diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.032951341851865414,
+  "eval_steps": 500,
+  "global_step": 1800,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "acc_char": 0.2961,
+      "acc_token": 0.6732,
+      "epoch": 0.0,
+      "loss_char": 0.7519,
+      "loss_token": 1.6949,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 1
+    },
+    {
+      "acc_char": 0.3069,
+      "acc_token": 0.6884,
+      "epoch": 0.0,
+      "loss_char": 0.7203,
+      "loss_token": 1.6238,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 2
+    },
+    {
+      "acc_char": 0.2884,
+      "acc_token": 0.6605,
+      "epoch": 0.0,
+      "loss_char": 0.7484,
+      "loss_token": 1.6386,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 3
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6459,
+      "epoch": 0.0,
+      "loss_char": 0.8252,
+      "loss_token": 1.7039,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 4
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6684,
+      "epoch": 0.0,
+      "loss_char": 0.7786,
+      "loss_token": 1.6981,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 5
+    },
+    {
+      "acc_char": 0.3093,
+      "acc_token": 0.7014,
+      "epoch": 0.0,
+      "loss_char": 0.6768,
+      "loss_token": 1.5207,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 6
+    },
+    {
+      "acc_char": 0.3094,
+      "acc_token": 0.689,
+      "epoch": 0.0,
+      "loss_char": 0.6929,
+      "loss_token": 1.5691,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 7
+    },
+    {
+      "acc_char": 0.2961,
+      "acc_token": 0.6754,
+      "epoch": 0.0,
+      "loss_char": 0.7308,
+      "loss_token": 1.6342,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 8
+    },
+    {
+      "acc_char": 0.2917,
+      "acc_token": 0.664,
+      "epoch": 0.0,
+      "loss_char": 0.7558,
+      "loss_token": 1.6594,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 9
+    },
+    {
+      "acc_char": 0.2949,
+      "acc_token": 0.6353,
+      "epoch": 0.0,
+      "loss_char": 0.8635,
+      "loss_token": 1.7367,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 10
+    },
+    {
+      "acc_char": 0.2912,
+      "acc_token": 0.6564,
+      "epoch": 0.0,
+      "loss_char": 0.7989,
+      "loss_token": 1.6703,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 11
+    },
+    {
+      "acc_char": 0.2945,
+      "acc_token": 0.6563,
+      "epoch": 0.0,
+      "loss_char": 0.815,
+      "loss_token": 1.7397,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 12
+    },
+    {
+      "acc_char": 0.3069,
+      "acc_token": 0.7084,
+      "epoch": 0.0,
+      "loss_char": 0.644,
+      "loss_token": 1.5543,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 13
+    },
+    {
+      "acc_char": 0.3055,
+      "acc_token": 0.677,
+      "epoch": 0.0,
+      "loss_char": 0.7408,
+      "loss_token": 1.619,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 14
+    },
+    {
+      "acc_char": 0.3079,
+      "acc_token": 0.6895,
+      "epoch": 0.0,
+      "loss_char": 0.7422,
+      "loss_token": 1.6489,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 15
+    },
+    {
+      "acc_char": 0.2986,
+      "acc_token": 0.6883,
+      "epoch": 0.0,
+      "loss_char": 0.6958,
+      "loss_token": 1.5508,
+      "lr": "0.00e+00",
+      "norm": "0.00e+00",
+      "step": 16
+    },
+    {
+      "acc_char": 0.2952,
+      "acc_token": 0.6751,
+      "epoch": 0.0,
+      "loss_char": 0.7564,
+      "loss_token": 1.7334,
+      "lr": "1.00e-07",
+      "norm": 5.9607,
+      "step": 17
+    },
+    {
+      "acc_char": 0.2883,
+      "acc_token": 0.6696,
+      "epoch": 0.0,
+      "loss_char": 0.7107,
+      "loss_token": 1.7119,
+      "lr": "1.00e-07",
+      "norm": 5.9607,
+      "step": 18
+    },
+    {
+      "acc_char": 0.2939,
+      "acc_token": 0.667,
+      "epoch": 0.0,
+      "loss_char": 0.7545,
+      "loss_token": 1.7812,
+      "lr": "2.00e-07",
+      "norm": 6.3926,
+      "step": 19
+    },
+    {
+      "acc_char": 0.3025,
+      "acc_token": 0.689,
+      "epoch": 0.0,
+      "loss_char": 0.7043,
+      "loss_token": 1.6273,
+      "lr": "2.00e-07",
+      "norm": 5.7886,
+      "step": 20
+    },
+    {
+      "acc_char": 0.29,
+      "acc_token": 0.6184,
+      "epoch": 0.0,
+      "loss_char": 0.9418,
+      "loss_token": 1.7781,
+      "lr": "2.00e-07",
+      "norm": 5.9238,
+      "step": 21
+    },
+    {
+      "acc_char": 0.3036,
+      "acc_token": 0.6826,
+      "epoch": 0.0,
+      "loss_char": 0.7337,
+      "loss_token": 1.6468,
+      "lr": "2.00e-07",
+      "norm": 6.2689,
+      "step": 22
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6894,
+      "epoch": 0.0,
+      "loss_char": 0.703,
+      "loss_token": 1.6869,
+      "lr": "2.00e-07",
+      "norm": 5.9656,
+      "step": 23
+    },
+    {
+      "acc_char": 0.303,
+      "acc_token": 0.6772,
+      "epoch": 0.0,
+      "loss_char": 0.7587,
+      "loss_token": 1.6943,
+      "lr": "2.00e-07",
+      "norm": 5.8126,
+      "step": 24
+    },
+    {
+      "acc_char": 0.2892,
+      "acc_token": 0.6357,
+      "epoch": 0.0,
+      "loss_char": 0.826,
+      "loss_token": 1.761,
+      "lr": "2.00e-07",
+      "norm": 5.7309,
+      "step": 25
+    },
+    {
+      "acc_char": 0.2755,
+      "acc_token": 0.6058,
+      "epoch": 0.0,
+      "loss_char": 0.932,
+      "loss_token": 1.8898,
+      "lr": "2.00e-07",
+      "norm": 5.7804,
+      "step": 26
+    },
+    {
+      "acc_char": 0.2905,
+      "acc_token": 0.6638,
+      "epoch": 0.0,
+      "loss_char": 0.7657,
+      "loss_token": 1.6386,
+      "lr": "2.00e-07",
+      "norm": 5.6621,
+      "step": 27
+    },
+    {
+      "acc_char": 0.2955,
+      "acc_token": 0.6773,
+      "epoch": 0.0,
+      "loss_char": 0.72,
+      "loss_token": 1.7086,
+      "lr": "2.00e-07",
+      "norm": 5.9964,
+      "step": 28
+    },
+    {
+      "acc_char": 0.2825,
+      "acc_token": 0.6586,
+      "epoch": 0.0,
+      "loss_char": 0.7422,
+      "loss_token": 1.7517,
+      "lr": "2.00e-07",
+      "norm": 5.8231,
+      "step": 29
+    },
+    {
+      "acc_char": 0.2974,
+      "acc_token": 0.6798,
+      "epoch": 0.0,
+      "loss_char": 0.7219,
+      "loss_token": 1.7468,
+      "lr": "2.00e-07",
+      "norm": 5.9514,
+      "step": 30
+    },
+    {
+      "acc_char": 0.2956,
+      "acc_token": 0.6434,
+      "epoch": 0.0,
+      "loss_char": 0.8223,
+      "loss_token": 1.6483,
+      "lr": "2.00e-07",
+      "norm": 5.6608,
+      "step": 31
+    },
+    {
+      "acc_char": 0.3033,
+      "acc_token": 0.6829,
+      "epoch": 0.0,
+      "loss_char": 0.7012,
+      "loss_token": 1.5172,
+      "lr": "2.00e-07",
+      "norm": 5.652,
+      "step": 32
+    },
+    {
+      "acc_char": 0.3007,
+      "acc_token": 0.6764,
+      "epoch": 0.0,
+      "loss_char": 0.7588,
+      "loss_token": 1.7038,
+      "lr": "2.00e-07",
+      "norm": 5.8963,
+      "step": 33
+    },
+    {
+      "acc_char": 0.2916,
+      "acc_token": 0.6639,
+      "epoch": 0.0,
+      "loss_char": 0.7669,
+      "loss_token": 1.7697,
+      "lr": "2.00e-07",
+      "norm": 6.2551,
+      "step": 34
+    },
+    {
+      "acc_char": 0.2868,
+      "acc_token": 0.6746,
+      "epoch": 0.0,
+      "loss_char": 0.7113,
+      "loss_token": 1.7143,
+      "lr": "2.00e-07",
+      "norm": 5.8972,
+      "step": 35
+    },
+    {
+      "acc_char": 0.2844,
+      "acc_token": 0.6591,
+      "epoch": 0.0,
+      "loss_char": 0.7909,
+      "loss_token": 1.8081,
+      "lr": "2.00e-07",
+      "norm": 6.105,
+      "step": 36
+    },
+    {
+      "acc_char": 0.3036,
+      "acc_token": 0.6838,
+      "epoch": 0.0,
+      "loss_char": 0.7246,
+      "loss_token": 1.6531,
+      "lr": "2.00e-07",
+      "norm": 5.7111,
+      "step": 37
+    },
+    {
+      "acc_char": 0.2848,
+      "acc_token": 0.6208,
+      "epoch": 0.0,
+      "loss_char": 0.8921,
+      "loss_token": 1.8735,
+      "lr": "2.00e-07",
+      "norm": 5.7803,
+      "step": 38
+    },
+    {
+      "acc_char": 0.2753,
+      "acc_token": 0.5703,
+      "epoch": 0.0,
+      "loss_char": 1.0188,
+      "loss_token": 1.9257,
+      "lr": "2.00e-07",
+      "norm": 5.8799,
+      "step": 39
+    },
+    {
+      "acc_char": 0.3044,
+      "acc_token": 0.6826,
+      "epoch": 0.0,
+      "loss_char": 0.7192,
+      "loss_token": 1.6524,
+      "lr": "2.00e-07",
+      "norm": 5.8362,
+      "step": 40
+    },
+    {
+      "acc_char": 0.313,
+      "acc_token": 0.6983,
+      "epoch": 0.0,
+      "loss_char": 0.6946,
+      "loss_token": 1.6673,
+      "lr": "2.00e-07",
+      "norm": 5.733,
+      "step": 41
+    },
+    {
+      "acc_char": 0.3086,
+      "acc_token": 0.7071,
+      "epoch": 0.0,
+      "loss_char": 0.668,
+      "loss_token": 1.5743,
+      "lr": "2.00e-07",
+      "norm": 5.5149,
+      "step": 42
+    },
+    {
+      "acc_char": 0.2965,
+      "acc_token": 0.6688,
+      "epoch": 0.0,
+      "loss_char": 0.7779,
+      "loss_token": 1.6748,
+      "lr": "2.00e-07",
+      "norm": 5.7856,
+      "step": 43
+    },
+    {
+      "acc_char": 0.298,
+      "acc_token": 0.6751,
+      "epoch": 0.0,
+      "loss_char": 0.7445,
+      "loss_token": 1.7059,
+      "lr": "2.00e-07",
+      "norm": 5.9937,
+      "step": 44
+    },
+    {
+      "acc_char": 0.2831,
+      "acc_token": 0.6538,
+      "epoch": 0.0,
+      "loss_char": 0.7752,
+      "loss_token": 1.6552,
+      "lr": "2.00e-07",
+      "norm": 5.5977,
+      "step": 45
+    },
+    {
+      "acc_char": 0.2992,
+      "acc_token": 0.677,
+      "epoch": 0.0,
+      "loss_char": 0.7412,
+      "loss_token": 1.6394,
+      "lr": "2.00e-07",
+      "norm": 5.6584,
+      "step": 46
+    },
+    {
+      "acc_char": 0.2969,
+      "acc_token": 0.6693,
+      "epoch": 0.0,
+      "loss_char": 0.7711,
+      "loss_token": 1.6313,
+      "lr": "2.00e-07",
+      "norm": 5.8419,
+      "step": 47
+    },
+    {
+      "acc_char": 0.2975,
+      "acc_token": 0.6811,
+      "epoch": 0.0,
+      "loss_char": 0.7134,
+      "loss_token": 1.7313,
+      "lr": "2.00e-07",
+      "norm": 5.9012,
+      "step": 48
+    },
+    {
+      "acc_char": 0.3007,
+      "acc_token": 0.6828,
+      "epoch": 0.0,
+      "loss_char": 0.7124,
+      "loss_token": 1.6284,
+      "lr": "2.00e-07",
+      "norm": 5.6233,
