      "learning_rate": 4.243322798055172e-08,
+      "loss": 0.2552,
+      "step": 7581
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.0,
+      "learning_rate": 4.205207519278998e-08,
+      "loss": 0.248,
+      "step": 7582
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.0,
+      "learning_rate": 4.167263836575286e-08,
+      "loss": 0.2308,
+      "step": 7583
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.0,
+      "learning_rate": 4.129491756482362e-08,
+      "loss": 0.2353,
+      "step": 7584
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.0,
+      "learning_rate": 4.091891285509353e-08,
+      "loss": 0.2802,
+      "step": 7585
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.0,
+      "learning_rate": 4.054462430135298e-08,
+      "loss": 0.2367,
+      "step": 7586
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0172051968101474e-08,
+      "loss": 0.2295,
+      "step": 7587
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.980119591954101e-08,
+      "loss": 0.2395,
+      "step": 7588
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.943205621957713e-08,
+      "loss": 0.2523,
+      "step": 7589
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9064632931821165e-08,
+      "loss": 0.287,
+      "step": 7590
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.869892611958692e-08,
+      "loss": 0.2212,
+      "step": 7591
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.833493584589398e-08,
+      "loss": 0.2697,
+      "step": 7592
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.797266217346329e-08,
+      "loss": 0.3112,
+      "step": 7593
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.761210516472602e-08,
+      "loss": 0.318,
+      "step": 7594
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7253264881809137e-08,
+      "loss": 0.2588,
+      "step": 7595
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6896141386552065e-08,
+      "loss": 0.2466,
+      "step": 7596
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.654073474049225e-08,
+      "loss": 0.2514,
+      "step": 7597
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.618704500487402e-08,
+      "loss": 0.2664,
+      "step": 7598
+    },
+    {
+      "epoch": 8.76,
+      "grad_norm": 0.0,
+      "learning_rate": 3.583507224064642e-08,
+      "loss": 0.2746,
+      "step": 7599
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.548481650846092e-08,
+      "loss": 0.2788,
+      "step": 7600
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5136277868674795e-08,
+      "loss": 0.2161,
+      "step": 7601
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4789456381346675e-08,
+      "loss": 0.2357,
+      "step": 7602
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4444352106242086e-08,
+      "loss": 0.2449,
+      "step": 7603
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4100965102829007e-08,
+      "loss": 0.2598,
+      "step": 7604
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3759295430281226e-08,
+      "loss": 0.2572,
+      "step": 7605
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.341934314747497e-08,
+      "loss": 0.2976,
+      "step": 7606
+    },
+    {
+      "epoch": 8.77,
+      "grad_norm": 0.0,
+      "learning_rate": 3.308110831299116e-08,
+      "loss": 0.262,
+      "step": 7607
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.274459098511429e-08,
+      "loss": 0.2576,
+      "step": 7608
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.240979122183352e-08,
+      "loss": 0.2553,
+      "step": 7609
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.207670908084159e-08,
+      "loss": 0.254,
+      "step": 7610
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.174534461953593e-08,
+      "loss": 0.2562,
+      "step": 7611
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.141569789501753e-08,
+      "loss": 0.2968,
+      "step": 7612
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1087768964092046e-08,
+      "loss": 0.2545,
+      "step": 7613
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0761557883265406e-08,
+      "loss": 0.2303,
+      "step": 7614
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.043706470875374e-08,
+      "loss": 0.258,
+      "step": 7615
+    },
+    {
+      "epoch": 8.78,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0114289496472325e-08,
+      "loss": 0.2466,
+      "step": 7616
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9793232302043342e-08,
+      "loss": 0.2386,
+      "step": 7617
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9473893180790304e-08,
+      "loss": 0.2269,
+      "step": 7618
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.915627218774142e-08,
+      "loss": 0.2421,
+      "step": 7619