+      "step": 49
+    },
+    {
+      "acc_char": 0.2972,
+      "acc_token": 0.6354,
+      "epoch": 0.0,
+      "loss_char": 0.8823,
+      "loss_token": 1.7413,
+      "lr": "2.00e-07",
+      "norm": 5.3973,
+      "step": 50
+    },
+    {
+      "acc_char": 0.3022,
+      "acc_token": 0.6905,
+      "epoch": 0.0,
+      "loss_char": 0.7153,
+      "loss_token": 1.5482,
+      "lr": "2.00e-07",
+      "norm": 5.6895,
+      "step": 51
+    },
+    {
+      "acc_char": 0.2982,
+      "acc_token": 0.6686,
+      "epoch": 0.0,
+      "loss_char": 0.7733,
+      "loss_token": 1.6329,
+      "lr": "2.00e-07",
+      "norm": 6.9704,
+      "step": 52
+    },
+    {
+      "acc_char": 0.2877,
+      "acc_token": 0.6616,
+      "epoch": 0.0,
+      "loss_char": 0.7528,
+      "loss_token": 1.6389,
+      "lr": "2.00e-07",
+      "norm": 6.0386,
+      "step": 53
+    },
+    {
+      "acc_char": 0.298,
+      "acc_token": 0.6711,
+      "epoch": 0.0,
+      "loss_char": 0.7559,
+      "loss_token": 1.7198,
+      "lr": "2.00e-07",
+      "norm": 5.7229,
+      "step": 54
+    },
+    {
+      "acc_char": 0.2947,
+      "acc_token": 0.6745,
+      "epoch": 0.0,
+      "loss_char": 0.7399,
+      "loss_token": 1.7187,
+      "lr": "2.00e-07",
+      "norm": 5.7068,
+      "step": 55
+    },
+    {
+      "acc_char": 0.3029,
+      "acc_token": 0.6835,
+      "epoch": 0.0,
+      "loss_char": 0.7238,
+      "loss_token": 1.5982,
+      "lr": "2.00e-07",
+      "norm": 5.517,
+      "step": 56
+    },
+    {
+      "acc_char": 0.3065,
+      "acc_token": 0.6897,
+      "epoch": 0.0,
+      "loss_char": 0.7217,
+      "loss_token": 1.6287,
+      "lr": "2.00e-07",
+      "norm": 5.6666,
+      "step": 57
+    },
+    {
+      "acc_char": 0.3045,
+      "acc_token": 0.6686,
+      "epoch": 0.0,
+      "loss_char": 0.7933,
+      "loss_token": 1.6452,
+      "lr": "2.00e-07",
+      "norm": 5.8336,
+      "step": 58
+    },
+    {
+      "acc_char": 0.2997,
+      "acc_token": 0.6734,
+      "epoch": 0.0,
+      "loss_char": 0.75,
+      "loss_token": 1.6541,
+      "lr": "2.00e-07",
+      "norm": 5.731,
+      "step": 59
+    },
+    {
+      "acc_char": 0.2929,
+      "acc_token": 0.6754,
+      "epoch": 0.0,
+      "loss_char": 0.742,
+      "loss_token": 1.7271,
+      "lr": "2.00e-07",
+      "norm": 6.1144,
+      "step": 60
+    },
+    {
+      "acc_char": 0.2977,
+      "acc_token": 0.6707,
+      "epoch": 0.0,
+      "loss_char": 0.7895,
+      "loss_token": 1.7199,
+      "lr": "2.00e-07",
+      "norm": 5.7852,
+      "step": 61
+    },
+    {
+      "acc_char": 0.299,
+      "acc_token": 0.6724,
+      "epoch": 0.0,
+      "loss_char": 0.7415,
+      "loss_token": 1.6173,
+      "lr": "2.00e-07",
+      "norm": 5.7001,
+      "step": 62
+    },
+    {
+      "acc_char": 0.3119,
+      "acc_token": 0.7042,
+      "epoch": 0.0,
+      "loss_char": 0.6822,
+      "loss_token": 1.6126,
+      "lr": "2.00e-07",
+      "norm": 5.7258,
+      "step": 63
+    },
+    {
+      "acc_char": 0.3012,
+      "acc_token": 0.6924,
+      "epoch": 0.0,
+      "loss_char": 0.6868,
+      "loss_token": 1.6205,
+      "lr": "2.00e-07",
+      "norm": 5.8205,
+      "step": 64
+    },
+    {
+      "acc_char": 0.3083,
+      "acc_token": 0.6987,
+      "epoch": 0.0,
+      "loss_char": 0.6842,
+      "loss_token": 1.6814,
+      "lr": "2.00e-07",
+      "norm": 5.8432,
+      "step": 65
+    },
+    {
+      "acc_char": 0.31,
+      "acc_token": 0.7027,
+      "epoch": 0.0,
+      "loss_char": 0.673,
+      "loss_token": 1.6156,
+      "lr": "2.00e-07",
+      "norm": 5.914,
+      "step": 66
+    },
+    {
+      "acc_char": 0.3223,
+      "acc_token": 0.7415,
+      "epoch": 0.0,
+      "loss_char": 0.5721,
+      "loss_token": 1.4831,
+      "lr": "2.00e-07",
+      "norm": 5.512,
+      "step": 67
+    },
+    {
+      "acc_char": 0.2937,
+      "acc_token": 0.6344,
+      "epoch": 0.0,
+      "loss_char": 0.8655,
+      "loss_token": 1.815,
+      "lr": "2.00e-07",
+      "norm": 5.8184,
+      "step": 68
+    },
+    {
+      "acc_char": 0.3014,
+      "acc_token": 0.6775,
+      "epoch": 0.0,
+      "loss_char": 0.7422,
+      "loss_token": 1.682,
+      "lr": "2.00e-07",
+      "norm": 5.5313,
+      "step": 69
+    },
+    {
+      "acc_char": 0.3014,
+      "acc_token": 0.6931,
+      "epoch": 0.0,
+      "loss_char": 0.6702,
+      "loss_token": 1.5349,
+      "lr": "2.00e-07",
+      "norm": 5.5487,
+      "step": 70
+    },
+    {
+      "acc_char": 0.2931,
+      "acc_token": 0.6635,
+      "epoch": 0.0,
+      "loss_char": 0.7725,
+      "loss_token": 1.6859,
+      "lr": "2.00e-07",
+      "norm": 5.7526,
+      "step": 71
+    },
+    {
+      "acc_char": 0.3132,
+      "acc_token": 0.7059,
+      "epoch": 0.0,
+      "loss_char": 0.6693,
+      "loss_token": 1.5245,
+      "lr": "2.00e-07",
+      "norm": 5.511,
+      "step": 72
+    },
+    {
+      "acc_char": 0.2954,
+      "acc_token": 0.6704,
+      "epoch": 0.0,
+      "loss_char": 0.7482,
+      "loss_token": 1.6966,
+      "lr": "2.00e-07",
+      "norm": 5.7717,
+      "step": 73
+    },
+    {
+      "acc_char": 0.2911,
+      "acc_token": 0.6618,
+      "epoch": 0.0,
+      "loss_char": 0.7721,
+      "loss_token": 1.7499,
+      "lr": "2.00e-07",
+      "norm": 5.9435,
+      "step": 74
+    },
+    {
+      "acc_char": 0.3034,
+      "acc_token": 0.6888,
+      "epoch": 0.0,
+      "loss_char": 0.698,
+      "loss_token": 1.5809,
+      "lr": "2.00e-07",
+      "norm": 5.2705,
+      "step": 75
+    },
+    {
+      "acc_char": 0.3086,
+      "acc_token": 0.6984,
+      "epoch": 0.0,
+      "loss_char": 0.6715,
+      "loss_token": 1.6068,
+      "lr": "2.00e-07",
+      "norm": 5.429,
+      "step": 76
+    },
+    {
+      "acc_char": 0.2997,
+      "acc_token": 0.6732,
+      "epoch": 0.0,
+      "loss_char": 0.7512,
+      "loss_token": 1.649,
+      "lr": "2.00e-07",
+      "norm": 5.7104,
+      "step": 77
+    },
+    {
+      "acc_char": 0.2966,
+      "acc_token": 0.6677,
+      "epoch": 0.0,
+      "loss_char": 0.7483,
+      "loss_token": 1.6908,
+      "lr": "2.00e-07",
+      "norm": 5.844,
+      "step": 78
+    },
+    {
+      "acc_char": 0.3089,
+      "acc_token": 0.6839,
+      "epoch": 0.0,
+      "loss_char": 0.7344,
+      "loss_token": 1.6171,
+      "lr": "2.00e-07",
+      "norm": 5.5551,
+      "step": 79
+    },
+    {
+      "acc_char": 0.3002,
+      "acc_token": 0.6858,
+      "epoch": 0.0,
+      "loss_char": 0.7015,
+      "loss_token": 1.5875,
+      "lr": "2.00e-07",
+      "norm": 5.7256,
+      "step": 80
+    },
+    {
+      "acc_char": 0.306,
+      "acc_token": 0.6897,
+      "epoch": 0.0,
+      "loss_char": 0.7126,
+      "loss_token": 1.6529,
+      "lr": "2.00e-07",
+      "norm": 5.757,
+      "step": 81
+    },
+    {
+      "acc_char": 0.29,
+      "acc_token": 0.6649,
+      "epoch": 0.0,
+      "loss_char": 0.7649,
+      "loss_token": 1.6658,
+      "lr": "2.00e-07",
+      "norm": 5.6632,
+      "step": 82
+    },
+    {
+      "acc_char": 0.3139,
+      "acc_token": 0.692,
+      "epoch": 0.0,
+      "loss_char": 0.7281,
+      "loss_token": 1.563,
+      "lr": "2.00e-07",
+      "norm": 5.7469,
+      "step": 83
+    },
+    {
+      "acc_char": 0.3093,
+      "acc_token": 0.6999,
+      "epoch": 0.0,
+      "loss_char": 0.6736,
+      "loss_token": 1.5781,
+      "lr": "2.00e-07",
+      "norm": 6.2755,
+      "step": 84
+    },
+    {
+      "acc_char": 0.3039,
+      "acc_token": 0.6809,
+      "epoch": 0.0,
+      "loss_char": 0.7506,
+      "loss_token": 1.7318,
+      "lr": "2.00e-07",
+      "norm": 5.5039,
+      "step": 85
+    },
+    {
+      "acc_char": 0.2907,
+      "acc_token": 0.6696,
+      "epoch": 0.0,
+      "loss_char": 0.7501,
+      "loss_token": 1.7583,
+      "lr": "2.00e-07",
+      "norm": 5.8564,
+      "step": 86
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.6717,
+      "epoch": 0.0,
+      "loss_char": 0.7485,
+      "loss_token": 1.7149,
+      "lr": "2.00e-07",
+      "norm": 5.8721,
+      "step": 87
+    },
+    {
+      "acc_char": 0.3062,
+      "acc_token": 0.685,
+      "epoch": 0.0,
+      "loss_char": 0.7216,
+      "loss_token": 1.6338,
+      "lr": "2.00e-07",
+      "norm": 5.7368,
+      "step": 88
+    },
+    {
+      "acc_char": 0.3074,
+      "acc_token": 0.6946,
+      "epoch": 0.0,
+      "loss_char": 0.698,
+      "loss_token": 1.6557,
+      "lr": "2.00e-07",
+      "norm": 5.6536,
+      "step": 89
+    },
+    {
+      "acc_char": 0.3309,
+      "acc_token": 0.7018,
+      "epoch": 0.0,
+      "loss_char": 0.6798,
+      "loss_token": 1.5062,
+      "lr": "2.00e-07",
+      "norm": 5.6085,
+      "step": 90
+    },
+    {
+      "acc_char": 0.3205,
+      "acc_token": 0.7148,
+      "epoch": 0.0,
+      "loss_char": 0.6533,
+      "loss_token": 1.5137,
+      "lr": "2.00e-07",
+      "norm": 5.9149,
+      "step": 91
+    },
+    {
+      "acc_char": 0.2983,
+      "acc_token": 0.6762,
+      "epoch": 0.0,
+      "loss_char": 0.7395,
+      "loss_token": 1.588,
+      "lr": "2.00e-07",
+      "norm": 5.5893,
+      "step": 92
+    },
+    {
+      "acc_char": 0.2925,
+      "acc_token": 0.6386,
+      "epoch": 0.0,
+      "loss_char": 0.8278,
+      "loss_token": 1.7404,
+      "lr": "2.00e-07",
+      "norm": 5.391,
+      "step": 93
+    },
+    {
+      "acc_char": 0.2976,
+      "acc_token": 0.6813,
+      "epoch": 0.0,
+      "loss_char": 0.7177,
+      "loss_token": 1.6489,
+      "lr": "2.00e-07",
+      "norm": 5.5271,
+      "step": 94
+    },
+    {
+      "acc_char": 0.3023,
+      "acc_token": 0.6715,
+      "epoch": 0.0,
+      "loss_char": 0.7614,
+      "loss_token": 1.6439,
+      "lr": "2.00e-07",
+      "norm": 5.7352,
+      "step": 95
+    },
+    {
+      "acc_char": 0.3102,
+      "acc_token": 0.7062,
+      "epoch": 0.0,
+      "loss_char": 0.6556,
+      "loss_token": 1.5339,
+      "lr": "2.00e-07",
+      "norm": 5.5893,
+      "step": 96
+    },
+    {
+      "acc_char": 0.3,
+      "acc_token": 0.6866,
+      "epoch": 0.0,
+      "loss_char": 0.7047,
+      "loss_token": 1.669,
+      "lr": "2.00e-07",
+      "norm": 5.4678,
+      "step": 97
+    },
+    {
+      "acc_char": 0.2867,
+      "acc_token": 0.6561,
+      "epoch": 0.0,
+      "loss_char": 0.7706,
+      "loss_token": 1.742,
+      "lr": "2.00e-07",
+      "norm": 6.2133,
+      "step": 98
+    },
+    {
+      "acc_char": 0.2934,
+      "acc_token": 0.6788,
+      "epoch": 0.0,
+      "loss_char": 0.7435,
+      "loss_token": 1.5761,
+      "lr": "2.00e-07",
+      "norm": 5.9936,
+      "step": 99
+    },
+    {
+      "acc_char": 0.3152,
+      "acc_token": 0.7196,
+      "epoch": 0.0,
+      "loss_char": 0.6295,
+      "loss_token": 1.4385,
+      "lr": "2.00e-07",
+      "norm": 5.3239,
+      "step": 100
+    },
+    {
+      "acc_char": 0.2895,
+      "acc_token": 0.6691,
+      "epoch": 0.0,
+      "loss_char": 0.7532,
+      "loss_token": 1.667,
+      "lr": "2.00e-07",
+      "norm": 5.7719,
+      "step": 101
+    },
+    {
+      "acc_char": 0.2954,
+      "acc_token": 0.6687,
+      "epoch": 0.0,
+      "loss_char": 0.7625,
+      "loss_token": 1.6473,
+      "lr": "2.00e-07",
+      "norm": 5.8406,
+      "step": 102
+    },
+    {
+      "acc_char": 0.2919,
+      "acc_token": 0.6414,
+      "epoch": 0.0,
+      "loss_char": 0.8862,
+      "loss_token": 1.8522,
+      "lr": "2.00e-07",
+      "norm": 5.8152,
+      "step": 103
+    },
+    {
+      "acc_char": 0.2972,
+      "acc_token": 0.688,
+      "epoch": 0.0,
+      "loss_char": 0.7107,
+      "loss_token": 1.6225,
+      "lr": "2.00e-07",
+      "norm": 5.7036,
+      "step": 104
+    },
+    {
+      "acc_char": 0.3106,
+      "acc_token": 0.7055,
+      "epoch": 0.0,
+      "loss_char": 0.6515,
+      "loss_token": 1.5368,
+      "lr": "2.00e-07",
+      "norm": 5.5132,
+      "step": 105
+    },
+    {
+      "acc_char": 0.3163,
+      "acc_token": 0.7085,
+      "epoch": 0.0,
+      "loss_char": 0.6619,
+      "loss_token": 1.5181,
+      "lr": "2.00e-07",
+      "norm": 5.7377,
+      "step": 106
+    },
+    {
+      "acc_char": 0.296,
+      "acc_token": 0.6795,
+      "epoch": 0.0,
+      "loss_char": 0.724,
+      "loss_token": 1.6804,
+      "lr": "2.00e-07",
+      "norm": 5.6408,
+      "step": 107
+    },
+    {
+      "acc_char": 0.3114,
+      "acc_token": 0.6911,
+      "epoch": 0.0,
+      "loss_char": 0.7201,
+      "loss_token": 1.4931,
+      "lr": "2.00e-07",
+      "norm": 5.6982,
+      "step": 108
+    },
+    {
+      "acc_char": 0.3197,
+      "acc_token": 0.7152,
+      "epoch": 0.0,
+      "loss_char": 0.6508,
+      "loss_token": 1.5723,
+      "lr": "2.00e-07",
+      "norm": 5.9246,
+      "step": 109
+    },
+    {
+      "acc_char": 0.3119,
+      "acc_token": 0.6919,
+      "epoch": 0.0,
+      "loss_char": 0.7435,
+      "loss_token": 1.6627,
+      "lr": "2.00e-07",
+      "norm": 6.0751,
+      "step": 110
+    },
+    {
+      "acc_char": 0.2963,
+      "acc_token": 0.648,
+      "epoch": 0.0,
+      "loss_char": 0.8038,
+      "loss_token": 1.7077,
+      "lr": "2.00e-07",
+      "norm": 5.5694,
+      "step": 111
+    },
+    {
+      "acc_char": 0.2931,
+      "acc_token": 0.6728,
+      "epoch": 0.0,
+      "loss_char": 0.7443,
+      "loss_token": 1.6811,
+      "lr": "2.00e-07",
+      "norm": 6.3389,
+      "step": 112
+    },
+    {
+      "acc_char": 0.2902,
+      "acc_token": 0.6706,
+      "epoch": 0.0,
+      "loss_char": 0.7368,
+      "loss_token": 1.5372,
+      "lr": "2.00e-07",
+      "norm": 5.4269,
+      "step": 113
+    },
+    {
+      "acc_char": 0.2961,
+      "acc_token": 0.6749,
+      "epoch": 0.0,
+      "loss_char": 0.762,
+      "loss_token": 1.76,
+      "lr": "2.00e-07",
+      "norm": 5.9477,
+      "step": 114
+    },
+    {
+      "acc_char": 0.3006,
+      "acc_token": 0.6797,
+      "epoch": 0.0,
+      "loss_char": 0.7402,
+      "loss_token": 1.6541,
+      "lr": "2.00e-07",
+      "norm": 5.6794,
+      "step": 115
+    },
+    {
+      "acc_char": 0.3053,
+      "acc_token": 0.6901,
+      "epoch": 0.0,
+      "loss_char": 0.6922,
+      "loss_token": 1.5729,
+      "lr": "2.00e-07",
+      "norm": 5.4315,
+      "step": 116
+    },
+    {
+      "acc_char": 0.3027,
+      "acc_token": 0.676,
+      "epoch": 0.0,
+      "loss_char": 0.7406,
+      "loss_token": 1.7055,
+      "lr": "2.00e-07",
+      "norm": 6.0586,
+      "step": 117
+    },
+    {
+      "acc_char": 0.3071,
+      "acc_token": 0.6983,
+      "epoch": 0.0,
+      "loss_char": 0.6718,
+      "loss_token": 1.6349,
+      "lr": "2.00e-07",
+      "norm": 5.6116,
+      "step": 118
+    },
+    {
+      "acc_char": 0.2997,
+      "acc_token": 0.6823,
+      "epoch": 0.0,
+      "loss_char": 0.7209,
+      "loss_token": 1.7037,
+      "lr": "2.00e-07",
+      "norm": 5.8724,
+      "step": 119
+    },
+    {
+      "acc_char": 0.3055,
+      "acc_token": 0.6851,
+      "epoch": 0.0,
+      "loss_char": 0.7164,
+      "loss_token": 1.5718,
+      "lr": "2.00e-07",
+      "norm": 6.0997,
+      "step": 120
+    },
+    {
+      "acc_char": 0.2934,
+      "acc_token": 0.6588,
+      "epoch": 0.0,
+      "loss_char": 0.8186,
+      "loss_token": 1.8209,
+      "lr": "2.00e-07",
+      "norm": 6.1657,
+      "step": 121
+    },
+    {
+      "acc_char": 0.3114,
+      "acc_token": 0.6917,
+      "epoch": 0.0,
+      "loss_char": 0.7182,
+      "loss_token": 1.6419,
+      "lr": "2.00e-07",
+      "norm": 5.763,
+      "step": 122
+    },
+    {
+      "acc_char": 0.3019,
+      "acc_token": 0.693,
+      "epoch": 0.0,
+      "loss_char": 0.6769,
+      "loss_token": 1.5495,
+      "lr": "2.00e-07",
+      "norm": 5.6592,
+      "step": 123
+    },
+    {
+      "acc_char": 0.2885,
+      "acc_token": 0.6628,
+      "epoch": 0.0,
+      "loss_char": 0.7731,
+      "loss_token": 1.6606,
+      "lr": "2.00e-07",
+      "norm": 5.8882,
+      "step": 124
+    },
+    {
+      "acc_char": 0.3051,
+      "acc_token": 0.6748,
+      "epoch": 0.0,
+      "loss_char": 0.7521,
+      "loss_token": 1.7262,
+      "lr": "2.00e-07",
+      "norm": 5.8277,
+      "step": 125
+    },
+    {
+      "acc_char": 0.2974,
+      "acc_token": 0.6753,
+      "epoch": 0.0,
+      "loss_char": 0.7267,
+      "loss_token": 1.6828,
+      "lr": "2.00e-07",
+      "norm": 5.8624,
+      "step": 126
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6713,
+      "epoch": 0.0,
+      "loss_char": 0.7441,
+      "loss_token": 1.6326,
+      "lr": "2.00e-07",
+      "norm": 5.6445,
+      "step": 127
+    },
+    {
+      "acc_char": 0.3151,
+      "acc_token": 0.7024,
+      "epoch": 0.0,
+      "loss_char": 0.7068,
+      "loss_token": 1.5497,
+      "lr": "2.00e-07",
+      "norm": 5.731,
+      "step": 128
+    },
+    {
+      "acc_char": 0.3007,
+      "acc_token": 0.6809,
+      "epoch": 0.0,
+      "loss_char": 0.7264,
+      "loss_token": 1.654,
+      "lr": "2.00e-07",
+      "norm": 5.6777,
+      "step": 129
+    },
+    {
+      "acc_char": 0.3074,
+      "acc_token": 0.7082,
+      "epoch": 0.0,
+      "loss_char": 0.6399,
+      "loss_token": 1.5027,
+      "lr": "2.00e-07",
+      "norm": 5.864,
+      "step": 130
+    },
+    {
+      "acc_char": 0.2976,
+      "acc_token": 0.667,
+      "epoch": 0.0,
+      "loss_char": 0.7645,
+      "loss_token": 1.663,
+      "lr": "2.00e-07",
+      "norm": 6.123,
+      "step": 131
+    },
+    {
+      "acc_char": 0.3146,
+      "acc_token": 0.703,
+      "epoch": 0.0,
+      "loss_char": 0.6845,
+      "loss_token": 1.5722,
+      "lr": "2.00e-07",
+      "norm": 5.6312,
+      "step": 132
+    },
+    {
+      "acc_char": 0.3033,
+      "acc_token": 0.684,
+      "epoch": 0.0,
+      "loss_char": 0.7349,
+      "loss_token": 1.6021,
+      "lr": "2.00e-07",
+      "norm": 5.604,
+      "step": 133
+    },
+    {
+      "acc_char": 0.3056,
+      "acc_token": 0.6914,
+      "epoch": 0.0,
+      "loss_char": 0.7159,
+      "loss_token": 1.627,
+      "lr": "2.00e-07",
+      "norm": 5.8821,
+      "step": 134
+    },
+    {
+      "acc_char": 0.301,
+      "acc_token": 0.6744,
+      "epoch": 0.0,
+      "loss_char": 0.7437,
+      "loss_token": 1.6091,
+      "lr": "2.00e-07",
+      "norm": 5.7012,
+      "step": 135
+    },
+    {
+      "acc_char": 0.2999,
+      "acc_token": 0.6793,
+      "epoch": 0.0,
+      "loss_char": 0.7243,
+      "loss_token": 1.6167,
+      "lr": "2.00e-07",
+      "norm": 5.7057,
+      "step": 136
+    },
+    {
+      "acc_char": 0.322,
+      "acc_token": 0.7129,
+      "epoch": 0.0,
+      "loss_char": 0.6653,
+      "loss_token": 1.5636,
+      "lr": "2.00e-07",
+      "norm": 5.4494,
+      "step": 137
+    },
+    {
+      "acc_char": 0.2882,
+      "acc_token": 0.6662,
+      "epoch": 0.0,
+      "loss_char": 0.7417,
+      "loss_token": 1.6943,
+      "lr": "2.00e-07",
+      "norm": 5.88,
+      "step": 138
+    },
+    {
+      "acc_char": 0.2908,
+      "acc_token": 0.6602,
+      "epoch": 0.0,
+      "loss_char": 0.7742,
+      "loss_token": 1.7059,
+      "lr": "2.00e-07",
+      "norm": 5.6887,
+      "step": 139
+    },
+    {
+      "acc_char": 0.2988,
+      "acc_token": 0.6609,
+      "epoch": 0.0,
+      "loss_char": 0.8141,
+      "loss_token": 1.6664,
+      "lr": "2.00e-07",
+      "norm": 5.6956,
+      "step": 140
+    },
+    {
+      "acc_char": 0.3026,
+      "acc_token": 0.6739,
+      "epoch": 0.0,
+      "loss_char": 0.7646,
+      "loss_token": 1.7001,
+      "lr": "2.00e-07",
+      "norm": 5.7097,
+      "step": 141
+    },
+    {
+      "acc_char": 0.2945,
+      "acc_token": 0.6707,
+      "epoch": 0.0,
+      "loss_char": 0.7792,
+      "loss_token": 1.7446,
+      "lr": "2.00e-07",
+      "norm": 6.1457,
+      "step": 142
+    },
+    {
+      "acc_char": 0.298,
+      "acc_token": 0.6815,
+      "epoch": 0.0,
+      "loss_char": 0.7148,
+      "loss_token": 1.6377,
+      "lr": "2.00e-07",
+      "norm": 5.5933,
+      "step": 143
+    },
+    {
+      "acc_char": 0.3128,
+      "acc_token": 0.714,
+      "epoch": 0.0,
+      "loss_char": 0.6391,
+      "loss_token": 1.5338,
+      "lr": "2.00e-07",
+      "norm": 5.5648,
+      "step": 144
+    },
+    {
+      "acc_char": 0.3061,
+      "acc_token": 0.6973,
+      "epoch": 0.0,
+      "loss_char": 0.6774,
+      "loss_token": 1.5908,
+      "lr": "2.00e-07",
+      "norm": 5.6272,
+      "step": 145
+    },
+    {
+      "acc_char": 0.3008,
+      "acc_token": 0.6728,
+      "epoch": 0.0,
+      "loss_char": 0.7721,
+      "loss_token": 1.6977,
+      "lr": "2.00e-07",
+      "norm": 6.0004,
+      "step": 146
+    },
+    {
+      "acc_char": 0.2837,
+      "acc_token": 0.6515,