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8840369377631793e-08,
+      "loss": 0.2185,
+      "step": 7620
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.852618480489566e-08,
+      "loss": 0.2462,
+      "step": 7621
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8213718523674162e-08,
+      "loss": 0.2382,
+      "step": 7622
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7902970587813104e-08,
+      "loss": 0.2719,
+      "step": 7623
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.759394105085855e-08,
+      "loss": 0.2627,
+      "step": 7624
+    },
+    {
+      "epoch": 8.79,
+      "grad_norm": 0.0,
+      "learning_rate": 2.728662996606457e-08,
+      "loss": 0.2477,
+      "step": 7625
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6981037386385468e-08,
+      "loss": 0.2993,
+      "step": 7626
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.667716336448356e-08,
+      "loss": 0.2283,
+      "step": 7627
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.63750079527203e-08,
+      "loss": 0.2392,
+      "step": 7628
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6074571203165143e-08,
+      "loss": 0.2411,
+      "step": 7629
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5775853167587793e-08,
+      "loss": 0.291,
+      "step": 7630
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.547885389746485e-08,
+      "loss": 0.2314,
+      "step": 7631
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5183573443975372e-08,
+      "loss": 0.2539,
+      "step": 7632
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4890011858000886e-08,
+      "loss": 0.2206,
+      "step": 7633
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4598169190129807e-08,
+      "loss": 0.3009,
+      "step": 7634
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.430804549065302e-08,
+      "loss": 0.2533,
+      "step": 7635
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.401964080956276e-08,
+      "loss": 0.251,
+      "step": 7636
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.373295519655927e-08,
+      "loss": 0.2976,
+      "step": 7637
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3447988701043036e-08,
+      "loss": 0.2904,
+      "step": 7638
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3164741372119215e-08,
+      "loss": 0.2152,
+      "step": 7639
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2883213258598767e-08,
+      "loss": 0.2255,
+      "step": 7640
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2603404408994e-08,
+      "loss": 0.233,
+      "step": 7641
+    },
+    {
+      "epoch": 8.81,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2325314871521898e-08,
+      "loss": 0.2509,
+      "step": 7642
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2048944694104123e-08,
+      "loss": 0.2555,
+      "step": 7643
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.177429392436259e-08,
+      "loss": 0.2915,
+      "step": 7644
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.150136260962721e-08,
+      "loss": 0.2548,
+      "step": 7645
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1230150796929248e-08,
+      "loss": 0.2528,
+      "step": 7646
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0960658533003554e-08,
+      "loss": 0.2392,
+      "step": 7647
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0692885864289637e-08,
+      "loss": 0.2571,
+      "step": 7648
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0426832836930588e-08,
+      "loss": 0.2815,
+      "step": 7649
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0162499496771958e-08,
+      "loss": 0.2473,
+      "step": 7650
+    },
+    {
+      "epoch": 8.82,
+      "grad_norm": 0.0,
+      "learning_rate": 1.989988588936509e-08,
+      "loss": 0.2536,
+      "step": 7651
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.963899205996378e-08,
+      "loss": 0.2466,
+      "step": 7652
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9379818053523182e-08,
+      "loss": 0.2529,
+      "step": 7653
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9122363914707565e-08,
+      "loss": 0.2283,
+      "step": 7654
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8866629687879222e-08,
+      "loss": 0.2625,
+      "step": 7655
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8612615417107348e-08,
+      "loss": 0.2579,
+      "step": 7656
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8360321146163595e-08,
+      "loss": 0.2264,
+      "step": 7657
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8109746918524295e-08,
+      "loss": 0.2629,
+      "step": 7658
+    },
+    {
+      "epoch": 8.83,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7860892777367133e-08,
+      "loss": 0.2601,
+      "step": 7659