+      "epoch": 0.0,
+      "loss_char": 0.8178,
+      "loss_token": 1.7521,
+      "lr": "2.00e-07",
+      "norm": 5.9622,
+      "step": 147
+    },
+    {
+      "acc_char": 0.2911,
+      "acc_token": 0.6604,
+      "epoch": 0.0,
+      "loss_char": 0.8271,
+      "loss_token": 1.6787,
+      "lr": "2.00e-07",
+      "norm": 5.9203,
+      "step": 148
+    },
+    {
+      "acc_char": 0.2951,
+      "acc_token": 0.6641,
+      "epoch": 0.0,
+      "loss_char": 0.7608,
+      "loss_token": 1.7659,
+      "lr": "2.00e-07",
+      "norm": 5.8751,
+      "step": 149
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6759,
+      "epoch": 0.0,
+      "loss_char": 0.7553,
+      "loss_token": 1.6277,
+      "lr": "2.00e-07",
+      "norm": 5.7304,
+      "step": 150
+    },
+    {
+      "acc_char": 0.3175,
+      "acc_token": 0.7099,
+      "epoch": 0.0,
+      "loss_char": 0.6529,
+      "loss_token": 1.5657,
+      "lr": "2.00e-07",
+      "norm": 5.6377,
+      "step": 151
+    },
+    {
+      "acc_char": 0.3109,
+      "acc_token": 0.6978,
+      "epoch": 0.0,
+      "loss_char": 0.6993,
+      "loss_token": 1.6537,
+      "lr": "2.00e-07",
+      "norm": 5.6356,
+      "step": 152
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6638,
+      "epoch": 0.0,
+      "loss_char": 0.7891,
+      "loss_token": 1.5923,
+      "lr": "2.00e-07",
+      "norm": 5.7508,
+      "step": 153
+    },
+    {
+      "acc_char": 0.3023,
+      "acc_token": 0.7008,
+      "epoch": 0.0,
+      "loss_char": 0.6588,
+      "loss_token": 1.634,
+      "lr": "2.00e-07",
+      "norm": 5.6687,
+      "step": 154
+    },
+    {
+      "acc_char": 0.2921,
+      "acc_token": 0.6568,
+      "epoch": 0.0,
+      "loss_char": 0.818,
+      "loss_token": 1.7285,
+      "lr": "2.00e-07",
+      "norm": 5.7942,
+      "step": 155
+    },
+    {
+      "acc_char": 0.3148,
+      "acc_token": 0.7104,
+      "epoch": 0.0,
+      "loss_char": 0.6577,
+      "loss_token": 1.5699,
+      "lr": "2.00e-07",
+      "norm": 5.7228,
+      "step": 156
+    },
+    {
+      "acc_char": 0.2811,
+      "acc_token": 0.6123,
+      "epoch": 0.0,
+      "loss_char": 0.9048,
+      "loss_token": 1.832,
+      "lr": "2.00e-07",
+      "norm": 5.4958,
+      "step": 157
+    },
+    {
+      "acc_char": 0.325,
+      "acc_token": 0.7288,
+      "epoch": 0.0,
+      "loss_char": 0.6283,
+      "loss_token": 1.5255,
+      "lr": "2.00e-07",
+      "norm": 5.4465,
+      "step": 158
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6789,
+      "epoch": 0.0,
+      "loss_char": 0.7332,
+      "loss_token": 1.5868,
+      "lr": "2.00e-07",
+      "norm": 5.4336,
+      "step": 159
+    },
+    {
+      "acc_char": 0.2975,
+      "acc_token": 0.6689,
+      "epoch": 0.0,
+      "loss_char": 0.7776,
+      "loss_token": 1.7173,
+      "lr": "2.00e-07",
+      "norm": 5.8854,
+      "step": 160
+    },
+    {
+      "acc_char": 0.2975,
+      "acc_token": 0.6693,
+      "epoch": 0.0,
+      "loss_char": 0.7565,
+      "loss_token": 1.7072,
+      "lr": "2.00e-07",
+      "norm": 6.249,
+      "step": 161
+    },
+    {
+      "acc_char": 0.2866,
+      "acc_token": 0.6611,
+      "epoch": 0.0,
+      "loss_char": 0.791,
+      "loss_token": 1.7247,
+      "lr": "2.00e-07",
+      "norm": 5.859,
+      "step": 162
+    },
+    {
+      "acc_char": 0.3141,
+      "acc_token": 0.7247,
+      "epoch": 0.0,
+      "loss_char": 0.5933,
+      "loss_token": 1.5428,
+      "lr": "2.00e-07",
+      "norm": 5.4944,
+      "step": 163
+    },
+    {
+      "acc_char": 0.294,
+      "acc_token": 0.6729,
+      "epoch": 0.0,
+      "loss_char": 0.7727,
+      "loss_token": 1.6252,
+      "lr": "2.00e-07",
+      "norm": 6.101,
+      "step": 164
+    },
+    {
+      "acc_char": 0.2924,
+      "acc_token": 0.6723,
+      "epoch": 0.0,
+      "loss_char": 0.7185,
+      "loss_token": 1.6621,
+      "lr": "2.00e-07",
+      "norm": 5.9544,
+      "step": 165
+    },
+    {
+      "acc_char": 0.3052,
+      "acc_token": 0.689,
+      "epoch": 0.0,
+      "loss_char": 0.7298,
+      "loss_token": 1.5926,
+      "lr": "2.00e-07",
+      "norm": 5.7368,
+      "step": 166
+    },
+    {
+      "acc_char": 0.3097,
+      "acc_token": 0.6985,
+      "epoch": 0.0,
+      "loss_char": 0.6821,
+      "loss_token": 1.6133,
+      "lr": "2.00e-07",
+      "norm": 5.7521,
+      "step": 167
+    },
+    {
+      "acc_char": 0.3013,
+      "acc_token": 0.6768,
+      "epoch": 0.0,
+      "loss_char": 0.7478,
+      "loss_token": 1.6652,
+      "lr": "2.00e-07",
+      "norm": 5.5591,
+      "step": 168
+    },
+    {
+      "acc_char": 0.2904,
+      "acc_token": 0.6626,
+      "epoch": 0.0,
+      "loss_char": 0.7737,
+      "loss_token": 1.6033,
+      "lr": "2.00e-07",
+      "norm": 5.6183,
+      "step": 169
+    },
+    {
+      "acc_char": 0.2984,
+      "acc_token": 0.6925,
+      "epoch": 0.0,
+      "loss_char": 0.6733,
+      "loss_token": 1.5176,
+      "lr": "2.00e-07",
+      "norm": 5.4951,
+      "step": 170
+    },
+    {
+      "acc_char": 0.3039,
+      "acc_token": 0.6942,
+      "epoch": 0.0,
+      "loss_char": 0.6754,
+      "loss_token": 1.6154,
+      "lr": "2.00e-07",
+      "norm": 5.7609,
+      "step": 171
+    },
+    {
+      "acc_char": 0.3067,
+      "acc_token": 0.6891,
+      "epoch": 0.0,
+      "loss_char": 0.7122,
+      "loss_token": 1.6741,
+      "lr": "2.00e-07",
+      "norm": 5.5804,
+      "step": 172
+    },
+    {
+      "acc_char": 0.2927,
+      "acc_token": 0.6688,
+      "epoch": 0.0,
+      "loss_char": 0.7944,
+      "loss_token": 1.7406,
+      "lr": "2.00e-07",
+      "norm": 5.8865,
+      "step": 173
+    },
+    {
+      "acc_char": 0.2974,
+      "acc_token": 0.6814,
+      "epoch": 0.0,
+      "loss_char": 0.7164,
+      "loss_token": 1.6779,
+      "lr": "2.00e-07",
+      "norm": 5.9016,
+      "step": 174
+    },
+    {
+      "acc_char": 0.3147,
+      "acc_token": 0.7079,
+      "epoch": 0.0,
+      "loss_char": 0.6844,
+      "loss_token": 1.6573,
+      "lr": "2.00e-07",
+      "norm": 5.9278,
+      "step": 175
+    },
+    {
+      "acc_char": 0.3126,
+      "acc_token": 0.7029,
+      "epoch": 0.0,
+      "loss_char": 0.6736,
+      "loss_token": 1.615,
+      "lr": "2.00e-07",
+      "norm": 5.5758,
+      "step": 176
+    },
+    {
+      "acc_char": 0.2953,
+      "acc_token": 0.6622,
+      "epoch": 0.0,
+      "loss_char": 0.7834,
+      "loss_token": 1.7336,
+      "lr": "2.00e-07",
+      "norm": 5.8171,
+      "step": 177
+    },
+    {
+      "acc_char": 0.2839,
+      "acc_token": 0.6287,
+      "epoch": 0.0,
+      "loss_char": 0.8726,
+      "loss_token": 1.7605,
+      "lr": "2.00e-07",
+      "norm": 5.4992,
+      "step": 178
+    },
+    {
+      "acc_char": 0.2985,
+      "acc_token": 0.6735,
+      "epoch": 0.0,
+      "loss_char": 0.7581,
+      "loss_token": 1.6939,
+      "lr": "2.00e-07",
+      "norm": 5.8877,
+      "step": 179
+    },
+    {
+      "acc_char": 0.3101,
+      "acc_token": 0.6953,
+      "epoch": 0.0,
+      "loss_char": 0.6897,
+      "loss_token": 1.5523,
+      "lr": "2.00e-07",
+      "norm": 5.3722,
+      "step": 180
+    },
+    {
+      "acc_char": 0.3193,
+      "acc_token": 0.6913,
+      "epoch": 0.0,
+      "loss_char": 0.6644,
+      "loss_token": 1.6021,
+      "lr": "2.00e-07",
+      "norm": 5.3646,
+      "step": 181
+    },
+    {
+      "acc_char": 0.3162,
+      "acc_token": 0.7104,
+      "epoch": 0.0,
+      "loss_char": 0.6554,
+      "loss_token": 1.6126,
+      "lr": "2.00e-07",
+      "norm": 5.4689,
+      "step": 182
+    },
+    {
+      "acc_char": 0.2991,
+      "acc_token": 0.6559,
+      "epoch": 0.0,
+      "loss_char": 0.7793,
+      "loss_token": 1.7622,
+      "lr": "2.00e-07",
+      "norm": 5.4834,
+      "step": 183
+    },
+    {
+      "acc_char": 0.3046,
+      "acc_token": 0.6817,
+      "epoch": 0.0,
+      "loss_char": 0.7293,
+      "loss_token": 1.6286,
+      "lr": "2.00e-07",
+      "norm": 5.63,
+      "step": 184
+    },
+    {
+      "acc_char": 0.2967,
+      "acc_token": 0.6837,
+      "epoch": 0.0,
+      "loss_char": 0.6733,
+      "loss_token": 1.6232,
+      "lr": "2.00e-07",
+      "norm": 5.4997,
+      "step": 185
+    },
+    {
+      "acc_char": 0.3174,
+      "acc_token": 0.7007,
+      "epoch": 0.0,
+      "loss_char": 0.6977,
+      "loss_token": 1.5801,
+      "lr": "2.00e-07",
+      "norm": 5.533,
+      "step": 186
+    },
+    {
+      "acc_char": 0.2927,
+      "acc_token": 0.6678,
+      "epoch": 0.0,
+      "loss_char": 0.7497,
+      "loss_token": 1.6652,
+      "lr": "2.00e-07",
+      "norm": 5.7318,
+      "step": 187
+    },
+    {
+      "acc_char": 0.2958,
+      "acc_token": 0.6793,
+      "epoch": 0.0,
+      "loss_char": 0.7117,
+      "loss_token": 1.6247,
+      "lr": "2.00e-07",
+      "norm": 5.558,
+      "step": 188
+    },
+    {
+      "acc_char": 0.299,
+      "acc_token": 0.6745,
+      "epoch": 0.0,
+      "loss_char": 0.7525,
+      "loss_token": 1.6932,
+      "lr": "2.00e-07",
+      "norm": 5.8052,
+      "step": 189
+    },
+    {
+      "acc_char": 0.3185,
+      "acc_token": 0.7051,
+      "epoch": 0.0,
+      "loss_char": 0.6852,
+      "loss_token": 1.5308,
+      "lr": "2.00e-07",
+      "norm": 5.7116,
+      "step": 190
+    },
+    {
+      "acc_char": 0.2858,
+      "acc_token": 0.6439,
+      "epoch": 0.0,
+      "loss_char": 0.8477,
+      "loss_token": 1.7375,
+      "lr": "2.00e-07",
+      "norm": 5.8329,
+      "step": 191
+    },
+    {
+      "acc_char": 0.2996,
+      "acc_token": 0.6782,
+      "epoch": 0.0,
+      "loss_char": 0.7316,
+      "loss_token": 1.6276,
+      "lr": "2.00e-07",
+      "norm": 5.6723,
+      "step": 192
+    },
+    {
+      "acc_char": 0.296,
+      "acc_token": 0.6798,
+      "epoch": 0.0,
+      "loss_char": 0.7162,
+      "loss_token": 1.6456,
+      "lr": "2.00e-07",
+      "norm": 5.5331,
+      "step": 193
+    },
+    {
+      "acc_char": 0.3012,
+      "acc_token": 0.6737,
+      "epoch": 0.0,
+      "loss_char": 0.753,
+      "loss_token": 1.6082,
+      "lr": "2.00e-07",
+      "norm": 5.8935,
+      "step": 194
+    },
+    {
+      "acc_char": 0.2792,
+      "acc_token": 0.6171,
+      "epoch": 0.0,
+      "loss_char": 0.8842,
+      "loss_token": 1.8427,
+      "lr": "2.00e-07",
+      "norm": 5.3946,
+      "step": 195
+    },
+    {
+      "acc_char": 0.2965,
+      "acc_token": 0.6732,
+      "epoch": 0.0,
+      "loss_char": 0.7525,
+      "loss_token": 1.6317,
+      "lr": "2.00e-07",
+      "norm": 5.5915,
+      "step": 196
+    },
+    {
+      "acc_char": 0.296,
+      "acc_token": 0.6791,
+      "epoch": 0.0,
+      "loss_char": 0.7122,
+      "loss_token": 1.6854,
+      "lr": "2.00e-07",
+      "norm": 5.656,
+      "step": 197
+    },
+    {
+      "acc_char": 0.3108,
+      "acc_token": 0.6935,
+      "epoch": 0.0,
+      "loss_char": 0.7203,
+      "loss_token": 1.6353,
+      "lr": "2.00e-07",
+      "norm": 5.8489,
+      "step": 198
+    },
+    {
+      "acc_char": 0.3209,
+      "acc_token": 0.6995,
+      "epoch": 0.0,
+      "loss_char": 0.6958,
+      "loss_token": 1.5762,
+      "lr": "2.00e-07",
+      "norm": 5.6491,
+      "step": 199
+    },
+    {
+      "acc_char": 0.2956,
+      "acc_token": 0.6688,
+      "epoch": 0.0,
+      "loss_char": 0.7339,
+      "loss_token": 1.6059,
+      "lr": "2.00e-07",
+      "norm": 5.7993,
+      "step": 200
+    },
+    {
+      "acc_char": 0.3025,
+      "acc_token": 0.6711,
+      "epoch": 0.0,
+      "loss_char": 0.7747,
+      "loss_token": 1.6607,
+      "lr": "2.00e-07",
+      "norm": 5.8025,
+      "step": 201
+    },
+    {
+      "acc_char": 0.3159,
+      "acc_token": 0.7115,
+      "epoch": 0.0,
+      "loss_char": 0.6561,
+      "loss_token": 1.5915,
+      "lr": "2.00e-07",
+      "norm": 5.5739,
+      "step": 202
+    },
+    {
+      "acc_char": 0.314,
+      "acc_token": 0.6954,
+      "epoch": 0.0,
+      "loss_char": 0.7071,
+      "loss_token": 1.5973,
+      "lr": "2.00e-07",
+      "norm": 5.7454,
+      "step": 203
+    },
+    {
+      "acc_char": 0.2964,
+      "acc_token": 0.6685,
+      "epoch": 0.0,
+      "loss_char": 0.7321,
+      "loss_token": 1.6484,
+      "lr": "2.00e-07",
+      "norm": 6.565,
+      "step": 204
+    },
+    {
+      "acc_char": 0.2827,
+      "acc_token": 0.6502,
+      "epoch": 0.0,
+      "loss_char": 0.7909,
+      "loss_token": 1.7053,
+      "lr": "2.00e-07",
+      "norm": 5.8039,
+      "step": 205
+    },
+    {
+      "acc_char": 0.3067,
+      "acc_token": 0.7046,
+      "epoch": 0.0,
+      "loss_char": 0.6551,
+      "loss_token": 1.5926,
+      "lr": "2.00e-07",
+      "norm": 5.8913,
+      "step": 206
+    },
+    {
+      "acc_char": 0.3069,
+      "acc_token": 0.6929,
+      "epoch": 0.0,
+      "loss_char": 0.6949,
+      "loss_token": 1.575,
+      "lr": "2.00e-07",
+      "norm": 5.4935,
+      "step": 207
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6792,
+      "epoch": 0.0,
+      "loss_char": 0.7197,
+      "loss_token": 1.6739,
+      "lr": "2.00e-07",
+      "norm": 5.7478,
+      "step": 208
+    },
+    {
+      "acc_char": 0.3042,
+      "acc_token": 0.6711,
+      "epoch": 0.0,
+      "loss_char": 0.7623,
+      "loss_token": 1.6577,
+      "lr": "2.00e-07",
+      "norm": 5.6784,
+      "step": 209
+    },
+    {
+      "acc_char": 0.2935,
+      "acc_token": 0.6707,
+      "epoch": 0.0,
+      "loss_char": 0.7702,
+      "loss_token": 1.6913,
+      "lr": "2.00e-07",
+      "norm": 5.8799,
+      "step": 210
+    },
+    {
+      "acc_char": 0.3207,
+      "acc_token": 0.7217,
+      "epoch": 0.0,
+      "loss_char": 0.6361,
+      "loss_token": 1.5228,
+      "lr": "2.00e-07",
+      "norm": 5.6157,
+      "step": 211
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.6767,
+      "epoch": 0.0,
+      "loss_char": 0.7229,
+      "loss_token": 1.5723,
+      "lr": "2.00e-07",
+      "norm": 5.5983,
+      "step": 212
+    },
+    {
+      "acc_char": 0.3048,
+      "acc_token": 0.6875,
+      "epoch": 0.0,
+      "loss_char": 0.7076,
+      "loss_token": 1.6463,
+      "lr": "2.00e-07",
+      "norm": 5.891,
+      "step": 213
+    },
+    {
+      "acc_char": 0.3108,
+      "acc_token": 0.7008,
+      "epoch": 0.0,
+      "loss_char": 0.6808,
+      "loss_token": 1.6064,
+      "lr": "2.00e-07",
+      "norm": 5.5738,
+      "step": 214
+    },
+    {
+      "acc_char": 0.3038,
+      "acc_token": 0.684,
+      "epoch": 0.0,
+      "loss_char": 0.7254,
+      "loss_token": 1.6527,
+      "lr": "2.00e-07",
+      "norm": 5.8218,
+      "step": 215
+    },
+    {
+      "acc_char": 0.2914,
+      "acc_token": 0.6668,
+      "epoch": 0.0,
+      "loss_char": 0.7661,
+      "loss_token": 1.6431,
+      "lr": "2.00e-07",
+      "norm": 5.4766,
+      "step": 216
+    },
+    {
+      "acc_char": 0.3127,
+      "acc_token": 0.703,
+      "epoch": 0.0,
+      "loss_char": 0.6885,
+      "loss_token": 1.6306,
+      "lr": "2.00e-07",
+      "norm": 5.6474,
+      "step": 217
+    },
+    {
+      "acc_char": 0.2957,
+      "acc_token": 0.662,
+      "epoch": 0.0,
+      "loss_char": 0.7833,
+      "loss_token": 1.6566,
+      "lr": "2.00e-07",
+      "norm": 7.2402,
+      "step": 218
+    },
+    {
+      "acc_char": 0.3161,
+      "acc_token": 0.6974,
+      "epoch": 0.0,
+      "loss_char": 0.6961,
+      "loss_token": 1.5986,
+      "lr": "2.00e-07",
+      "norm": 6.046,
+      "step": 219
+    },
+    {
+      "acc_char": 0.2773,
+      "acc_token": 0.6116,
+      "epoch": 0.0,
+      "loss_char": 0.8903,
+      "loss_token": 1.7691,
+      "lr": "2.00e-07",
+      "norm": 5.8258,
+      "step": 220
+    },
+    {
+      "acc_char": 0.3011,
+      "acc_token": 0.6851,
+      "epoch": 0.0,
+      "loss_char": 0.7189,
+      "loss_token": 1.6715,
+      "lr": "2.00e-07",
+      "norm": 5.7991,
+      "step": 221
+    },
+    {
+      "acc_char": 0.2985,
+      "acc_token": 0.6706,
+      "epoch": 0.0,
+      "loss_char": 0.7537,
+      "loss_token": 1.657,
+      "lr": "2.00e-07",
+      "norm": 5.9313,
+      "step": 222
+    },
+    {
+      "acc_char": 0.2766,
+      "acc_token": 0.6166,
+      "epoch": 0.0,
+      "loss_char": 0.8881,
+      "loss_token": 1.8768,
+      "lr": "2.00e-07",
+      "norm": 5.6931,
+      "step": 223
+    },
+    {
+      "acc_char": 0.2935,
+      "acc_token": 0.6561,
+      "epoch": 0.0,
+      "loss_char": 0.7996,
+      "loss_token": 1.6949,
+      "lr": "2.00e-07",
+      "norm": 6.053,
+      "step": 224
+    },
+    {
+      "acc_char": 0.305,
+      "acc_token": 0.6896,
+      "epoch": 0.0,
+      "loss_char": 0.7181,
+      "loss_token": 1.6583,
+      "lr": "2.00e-07",
+      "norm": 5.6742,
+      "step": 225
+    },
+    {
+      "acc_char": 0.2892,
+      "acc_token": 0.6651,
+      "epoch": 0.0,
+      "loss_char": 0.7748,
+      "loss_token": 1.681,
+      "lr": "2.00e-07",
+      "norm": 5.7138,
+      "step": 226
+    },
+    {
+      "acc_char": 0.282,
+      "acc_token": 0.6174,
+      "epoch": 0.0,
+      "loss_char": 0.9016,
+      "loss_token": 1.7903,
+      "lr": "2.00e-07",
+      "norm": 5.938,
+      "step": 227
+    },
+    {
+      "acc_char": 0.2972,
+      "acc_token": 0.6766,
+      "epoch": 0.0,
+      "loss_char": 0.7571,
+      "loss_token": 1.7066,
+      "lr": "2.00e-07",
+      "norm": 5.8829,
+      "step": 228
+    },
+    {
+      "acc_char": 0.3053,
+      "acc_token": 0.6894,
+      "epoch": 0.0,
+      "loss_char": 0.7142,
+      "loss_token": 1.6113,
+      "lr": "2.00e-07",
+      "norm": 5.8733,
+      "step": 229
+    },
+    {
+      "acc_char": 0.2838,
+      "acc_token": 0.656,
+      "epoch": 0.0,
+      "loss_char": 0.7629,
+      "loss_token": 1.7732,
+      "lr": "2.00e-07",
+      "norm": 6.0202,
+      "step": 230
+    },
+    {
+      "acc_char": 0.3108,
+      "acc_token": 0.7016,
+      "epoch": 0.0,
+      "loss_char": 0.6778,
+      "loss_token": 1.6731,
+      "lr": "2.00e-07",
+      "norm": 5.672,
+      "step": 231
+    },
+    {
+      "acc_char": 0.308,
+      "acc_token": 0.703,
+      "epoch": 0.0,
+      "loss_char": 0.6667,
+      "loss_token": 1.5779,
+      "lr": "2.00e-07",
+      "norm": 5.6096,
+      "step": 232
+    },
+    {
+      "acc_char": 0.304,
+      "acc_token": 0.6848,
+      "epoch": 0.0,
+      "loss_char": 0.7392,
+      "loss_token": 1.6087,
+      "lr": "2.00e-07",
+      "norm": 5.634,
+      "step": 233
+    },
+    {
+      "acc_char": 0.3052,
+      "acc_token": 0.689,
+      "epoch": 0.0,
+      "loss_char": 0.6903,
+      "loss_token": 1.6823,
+      "lr": "2.00e-07",
+      "norm": 5.6758,
+      "step": 234
+    },
+    {
+      "acc_char": 0.3041,
+      "acc_token": 0.6942,
+      "epoch": 0.0,
+      "loss_char": 0.6867,
+      "loss_token": 1.5694,
+      "lr": "2.00e-07",
+      "norm": 5.8105,
+      "step": 235
+    },
+    {
+      "acc_char": 0.2992,
+      "acc_token": 0.6725,
+      "epoch": 0.0,
+      "loss_char": 0.7675,
+      "loss_token": 1.8046,
+      "lr": "2.00e-07",
+      "norm": 5.8105,
+      "step": 236
+    },
+    {
+      "acc_char": 0.3097,
+      "acc_token": 0.6881,
+      "epoch": 0.0,
+      "loss_char": 0.7139,
+      "loss_token": 1.6435,
+      "lr": "2.00e-07",
+      "norm": 5.7911,
+      "step": 237
+    },
+    {
+      "acc_char": 0.2896,
+      "acc_token": 0.6658,
+      "epoch": 0.0,
+      "loss_char": 0.7611,
+      "loss_token": 1.6792,
+      "lr": "2.00e-07",
+      "norm": 5.6335,
+      "step": 238
+    },
+    {
+      "acc_char": 0.3009,
+      "acc_token": 0.6747,
+      "epoch": 0.0,
+      "loss_char": 0.7615,
+      "loss_token": 1.6733,
+      "lr": "2.00e-07",
+      "norm": 5.9079,
+      "step": 239
+    },
+    {
+      "acc_char": 0.302,
+      "acc_token": 0.6832,
+      "epoch": 0.0,
+      "loss_char": 0.7202,
+      "loss_token": 1.6677,
+      "lr": "2.00e-07",
+      "norm": 5.9424,
+      "step": 240
+    },
+    {
+      "acc_char": 0.302,
+      "acc_token": 0.6742,
+      "epoch": 0.0,
+      "loss_char": 0.7599,
+      "loss_token": 1.6017,
+      "lr": "2.00e-07",
+      "norm": 6.373,
+      "step": 241
+    },
+    {
+      "acc_char": 0.3049,
+      "acc_token": 0.7066,
+      "epoch": 0.0,
+      "loss_char": 0.6629,
+      "loss_token": 1.6348,
+      "lr": "2.00e-07",
+      "norm": 5.6642,
+      "step": 242
+    },
+    {
+      "acc_char": 0.291,
+      "acc_token": 0.6323,
+      "epoch": 0.0,
+      "loss_char": 0.8309,
+      "loss_token": 1.6843,
+      "lr": "2.00e-07",
+      "norm": 7.0208,
+      "step": 243
+    },
+    {
+      "acc_char": 0.2987,
+      "acc_token": 0.6944,
+      "epoch": 0.0,
+      "loss_char": 0.6775,