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7613758765576694e-08,
+      "loss": 0.2481,
+      "step": 7660
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7368344925736692e-08,
+      "loss": 0.2772,
+      "step": 7661
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.712465130013996e-08,
+      "loss": 0.2718,
+      "step": 7662
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6882677930777357e-08,
+      "loss": 0.2588,
+      "step": 7663
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.664242485934775e-08,
+      "loss": 0.2727,
+      "step": 7664
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6403892127251352e-08,
+      "loss": 0.2546,
+      "step": 7665
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6167079775591956e-08,
+      "loss": 0.2358,
+      "step": 7666
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5931987845176912e-08,
+      "loss": 0.3131,
+      "step": 7667
+    },
+    {
+      "epoch": 8.84,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5698616376518262e-08,
+      "loss": 0.2631,
+      "step": 7668
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.546696540982939e-08,
+      "loss": 0.2668,
+      "step": 7669
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5237034985030596e-08,
+      "loss": 0.2838,
+      "step": 7670
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.50088251417424e-08,
+      "loss": 0.2152,
+      "step": 7671
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4782335919288903e-08,
+      "loss": 0.2981,
+      "step": 7672
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4557567356702218e-08,
+      "loss": 0.2617,
+      "step": 7673
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4334519492711362e-08,
+      "loss": 0.2785,
+      "step": 7674
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.411319236575337e-08,
+      "loss": 0.2543,
+      "step": 7675
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3893586013967731e-08,
+      "loss": 0.2538,
+      "step": 7676
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3675700475196395e-08,
+      "loss": 0.2751,
+      "step": 7677
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3459535786987111e-08,
+      "loss": 0.244,
+      "step": 7678
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3245091986587855e-08,
+      "loss": 0.2393,
+      "step": 7679
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3032369110952403e-08,
+      "loss": 0.2789,
+      "step": 7680
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2821367196736988e-08,
+      "loss": 0.2522,
+      "step": 7681
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2612086280302527e-08,
+      "loss": 0.292,
+      "step": 7682
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2404526397711281e-08,
+      "loss": 0.2376,
+      "step": 7683
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2198687584730195e-08,
+      "loss": 0.3294,
+      "step": 7684
+    },
+    {
+      "epoch": 8.86,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1994569876830897e-08,
+      "loss": 0.2402,
+      "step": 7685
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1792173309186361e-08,
+      "loss": 0.2538,
+      "step": 7686
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1591497916673133e-08,
+      "loss": 0.2512,
+      "step": 7687
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.139254373387244e-08,
+      "loss": 0.2571,
+      "step": 7688
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.119531079506797e-08,
+      "loss": 0.2123,
+      "step": 7689
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.099979913424698e-08,
+      "loss": 0.2867,
+      "step": 7690
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0806008785100297e-08,
+      "loss": 0.2958,
+      "step": 7691
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.061393978102121e-08,
+      "loss": 0.2257,
+      "step": 7692
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0423592155108798e-08,
+      "loss": 0.2182,
+      "step": 7693
+    },
+    {
+      "epoch": 8.87,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0234965940162378e-08,
+      "loss": 0.2335,
+      "step": 7694
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0048061168688173e-08,
+      "loss": 0.2565,
+      "step": 7695
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.862877872891529e-09,
+      "loss": 0.2445,
+      "step": 7696
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.67941608468359e-09,
+      "loss": 0.2552,
+      "step": 7697
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.497675835680175e-09,
+      "loss": 0.2406,
+      "step": 7698
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.317657157197347e-09,
+      "loss": 0.2523,
+      "step": 7699
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.0,