+      "loss_token": 1.6127,
+      "lr": "2.00e-07",
+      "norm": 5.8339,
+      "step": 244
+    },
+    {
+      "acc_char": 0.2927,
+      "acc_token": 0.6734,
+      "epoch": 0.0,
+      "loss_char": 0.7555,
+      "loss_token": 1.7228,
+      "lr": "2.00e-07",
+      "norm": 5.8693,
+      "step": 245
+    },
+    {
+      "acc_char": 0.3026,
+      "acc_token": 0.6829,
+      "epoch": 0.0,
+      "loss_char": 0.7342,
+      "loss_token": 1.6397,
+      "lr": "2.00e-07",
+      "norm": 5.6664,
+      "step": 246
+    },
+    {
+      "acc_char": 0.3078,
+      "acc_token": 0.6956,
+      "epoch": 0.0,
+      "loss_char": 0.7104,
+      "loss_token": 1.713,
+      "lr": "2.00e-07",
+      "norm": 6.3569,
+      "step": 247
+    },
+    {
+      "acc_char": 0.2887,
+      "acc_token": 0.6516,
+      "epoch": 0.0,
+      "loss_char": 0.8043,
+      "loss_token": 1.7693,
+      "lr": "2.00e-07",
+      "norm": 6.1627,
+      "step": 248
+    },
+    {
+      "acc_char": 0.315,
+      "acc_token": 0.7136,
+      "epoch": 0.0,
+      "loss_char": 0.6514,
+      "loss_token": 1.5165,
+      "lr": "2.00e-07",
+      "norm": 5.6597,
+      "step": 249
+    },
+    {
+      "acc_char": 0.3168,
+      "acc_token": 0.7127,
+      "epoch": 0.0,
+      "loss_char": 0.6437,
+      "loss_token": 1.5805,
+      "lr": "2.00e-07",
+      "norm": 5.6713,
+      "step": 250
+    },
+    {
+      "acc_char": 0.3089,
+      "acc_token": 0.7065,
+      "epoch": 0.0,
+      "loss_char": 0.6508,
+      "loss_token": 1.5944,
+      "lr": "2.00e-07",
+      "norm": 5.5517,
+      "step": 251
+    },
+    {
+      "acc_char": 0.3234,
+      "acc_token": 0.7193,
+      "epoch": 0.0,
+      "loss_char": 0.6507,
+      "loss_token": 1.5453,
+      "lr": "2.00e-07",
+      "norm": 5.8191,
+      "step": 252
+    },
+    {
+      "acc_char": 0.2985,
+      "acc_token": 0.6924,
+      "epoch": 0.0,
+      "loss_char": 0.6841,
+      "loss_token": 1.6334,
+      "lr": "2.00e-07",
+      "norm": 5.5597,
+      "step": 253
+    },
+    {
+      "acc_char": 0.3017,
+      "acc_token": 0.6705,
+      "epoch": 0.0,
+      "loss_char": 0.7677,
+      "loss_token": 1.5948,
+      "lr": "2.00e-07",
+      "norm": 6.1439,
+      "step": 254
+    },
+    {
+      "acc_char": 0.3057,
+      "acc_token": 0.6895,
+      "epoch": 0.0,
+      "loss_char": 0.7055,
+      "loss_token": 1.5337,
+      "lr": "2.00e-07",
+      "norm": 5.4335,
+      "step": 255
+    },
+    {
+      "acc_char": 0.2927,
+      "acc_token": 0.6691,
+      "epoch": 0.0,
+      "loss_char": 0.7783,
+      "loss_token": 1.6823,
+      "lr": "2.00e-07",
+      "norm": 5.8874,
+      "step": 256
+    },
+    {
+      "acc_char": 0.3109,
+      "acc_token": 0.6977,
+      "epoch": 0.0,
+      "loss_char": 0.7025,
+      "loss_token": 1.522,
+      "lr": "2.00e-07",
+      "norm": 5.4743,
+      "step": 257
+    },
+    {
+      "acc_char": 0.2988,
+      "acc_token": 0.6697,
+      "epoch": 0.0,
+      "loss_char": 0.7659,
+      "loss_token": 1.6629,
+      "lr": "2.00e-07",
+      "norm": 5.773,
+      "step": 258
+    },
+    {
+      "acc_char": 0.3034,
+      "acc_token": 0.6779,
+      "epoch": 0.0,
+      "loss_char": 0.749,
+      "loss_token": 1.5817,
+      "lr": "2.00e-07",
+      "norm": 5.7499,
+      "step": 259
+    },
+    {
+      "acc_char": 0.3071,
+      "acc_token": 0.7046,
+      "epoch": 0.0,
+      "loss_char": 0.6483,
+      "loss_token": 1.5455,
+      "lr": "2.00e-07",
+      "norm": 5.4589,
+      "step": 260
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6709,
+      "epoch": 0.0,
+      "loss_char": 0.7631,
+      "loss_token": 1.7526,
+      "lr": "2.00e-07",
+      "norm": 5.8391,
+      "step": 261
+    },
+    {
+      "acc_char": 0.2992,
+      "acc_token": 0.6824,
+      "epoch": 0.0,
+      "loss_char": 0.7006,
+      "loss_token": 1.6149,
+      "lr": "2.00e-07",
+      "norm": 5.4798,
+      "step": 262
+    },
+    {
+      "acc_char": 0.2893,
+      "acc_token": 0.673,
+      "epoch": 0.0,
+      "loss_char": 0.7407,
+      "loss_token": 1.6473,
+      "lr": "2.00e-07",
+      "norm": 6.0615,
+      "step": 263
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6681,
+      "epoch": 0.0,
+      "loss_char": 0.7942,
+      "loss_token": 1.6474,
+      "lr": "2.00e-07",
+      "norm": 6.1056,
+      "step": 264
+    },
+    {
+      "acc_char": 0.3208,
+      "acc_token": 0.7085,
+      "epoch": 0.0,
+      "loss_char": 0.6816,
+      "loss_token": 1.6371,
+      "lr": "2.00e-07",
+      "norm": 5.5811,
+      "step": 265
+    },
+    {
+      "acc_char": 0.2961,
+      "acc_token": 0.6715,
+      "epoch": 0.0,
+      "loss_char": 0.736,
+      "loss_token": 1.7145,
+      "lr": "2.00e-07",
+      "norm": 5.7918,
+      "step": 266
+    },
+    {
+      "acc_char": 0.3118,
+      "acc_token": 0.6992,
+      "epoch": 0.0,
+      "loss_char": 0.6895,
+      "loss_token": 1.5566,
+      "lr": "2.00e-07",
+      "norm": 5.6788,
+      "step": 267
+    },
+    {
+      "acc_char": 0.3006,
+      "acc_token": 0.6892,
+      "epoch": 0.0,
+      "loss_char": 0.7024,
+      "loss_token": 1.6323,
+      "lr": "2.00e-07",
+      "norm": 5.6619,
+      "step": 268
+    },
+    {
+      "acc_char": 0.3243,
+      "acc_token": 0.7098,
+      "epoch": 0.0,
+      "loss_char": 0.6818,
+      "loss_token": 1.3966,
+      "lr": "2.00e-07",
+      "norm": 5.6444,
+      "step": 269
+    },
+    {
+      "acc_char": 0.2971,
+      "acc_token": 0.6789,
+      "epoch": 0.0,
+      "loss_char": 0.7399,
+      "loss_token": 1.6553,
+      "lr": "2.00e-07",
+      "norm": 5.9755,
+      "step": 270
+    },
+    {
+      "acc_char": 0.3015,
+      "acc_token": 0.687,
+      "epoch": 0.0,
+      "loss_char": 0.7078,
+      "loss_token": 1.6638,
+      "lr": "2.00e-07",
+      "norm": 5.673,
+      "step": 271
+    },
+    {
+      "acc_char": 0.3144,
+      "acc_token": 0.7131,
+      "epoch": 0.0,
+      "loss_char": 0.6484,
+      "loss_token": 1.5826,
+      "lr": "2.00e-07",
+      "norm": 5.4967,
+      "step": 272
+    },
+    {
+      "acc_char": 0.2987,
+      "acc_token": 0.6729,
+      "epoch": 0.0,
+      "loss_char": 0.744,
+      "loss_token": 1.6038,
+      "lr": "2.00e-07",
+      "norm": 6.0506,
+      "step": 273
+    },
+    {
+      "acc_char": 0.3145,
+      "acc_token": 0.7126,
+      "epoch": 0.01,
+      "loss_char": 0.6452,
+      "loss_token": 1.5216,
+      "lr": "2.00e-07",
+      "norm": 5.404,
+      "step": 274
+    },
+    {
+      "acc_char": 0.3076,
+      "acc_token": 0.696,
+      "epoch": 0.01,
+      "loss_char": 0.6986,
+      "loss_token": 1.5858,
+      "lr": "2.00e-07",
+      "norm": 5.7001,
+      "step": 275
+    },
+    {
+      "acc_char": 0.2755,
+      "acc_token": 0.6011,
+      "epoch": 0.01,
+      "loss_char": 0.9312,
+      "loss_token": 1.6963,
+      "lr": "2.00e-07",
+      "norm": 5.6149,
+      "step": 276
+    },
+    {
+      "acc_char": 0.3184,
+      "acc_token": 0.7168,
+      "epoch": 0.01,
+      "loss_char": 0.6472,
+      "loss_token": 1.5169,
+      "lr": "2.00e-07",
+      "norm": 5.729,
+      "step": 277
+    },
+    {
+      "acc_char": 0.2988,
+      "acc_token": 0.6767,
+      "epoch": 0.01,
+      "loss_char": 0.7273,
+      "loss_token": 1.6534,
+      "lr": "2.00e-07",
+      "norm": 5.7457,
+      "step": 278
+    },
+    {
+      "acc_char": 0.3217,
+      "acc_token": 0.7216,
+      "epoch": 0.01,
+      "loss_char": 0.6424,
+      "loss_token": 1.632,
+      "lr": "2.00e-07",
+      "norm": 5.6512,
+      "step": 279
+    },
+    {
+      "acc_char": 0.3261,
+      "acc_token": 0.7308,
+      "epoch": 0.01,
+      "loss_char": 0.631,
+      "loss_token": 1.5274,
+      "lr": "2.00e-07",
+      "norm": 5.8094,
+      "step": 280
+    },
+    {
+      "acc_char": 0.3047,
+      "acc_token": 0.6862,
+      "epoch": 0.01,
+      "loss_char": 0.7273,
+      "loss_token": 1.6198,
+      "lr": "2.00e-07",
+      "norm": 5.5336,
+      "step": 281
+    },
+    {
+      "acc_char": 0.306,
+      "acc_token": 0.6835,
+      "epoch": 0.01,
+      "loss_char": 0.7451,
+      "loss_token": 1.6363,
+      "lr": "2.00e-07",
+      "norm": 5.6988,
+      "step": 282
+    },
+    {
+      "acc_char": 0.3034,
+      "acc_token": 0.6924,
+      "epoch": 0.01,
+      "loss_char": 0.6886,
+      "loss_token": 1.5916,
+      "lr": "2.00e-07",
+      "norm": 5.468,
+      "step": 283
+    },
+    {
+      "acc_char": 0.3107,
+      "acc_token": 0.692,
+      "epoch": 0.01,
+      "loss_char": 0.7125,
+      "loss_token": 1.6719,
+      "lr": "2.00e-07",
+      "norm": 5.9679,
+      "step": 284
+    },
+    {
+      "acc_char": 0.2917,
+      "acc_token": 0.6532,
+      "epoch": 0.01,
+      "loss_char": 0.8194,
+      "loss_token": 1.7125,
+      "lr": "2.00e-07",
+      "norm": 6.0751,
+      "step": 285
+    },
+    {
+      "acc_char": 0.3283,
+      "acc_token": 0.7289,
+      "epoch": 0.01,
+      "loss_char": 0.6037,
+      "loss_token": 1.4546,
+      "lr": "2.00e-07",
+      "norm": 5.5938,
+      "step": 286
+    },
+    {
+      "acc_char": 0.3191,
+      "acc_token": 0.7103,
+      "epoch": 0.01,
+      "loss_char": 0.6685,
+      "loss_token": 1.5951,
+      "lr": "2.00e-07",
+      "norm": 5.5813,
+      "step": 287
+    },
+    {
+      "acc_char": 0.2991,
+      "acc_token": 0.6766,
+      "epoch": 0.01,
+      "loss_char": 0.7287,
+      "loss_token": 1.6345,
+      "lr": "2.00e-07",
+      "norm": 5.9596,
+      "step": 288
+    },
+    {
+      "acc_char": 0.2962,
+      "acc_token": 0.6717,
+      "epoch": 0.01,
+      "loss_char": 0.7433,
+      "loss_token": 1.6893,
+      "lr": "2.00e-07",
+      "norm": 5.7588,
+      "step": 289
+    },
+    {
+      "acc_char": 0.3013,
+      "acc_token": 0.6765,
+      "epoch": 0.01,
+      "loss_char": 0.7459,
+      "loss_token": 1.6535,
+      "lr": "2.00e-07",
+      "norm": 5.8478,
+      "step": 290
+    },
+    {
+      "acc_char": 0.3181,
+      "acc_token": 0.7165,
+      "epoch": 0.01,
+      "loss_char": 0.64,
+      "loss_token": 1.5437,
+      "lr": "2.00e-07",
+      "norm": 5.387,
+      "step": 291
+    },
+    {
+      "acc_char": 0.2807,
+      "acc_token": 0.6105,
+      "epoch": 0.01,
+      "loss_char": 0.8958,
+      "loss_token": 1.8091,
+      "lr": "2.00e-07",
+      "norm": 5.8084,
+      "step": 292
+    },
+    {
+      "acc_char": 0.3056,
+      "acc_token": 0.6883,
+      "epoch": 0.01,
+      "loss_char": 0.6985,
+      "loss_token": 1.5708,
+      "lr": "2.00e-07",
+      "norm": 5.5178,
+      "step": 293
+    },
+    {
+      "acc_char": 0.3006,
+      "acc_token": 0.6808,
+      "epoch": 0.01,
+      "loss_char": 0.7287,
+      "loss_token": 1.6012,
+      "lr": "2.00e-07",
+      "norm": 5.6108,
+      "step": 294
+    },
+    {
+      "acc_char": 0.2975,
+      "acc_token": 0.671,
+      "epoch": 0.01,
+      "loss_char": 0.7498,
+      "loss_token": 1.7116,
+      "lr": "2.00e-07",
+      "norm": 5.4779,
+      "step": 295
+    },
+    {
+      "acc_char": 0.3121,
+      "acc_token": 0.7031,
+      "epoch": 0.01,
+      "loss_char": 0.6763,
+      "loss_token": 1.5036,
+      "lr": "2.00e-07",
+      "norm": 5.49,
+      "step": 296
+    },
+    {
+      "acc_char": 0.3,
+      "acc_token": 0.6808,
+      "epoch": 0.01,
+      "loss_char": 0.7139,
+      "loss_token": 1.6118,
+      "lr": "2.00e-07",
+      "norm": 5.3666,
+      "step": 297
+    },
+    {
+      "acc_char": 0.2984,
+      "acc_token": 0.6774,
+      "epoch": 0.01,
+      "loss_char": 0.7645,
+      "loss_token": 1.6297,
+      "lr": "2.00e-07",
+      "norm": 5.8568,
+      "step": 298
+    },
+    {
+      "acc_char": 0.2996,
+      "acc_token": 0.6744,
+      "epoch": 0.01,
+      "loss_char": 0.7564,
+      "loss_token": 1.5995,
+      "lr": "2.00e-07",
+      "norm": 5.9268,
+      "step": 299
+    },
+    {
+      "acc_char": 0.3044,
+      "acc_token": 0.6908,
+      "epoch": 0.01,
+      "loss_char": 0.7074,
+      "loss_token": 1.5607,
+      "lr": "2.00e-07",
+      "norm": 5.5007,
+      "step": 300
+    },
+    {
+      "acc_char": 0.3061,
+      "acc_token": 0.6878,
+      "epoch": 0.01,
+      "loss_char": 0.7224,
+      "loss_token": 1.6879,
+      "lr": "2.00e-07",
+      "norm": 5.6493,
+      "step": 301
+    },
+    {
+      "acc_char": 0.2944,
+      "acc_token": 0.6705,
+      "epoch": 0.01,
+      "loss_char": 0.7456,
+      "loss_token": 1.609,
+      "lr": "2.00e-07",
+      "norm": 5.7389,
+      "step": 302
+    },
+    {
+      "acc_char": 0.2968,
+      "acc_token": 0.6759,
+      "epoch": 0.01,
+      "loss_char": 0.7284,
+      "loss_token": 1.6731,
+      "lr": "2.00e-07",
+      "norm": 5.523,
+      "step": 303
+    },
+    {
+      "acc_char": 0.2979,
+      "acc_token": 0.6824,
+      "epoch": 0.01,
+      "loss_char": 0.7128,
+      "loss_token": 1.5829,
+      "lr": "2.00e-07",
+      "norm": 5.5804,
+      "step": 304
+    },
+    {
+      "acc_char": 0.2975,
+      "acc_token": 0.6765,
+      "epoch": 0.01,
+      "loss_char": 0.7475,
+      "loss_token": 1.6629,
+      "lr": "2.00e-07",
+      "norm": 5.8821,
+      "step": 305
+    },
+    {
+      "acc_char": 0.3019,
+      "acc_token": 0.6788,
+      "epoch": 0.01,
+      "loss_char": 0.7458,
+      "loss_token": 1.6448,
+      "lr": "2.00e-07",
+      "norm": 5.7853,
+      "step": 306
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6883,
+      "epoch": 0.01,
+      "loss_char": 0.7007,
+      "loss_token": 1.5519,
+      "lr": "2.00e-07",
+      "norm": 5.3026,
+      "step": 307
+    },
+    {
+      "acc_char": 0.2937,
+      "acc_token": 0.6658,
+      "epoch": 0.01,
+      "loss_char": 0.7739,
+      "loss_token": 1.7083,
+      "lr": "2.00e-07",
+      "norm": 5.9143,
+      "step": 308
+    },
+    {
+      "acc_char": 0.3012,
+      "acc_token": 0.6801,
+      "epoch": 0.01,
+      "loss_char": 0.7371,
+      "loss_token": 1.6619,
+      "lr": "2.00e-07",
+      "norm": 5.7714,
+      "step": 309
+    },
+    {
+      "acc_char": 0.306,
+      "acc_token": 0.6964,
+      "epoch": 0.01,
+      "loss_char": 0.681,
+      "loss_token": 1.656,
+      "lr": "2.00e-07",
+      "norm": 5.8231,
+      "step": 310
+    },
+    {
+      "acc_char": 0.3065,
+      "acc_token": 0.7017,
+      "epoch": 0.01,
+      "loss_char": 0.6808,
+      "loss_token": 1.5909,
+      "lr": "2.00e-07",
+      "norm": 5.4979,
+      "step": 311
+    },
+    {
+      "acc_char": 0.3129,
+      "acc_token": 0.6958,
+      "epoch": 0.01,
+      "loss_char": 0.7117,
+      "loss_token": 1.619,
+      "lr": "2.00e-07",
+      "norm": 5.7152,
+      "step": 312
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6706,
+      "epoch": 0.01,
+      "loss_char": 0.7657,
+      "loss_token": 1.6683,
+      "lr": "2.00e-07",
+      "norm": 5.6728,
+      "step": 313
+    },
+    {
+      "acc_char": 0.3176,
+      "acc_token": 0.7069,
+      "epoch": 0.01,
+      "loss_char": 0.6768,
+      "loss_token": 1.6478,
+      "lr": "2.00e-07",
+      "norm": 5.7274,
+      "step": 314
+    },
+    {
+      "acc_char": 0.2875,
+      "acc_token": 0.6621,
+      "epoch": 0.01,
+      "loss_char": 0.7648,
+      "loss_token": 1.694,
+      "lr": "2.00e-07",
+      "norm": 5.683,
+      "step": 315
+    },
+    {
+      "acc_char": 0.301,
+      "acc_token": 0.678,
+      "epoch": 0.01,
+      "loss_char": 0.7439,
+      "loss_token": 1.6851,
+      "lr": "2.00e-07",
+      "norm": 5.7478,
+      "step": 316
+    },
+    {
+      "acc_char": 0.3038,
+      "acc_token": 0.6839,
+      "epoch": 0.01,
+      "loss_char": 0.7546,
+      "loss_token": 1.5604,
+      "lr": "2.00e-07",
+      "norm": 5.7545,
+      "step": 317
+    },
+    {
+      "acc_char": 0.3005,
+      "acc_token": 0.6839,
+      "epoch": 0.01,
+      "loss_char": 0.7142,
+      "loss_token": 1.6036,
+      "lr": "2.00e-07",
+      "norm": 5.8406,
+      "step": 318
+    },
+    {
+      "acc_char": 0.309,
+      "acc_token": 0.6981,
+      "epoch": 0.01,
+      "loss_char": 0.6622,
+      "loss_token": 1.5909,
+      "lr": "2.00e-07",
+      "norm": 5.6064,
+      "step": 319
+    },
+    {
+      "acc_char": 0.3106,
+      "acc_token": 0.6836,
+      "epoch": 0.01,
+      "loss_char": 0.7384,
+      "loss_token": 1.6081,
+      "lr": "2.00e-07",
+      "norm": 5.972,
+      "step": 320
+    },
+    {
+      "acc_char": 0.3107,
+      "acc_token": 0.6937,
+      "epoch": 0.01,
+      "loss_char": 0.6974,
+      "loss_token": 1.5829,
+      "lr": "2.00e-07",
+      "norm": 5.9956,
+      "step": 321
+    },
+    {
+      "acc_char": 0.2968,
+      "acc_token": 0.6709,
+      "epoch": 0.01,
+      "loss_char": 0.7633,
+      "loss_token": 1.606,
+      "lr": "2.00e-07",
+      "norm": 5.7139,
+      "step": 322
+    },
+    {
+      "acc_char": 0.3093,
+      "acc_token": 0.7082,
+      "epoch": 0.01,
+      "loss_char": 0.639,
+      "loss_token": 1.5356,
+      "lr": "2.00e-07",
+      "norm": 5.5638,
+      "step": 323
+    },
+    {
+      "acc_char": 0.3121,
+      "acc_token": 0.711,
+      "epoch": 0.01,
+      "loss_char": 0.6582,
+      "loss_token": 1.5754,
+      "lr": "2.00e-07",
+      "norm": 5.5749,
+      "step": 324
+    },
+    {
+      "acc_char": 0.3014,
+      "acc_token": 0.6806,
+      "epoch": 0.01,
+      "loss_char": 0.7369,
+      "loss_token": 1.612,
+      "lr": "2.00e-07",
+      "norm": 5.7404,
+      "step": 325
+    },
+    {
+      "acc_char": 0.3084,
+      "acc_token": 0.6974,
+      "epoch": 0.01,
+      "loss_char": 0.6704,
+      "loss_token": 1.5965,
+      "lr": "2.00e-07",
+      "norm": 5.8155,
+      "step": 326
+    },
+    {
+      "acc_char": 0.2937,
+      "acc_token": 0.6618,
+      "epoch": 0.01,
+      "loss_char": 0.7958,
+      "loss_token": 1.7527,
+      "lr": "2.00e-07",
+      "norm": 5.9642,
+      "step": 327
+    },
+    {
+      "acc_char": 0.3047,
+      "acc_token": 0.6887,
+      "epoch": 0.01,
+      "loss_char": 0.7168,
+      "loss_token": 1.6637,
+      "lr": "2.00e-07",
+      "norm": 5.7664,
+      "step": 328
+    },
+    {
+      "acc_char": 0.3076,
+      "acc_token": 0.6757,
+      "epoch": 0.01,
+      "loss_char": 0.7671,
+      "loss_token": 1.7099,
+      "lr": "2.00e-07",
+      "norm": 6.0867,
+      "step": 329
+    },
+    {
+      "acc_char": 0.3022,
+      "acc_token": 0.6847,
+      "epoch": 0.01,
+      "loss_char": 0.7003,
+      "loss_token": 1.5672,
+      "lr": "2.00e-07",
+      "norm": 5.9077,
+      "step": 330
+    },
+    {
+      "acc_char": 0.3147,
+      "acc_token": 0.7041,
+      "epoch": 0.01,
+      "loss_char": 0.6763,
+      "loss_token": 1.5775,
+      "lr": "2.00e-07",
+      "norm": 5.5642,
+      "step": 331
+    },
+    {
+      "acc_char": 0.3008,
+      "acc_token": 0.673,
+      "epoch": 0.01,
+      "loss_char": 0.7647,
+      "loss_token": 1.6656,
+      "lr": "2.00e-07",
+      "norm": 5.7092,
+      "step": 332
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6857,
+      "epoch": 0.01,
+      "loss_char": 0.6981,
+      "loss_token": 1.6254,
+      "lr": "2.00e-07",
+      "norm": 5.8577,
+      "step": 333
+    },
+    {
+      "acc_char": 0.2911,
+      "acc_token": 0.6324,
+      "epoch": 0.01,
+      "loss_char": 0.8438,
+      "loss_token": 1.6785,
+      "lr": "2.00e-07",
+      "norm": 5.2846,
+      "step": 334
+    },
+    {
+      "acc_char": 0.3018,
+      "acc_token": 0.6796,
+      "epoch": 0.01,
+      "loss_char": 0.748,
+      "loss_token": 1.6903,
+      "lr": "2.00e-07",
+      "norm": 6.3385,
+      "step": 335
+    },
+    {
+      "acc_char": 0.3005,
+      "acc_token": 0.6533,
+      "epoch": 0.01,
+      "loss_char": 0.7987,
+      "loss_token": 1.722,
+      "lr": "2.00e-07",
+      "norm": 5.7037,
+      "step": 336
+    },
+    {
+      "acc_char": 0.2927,
+      "acc_token": 0.6856,
+      "epoch": 0.01,
+      "loss_char": 0.676,
+      "loss_token": 1.6197,
+      "lr": "2.00e-07",
+      "norm": 5.504,
+      "step": 337
+    },
+    {
+      "acc_char": 0.2836,
+      "acc_token": 0.6138,
+      "epoch": 0.01,
+      "loss_char": 0.9242,
+      "loss_token": 1.892,
+      "lr": "2.00e-07",
+      "norm": 5.9519,
+      "step": 338
+    },
+    {
+      "acc_char": 0.3061,
+      "acc_token": 0.69,
+      "epoch": 0.01,
+      "loss_char": 0.7105,
+      "loss_token": 1.6633,
+      "lr": "2.00e-07",
+      "norm": 5.6888,
+      "step": 339
+    },
+    {
+      "acc_char": 0.3114,
+      "acc_token": 0.6959,
+      "epoch": 0.01,
+      "loss_char": 0.7106,
+      "loss_token": 1.6273,
+      "lr": "2.00e-07",
+      "norm": 5.9986,
+      "step": 340
+    },
+    {
+      "acc_char": 0.3007,
+      "acc_token": 0.6855,
+      "epoch": 0.01,
+      "loss_char": 0.7016,