+      "learning_rate": 9.139360080256953e-09,
+      "loss": 0.2356,
+      "step": 7700
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.0,
+      "learning_rate": 8.96278463558331e-09,
+      "loss": 0.2447,
+      "step": 7701
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.0,
+      "learning_rate": 8.787930853603188e-09,
+      "loss": 0.2442,
+      "step": 7702
+    },
+    {
+      "epoch": 8.88,
+      "grad_norm": 0.0,
+      "learning_rate": 8.61479876444804e-09,
+      "loss": 0.2118,
+      "step": 7703
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 0.0,
+      "learning_rate": 8.44338839795289e-09,
+      "loss": 0.241,
+      "step": 7704
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 0.0,
+      "learning_rate": 8.273699783652999e-09,
+      "loss": 0.2835,
+      "step": 7705
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 0.0,
+      "learning_rate": 8.105732950791645e-09,
+      "loss": 0.3117,
+      "step": 7706
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.93948792831234e-09,
+      "loss": 0.2236,
+      "step": 7707
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.774964744862167e-09,
+      "loss": 0.2439,
+      "step": 7708
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.612163428791786e-09,
+      "loss": 0.26,
+      "step": 7709
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.451084008155419e-09,
+      "loss": 0.2637,
+      "step": 7710
+    },
+    {
+      "epoch": 8.89,
+      "grad_norm": 0.0,
+      "learning_rate": 7.291726510710861e-09,
+      "loss": 0.2352,
+      "step": 7711
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 0.0,
+      "learning_rate": 7.134090963918372e-09,
+      "loss": 0.2306,
+      "step": 7712
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 0.0,
+      "learning_rate": 6.978177394940666e-09,
+      "loss": 0.2608,
+      "step": 7713
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 0.0,
+      "learning_rate": 6.823985830647362e-09,
+      "loss": 0.2519,
+      "step": 7714
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 0.0,
+      "learning_rate": 6.671516297606095e-09,
+      "loss": 0.2239,
+      "step": 7715
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 0.0,
+      "learning_rate": 6.520768822092516e-09,
+      "loss": 0.2302,
+      "step": 7716
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 0.0,
+      "learning_rate": 6.371743430082511e-09,
+      "loss": 0.2455,
+      "step": 7717
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 0.0,
+      "learning_rate": 6.2244401472577595e-09,
+      "loss": 0.2467,
+      "step": 7718
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 0.0,
+      "learning_rate": 6.078858998999071e-09,
+      "loss": 0.2661,
+      "step": 7719
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 0.0,
+      "learning_rate": 5.935000010396375e-09,
+      "loss": 0.2679,
+      "step": 7720
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.79286320623651e-09,
+      "loss": 0.2504,
+      "step": 7721
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.652448611013217e-09,
+      "loss": 0.2642,
+      "step": 7722
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.513756248924917e-09,
+      "loss": 0.268,
+      "step": 7723
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.376786143869162e-09,
+      "loss": 0.2635,
+      "step": 7724
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.241538319448181e-09,
+      "loss": 0.2614,
+      "step": 7725
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 0.0,
+      "learning_rate": 5.1080127989699964e-09,
+      "loss": 0.2867,
+      "step": 7726
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9762096054428724e-09,
+      "loss": 0.2347,
+      "step": 7727
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 0.0,
+      "learning_rate": 4.846128761579749e-09,
+      "loss": 0.251,
+      "step": 7728
+    },
+    {
+      "epoch": 8.91,
+      "grad_norm": 0.0,
+      "learning_rate": 4.717770289794921e-09,
+      "loss": 0.3454,
+      "step": 7729
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5911342122084705e-09,
+      "loss": 0.2278,
+      "step": 7730
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 0.0,
+      "learning_rate": 4.466220550641831e-09,
+      "loss": 0.2815,
+      "step": 7731
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 0.0,
+      "learning_rate": 4.343029326621118e-09,
+      "loss": 0.2445,
+      "step": 7732
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 0.0,
+      "learning_rate": 4.221560561372684e-09,
+      "loss": 0.2263,
+      "step": 7733
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1018142758308955e-09,
+      "loss": 0.2613,
+      "step": 7734
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 0.0,
+      "learning_rate": 3.983790490628137e-09,
+      "loss": 0.2606,
+      "step": 7735
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 0.0,
+      "learning_rate": 3.867489226102583e-09,
+      "loss": 0.2351,
+      "step": 7736
+    },
+    {
+      "epoch": 8.92,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7529105022970915e-09,
+      "loss": 0.2331,
+      "step": 7737
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 0.0,
+      "learning_rate": 3.640054338954757e-09,
+      "loss": 0.2585,
+      "step": 7738
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 0.0,
+      "learning_rate": 3.528920755523357e-09,
+      "loss": 0.2396,
+      "step": 7739
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 0.0,
+      "learning_rate": 3.419509771153129e-09,
+      "loss": 0.2477,
+      "step": 7740
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 0.0,
+      "learning_rate": 3.311821404697879e-09,
+      "loss": 0.2738,
+      "step": 7741
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 0.0,
+      "learning_rate": 3.205855674714986e-09,
+      "loss": 0.2294,
+      "step": 7742
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1016125994631776e-09,
+      "loss": 0.2695,
+      "step": 7743
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.999092196908082e-09,
+      "loss": 0.2363,
+      "step": 7744
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.898294484714459e-09,
+      "loss": 0.2846,
+      "step": 7745
+    },
+    {
+      "epoch": 8.93,
+      "grad_norm": 0.0,
+      "learning_rate": 2.799219480251747e-09,
+      "loss": 0.2654,
+      "step": 7746
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.701867200592956e-09,
+      "loss": 0.2599,
+      "step": 7747
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6062376625146657e-09,
+      "loss": 0.2459,
+      "step": 7748
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.512330882494807e-09,
+      "loss": 0.2459,
+      "step": 7749
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4201468767159895e-09,
+      "loss": 0.2768,
+      "step": 7750
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.329685661063286e-09,
+      "loss": 0.2417,
+      "step": 7751
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.240947251125336e-09,
+      "loss": 0.2762,
+      "step": 7752
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1539316621921326e-09,
+      "loss": 0.2986,
+      "step": 7753
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0686389092605674e-09,
+      "loss": 0.283,
+      "step": 7754
+    },
+    {
+      "epoch": 8.94,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9850690070266633e-09,
+      "loss": 0.3076,
+      "step": 7755
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9032219698911226e-09,
+      "loss": 0.2481,
+      "step": 7756
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.823097811959329e-09,
+      "loss": 0.2726,
+      "step": 7757
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.744696547036906e-09,
+      "loss": 0.252,
+      "step": 7758
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6680181886352676e-09,
+      "loss": 0.232,
+      "step": 7759
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.593062749967178e-09,
+      "loss": 0.2253,
+      "step": 7760
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.519830243948972e-09,
+      "loss": 0.2519,
+      "step": 7761
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4483206831994446e-09,
+      "loss": 0.2298,
+      "step": 7762
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 0.0,
+      "learning_rate": 1.378534080042071e-09,
+      "loss": 0.2181,
+      "step": 7763
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.310470446501677e-09,
+      "loss": 0.2877,
+      "step": 7764
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.244129794307769e-09,
+      "loss": 0.3071,
+      "step": 7765
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1795121348923134e-09,
+      "loss": 0.267,
+      "step": 7766
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1166174793908468e-09,
+      "loss": 0.2492,
+      "step": 7767
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0554458386391464e-09,
+      "loss": 0.2601,
+      "step": 7768
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.0,
+      "learning_rate": 9.959972231798898e-10,
+      "loss": 0.2759,
+      "step": 7769
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.0,
+      "learning_rate": 9.38271643258215e-10,
+      "loss": 0.2476,
+      "step": 7770
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.0,
+      "learning_rate": 8.822691088195001e-10,
+      "loss": 0.259,
+      "step": 7771
+    },
+    {
+      "epoch": 8.96,
+      "grad_norm": 0.0,
+      "learning_rate": 8.279896295160239e-10,
+      "loss": 0.2925,
+      "step": 7772
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 0.0,
+      "learning_rate": 7.75433214699195e-10,
+      "loss": 0.267,
+      "step": 7773
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 0.0,
+      "learning_rate": 7.24599873427323e-10,
+      "loss": 0.2573,
+      "step": 7774
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 0.0,
+      "learning_rate": 6.754896144589573e-10,
+      "loss": 0.2156,
+      "step": 7775
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 0.0,
+      "learning_rate": 6.281024462584384e-10,
+      "loss": 0.2668,
+      "step": 7776
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 0.0,
+      "learning_rate": 5.824383769892361e-10,
+      "loss": 0.2581,
+      "step": 7777
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 0.0,
+      "learning_rate": 5.384974145217214e-10,
+      "loss": 0.2337,
+      "step": 7778
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 0.0,
+      "learning_rate": 4.962795664265052e-10,
+      "loss": 0.3111,
+      "step": 7779
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5578483998109934e-10,
+      "loss": 0.2721,
+      "step": 7780
+    },
+    {
+      "epoch": 8.97,
+      "grad_norm": 0.0,
+      "learning_rate": 4.170132421610351e-10,
+      "loss": 0.2288,
+      "step": 7781
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 0.0,
+      "learning_rate": 3.799647796487449e-10,
+      "loss": 0.2519,
+      "step": 7782
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 0.0,
+      "learning_rate": 3.446394588280111e-10,
+      "loss": 0.2635,
+      "step": 7783
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 0.0,
+      "learning_rate": 3.110372857861865e-10,
+      "loss": 0.2885,
+      "step": 7784
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 0.0,
+      "learning_rate": 2.791582663141945e-10,
+      "loss": 0.2604,
+      "step": 7785
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 0.0,
+      "learning_rate": 2.490024059054186e-10,
+      "loss": 0.2195,
+      "step": 7786
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2056970975459223e-10,
+      "loss": 0.2646,
+      "step": 7787
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9386018276335017e-10,
+      "loss": 0.241,
+      "step": 7788
+    },
+    {
+      "epoch": 8.98,
+      "grad_norm": 0.0,
+      "learning_rate": 1.688738295335668e-10,
+      "loss": 0.2868,
+      "step": 7789
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.45610654370687e-10,
+      "loss": 0.3046,
+      "step": 7790
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2407066128372613e-10,
+      "loss": 0.2649,
+      "step": 7791
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0425385398304955e-10,
+      "loss": 0.2847,
+      "step": 7792
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 0.0,
+      "learning_rate": 8.61602358859237e-11,
+      "loss": 0.2673,
+      "step": 7793
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 0.0,
+      "learning_rate": 6.978981010985486e-11,
+      "loss": 0.2235,
+      "step": 7794
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 0.0,
+      "learning_rate": 5.514257947369928e-11,
+      "loss": 0.2457,
+      "step": 7795
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 0.0,
+      "learning_rate": 4.221854650321433e-11,
+      "loss": 0.2492,
+      "step": 7796
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 0.0,
+      "learning_rate": 3.101771342550741e-11,
+      "loss": 0.3082,
+      "step": 7797
+    },
+    {
+      "epoch": 8.99,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1540082170146137e-11,
+      "loss": 0.2753,
+      "step": 7798
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3785654370268575e-11,
+      "loss": 0.2439,
+      "step": 7799
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.0,
+      "learning_rate": 7.754431362583247e-12,
+      "loss": 0.2815,
+      "step": 7800
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4464141862589064e-12,
+      "loss": 0.2852,
+      "step": 7801
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.0,
+      "learning_rate": 8.616035840347536e-13,
+      "loss": 0.2566,
+      "step": 7802
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 0.2586,
+      "step": 7803
+    },
+    {
+      "epoch": 9.0,
+      "step": 7803,
+      "total_flos": 4.7906648348635955e+17,
+      "train_loss": 0.5852041420311129,
+      "train_runtime": 39282.0043,
+      "train_samples_per_second": 6.355,
+      "train_steps_per_second": 0.199
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 7803,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 9,
+  "save_steps": 2000,
+  "total_flos": 4.7906648348635955e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}