+      "loss_token": 1.6079,
+      "lr": "2.00e-07",
+      "norm": 5.857,
+      "step": 341
+    },
+    {
+      "acc_char": 0.2979,
+      "acc_token": 0.6707,
+      "epoch": 0.01,
+      "loss_char": 0.7816,
+      "loss_token": 1.7775,
+      "lr": "2.00e-07",
+      "norm": 6.0107,
+      "step": 342
+    },
+    {
+      "acc_char": 0.3073,
+      "acc_token": 0.6961,
+      "epoch": 0.01,
+      "loss_char": 0.6894,
+      "loss_token": 1.6348,
+      "lr": "2.00e-07",
+      "norm": 5.8313,
+      "step": 343
+    },
+    {
+      "acc_char": 0.3084,
+      "acc_token": 0.6846,
+      "epoch": 0.01,
+      "loss_char": 0.7451,
+      "loss_token": 1.574,
+      "lr": "2.00e-07",
+      "norm": 5.7596,
+      "step": 344
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6724,
+      "epoch": 0.01,
+      "loss_char": 0.7413,
+      "loss_token": 1.6613,
+      "lr": "2.00e-07",
+      "norm": 5.8325,
+      "step": 345
+    },
+    {
+      "acc_char": 0.3101,
+      "acc_token": 0.6854,
+      "epoch": 0.01,
+      "loss_char": 0.743,
+      "loss_token": 1.608,
+      "lr": "2.00e-07",
+      "norm": 5.5323,
+      "step": 346
+    },
+    {
+      "acc_char": 0.2904,
+      "acc_token": 0.6635,
+      "epoch": 0.01,
+      "loss_char": 0.7893,
+      "loss_token": 1.7397,
+      "lr": "2.00e-07",
+      "norm": 6.1739,
+      "step": 347
+    },
+    {
+      "acc_char": 0.3111,
+      "acc_token": 0.6931,
+      "epoch": 0.01,
+      "loss_char": 0.709,
+      "loss_token": 1.6418,
+      "lr": "2.00e-07",
+      "norm": 5.9963,
+      "step": 348
+    },
+    {
+      "acc_char": 0.2909,
+      "acc_token": 0.6682,
+      "epoch": 0.01,
+      "loss_char": 0.7396,
+      "loss_token": 1.6422,
+      "lr": "2.00e-07",
+      "norm": 5.6683,
+      "step": 349
+    },
+    {
+      "acc_char": 0.3238,
+      "acc_token": 0.7197,
+      "epoch": 0.01,
+      "loss_char": 0.6568,
+      "loss_token": 1.5539,
+      "lr": "2.00e-07",
+      "norm": 5.7899,
+      "step": 350
+    },
+    {
+      "acc_char": 0.3086,
+      "acc_token": 0.6668,
+      "epoch": 0.01,
+      "loss_char": 0.7888,
+      "loss_token": 1.6842,
+      "lr": "2.00e-07",
+      "norm": 5.2554,
+      "step": 351
+    },
+    {
+      "acc_char": 0.3072,
+      "acc_token": 0.6981,
+      "epoch": 0.01,
+      "loss_char": 0.6763,
+      "loss_token": 1.5788,
+      "lr": "2.00e-07",
+      "norm": 5.5028,
+      "step": 352
+    },
+    {
+      "acc_char": 0.2979,
+      "acc_token": 0.6875,
+      "epoch": 0.01,
+      "loss_char": 0.6671,
+      "loss_token": 1.6384,
+      "lr": "2.00e-07",
+      "norm": 6.1584,
+      "step": 353
+    },
+    {
+      "acc_char": 0.2829,
+      "acc_token": 0.6175,
+      "epoch": 0.01,
+      "loss_char": 0.8946,
+      "loss_token": 1.8867,
+      "lr": "2.00e-07",
+      "norm": 5.5444,
+      "step": 354
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.7051,
+      "epoch": 0.01,
+      "loss_char": 0.6552,
+      "loss_token": 1.6427,
+      "lr": "2.00e-07",
+      "norm": 5.7578,
+      "step": 355
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6836,
+      "epoch": 0.01,
+      "loss_char": 0.7052,
+      "loss_token": 1.6508,
+      "lr": "2.00e-07",
+      "norm": 5.6257,
+      "step": 356
+    },
+    {
+      "acc_char": 0.284,
+      "acc_token": 0.6558,
+      "epoch": 0.01,
+      "loss_char": 0.7454,
+      "loss_token": 1.6895,
+      "lr": "2.00e-07",
+      "norm": 5.9138,
+      "step": 357
+    },
+    {
+      "acc_char": 0.2886,
+      "acc_token": 0.6673,
+      "epoch": 0.01,
+      "loss_char": 0.7445,
+      "loss_token": 1.6311,
+      "lr": "2.00e-07",
+      "norm": 5.7174,
+      "step": 358
+    },
+    {
+      "acc_char": 0.3009,
+      "acc_token": 0.6745,
+      "epoch": 0.01,
+      "loss_char": 0.7696,
+      "loss_token": 1.666,
+      "lr": "2.00e-07",
+      "norm": 5.9078,
+      "step": 359
+    },
+    {
+      "acc_char": 0.2994,
+      "acc_token": 0.6862,
+      "epoch": 0.01,
+      "loss_char": 0.7015,
+      "loss_token": 1.5945,
+      "lr": "2.00e-07",
+      "norm": 5.9093,
+      "step": 360
+    },
+    {
+      "acc_char": 0.3052,
+      "acc_token": 0.6858,
+      "epoch": 0.01,
+      "loss_char": 0.731,
+      "loss_token": 1.6641,
+      "lr": "2.00e-07",
+      "norm": 5.9643,
+      "step": 361
+    },
+    {
+      "acc_char": 0.3113,
+      "acc_token": 0.7002,
+      "epoch": 0.01,
+      "loss_char": 0.6977,
+      "loss_token": 1.5529,
+      "lr": "2.00e-07",
+      "norm": 5.548,
+      "step": 362
+    },
+    {
+      "acc_char": 0.2994,
+      "acc_token": 0.6823,
+      "epoch": 0.01,
+      "loss_char": 0.7226,
+      "loss_token": 1.6606,
+      "lr": "2.00e-07",
+      "norm": 5.6489,
+      "step": 363
+    },
+    {
+      "acc_char": 0.3144,
+      "acc_token": 0.6987,
+      "epoch": 0.01,
+      "loss_char": 0.7038,
+      "loss_token": 1.6622,
+      "lr": "2.00e-07",
+      "norm": 5.6968,
+      "step": 364
+    },
+    {
+      "acc_char": 0.3049,
+      "acc_token": 0.6923,
+      "epoch": 0.01,
+      "loss_char": 0.6968,
+      "loss_token": 1.5419,
+      "lr": "2.00e-07",
+      "norm": 5.4769,
+      "step": 365
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6884,
+      "epoch": 0.01,
+      "loss_char": 0.6917,
+      "loss_token": 1.5916,
+      "lr": "2.00e-07",
+      "norm": 5.8595,
+      "step": 366
+    },
+    {
+      "acc_char": 0.2899,
+      "acc_token": 0.6622,
+      "epoch": 0.01,
+      "loss_char": 0.774,
+      "loss_token": 1.6747,
+      "lr": "2.00e-07",
+      "norm": 6.1018,
+      "step": 367
+    },
+    {
+      "acc_char": 0.2971,
+      "acc_token": 0.6756,
+      "epoch": 0.01,
+      "loss_char": 0.733,
+      "loss_token": 1.6098,
+      "lr": "2.00e-07",
+      "norm": 5.7129,
+      "step": 368
+    },
+    {
+      "acc_char": 0.3075,
+      "acc_token": 0.6922,
+      "epoch": 0.01,
+      "loss_char": 0.6952,
+      "loss_token": 1.628,
+      "lr": "2.00e-07",
+      "norm": 5.8883,
+      "step": 369
+    },
+    {
+      "acc_char": 0.3026,
+      "acc_token": 0.7124,
+      "epoch": 0.01,
+      "loss_char": 0.6241,
+      "loss_token": 1.5773,
+      "lr": "2.00e-07",
+      "norm": 5.5128,
+      "step": 370
+    },
+    {
+      "acc_char": 0.299,
+      "acc_token": 0.6668,
+      "epoch": 0.01,
+      "loss_char": 0.784,
+      "loss_token": 1.6683,
+      "lr": "2.00e-07",
+      "norm": 5.9957,
+      "step": 371
+    },
+    {
+      "acc_char": 0.2926,
+      "acc_token": 0.6731,
+      "epoch": 0.01,
+      "loss_char": 0.735,
+      "loss_token": 1.562,
+      "lr": "2.00e-07",
+      "norm": 5.4053,
+      "step": 372
+    },
+    {
+      "acc_char": 0.2935,
+      "acc_token": 0.6759,
+      "epoch": 0.01,
+      "loss_char": 0.7445,
+      "loss_token": 1.7457,
+      "lr": "2.00e-07",
+      "norm": 5.9811,
+      "step": 373
+    },
+    {
+      "acc_char": 0.2991,
+      "acc_token": 0.6718,
+      "epoch": 0.01,
+      "loss_char": 0.7618,
+      "loss_token": 1.6714,
+      "lr": "2.00e-07",
+      "norm": 5.7125,
+      "step": 374
+    },
+    {
+      "acc_char": 0.3105,
+      "acc_token": 0.6914,
+      "epoch": 0.01,
+      "loss_char": 0.7096,
+      "loss_token": 1.6491,
+      "lr": "2.00e-07",
+      "norm": 5.7288,
+      "step": 375
+    },
+    {
+      "acc_char": 0.3093,
+      "acc_token": 0.7048,
+      "epoch": 0.01,
+      "loss_char": 0.6826,
+      "loss_token": 1.6267,
+      "lr": "2.00e-07",
+      "norm": 5.7194,
+      "step": 376
+    },
+    {
+      "acc_char": 0.3067,
+      "acc_token": 0.6888,
+      "epoch": 0.01,
+      "loss_char": 0.6965,
+      "loss_token": 1.6184,
+      "lr": "2.00e-07",
+      "norm": 5.7391,
+      "step": 377
+    },
+    {
+      "acc_char": 0.2957,
+      "acc_token": 0.6721,
+      "epoch": 0.01,
+      "loss_char": 0.7448,
+      "loss_token": 1.5463,
+      "lr": "2.00e-07",
+      "norm": 5.3704,
+      "step": 378
+    },
+    {
+      "acc_char": 0.3098,
+      "acc_token": 0.6886,
+      "epoch": 0.01,
+      "loss_char": 0.7316,
+      "loss_token": 1.6599,
+      "lr": "2.00e-07",
+      "norm": 5.8194,
+      "step": 379
+    },
+    {
+      "acc_char": 0.2925,
+      "acc_token": 0.6664,
+      "epoch": 0.01,
+      "loss_char": 0.7882,
+      "loss_token": 1.8153,
+      "lr": "2.00e-07",
+      "norm": 5.9483,
+      "step": 380
+    },
+    {
+      "acc_char": 0.2972,
+      "acc_token": 0.6659,
+      "epoch": 0.01,
+      "loss_char": 0.7625,
+      "loss_token": 1.6956,
+      "lr": "2.00e-07",
+      "norm": 5.8517,
+      "step": 381
+    },
+    {
+      "acc_char": 0.3071,
+      "acc_token": 0.6989,
+      "epoch": 0.01,
+      "loss_char": 0.6656,
+      "loss_token": 1.6748,
+      "lr": "2.00e-07",
+      "norm": 5.7486,
+      "step": 382
+    },
+    {
+      "acc_char": 0.3091,
+      "acc_token": 0.7008,
+      "epoch": 0.01,
+      "loss_char": 0.6794,
+      "loss_token": 1.5323,
+      "lr": "2.00e-07",
+      "norm": 5.2689,
+      "step": 383
+    },
+    {
+      "acc_char": 0.3172,
+      "acc_token": 0.7194,
+      "epoch": 0.01,
+      "loss_char": 0.6259,
+      "loss_token": 1.4796,
+      "lr": "2.00e-07",
+      "norm": 5.2914,
+      "step": 384
+    },
+    {
+      "acc_char": 0.3008,
+      "acc_token": 0.6811,
+      "epoch": 0.01,
+      "loss_char": 0.7145,
+      "loss_token": 1.6898,
+      "lr": "2.00e-07",
+      "norm": 5.7635,
+      "step": 385
+    },
+    {
+      "acc_char": 0.3045,
+      "acc_token": 0.6842,
+      "epoch": 0.01,
+      "loss_char": 0.7293,
+      "loss_token": 1.6527,
+      "lr": "2.00e-07",
+      "norm": 6.0697,
+      "step": 386
+    },
+    {
+      "acc_char": 0.2971,
+      "acc_token": 0.6735,
+      "epoch": 0.01,
+      "loss_char": 0.7487,
+      "loss_token": 1.6422,
+      "lr": "2.00e-07",
+      "norm": 5.7503,
+      "step": 387
+    },
+    {
+      "acc_char": 0.2948,
+      "acc_token": 0.6757,
+      "epoch": 0.01,
+      "loss_char": 0.7323,
+      "loss_token": 1.6915,
+      "lr": "2.00e-07",
+      "norm": 5.9776,
+      "step": 388
+    },
+    {
+      "acc_char": 0.307,
+      "acc_token": 0.6914,
+      "epoch": 0.01,
+      "loss_char": 0.7019,
+      "loss_token": 1.5222,
+      "lr": "2.00e-07",
+      "norm": 5.7667,
+      "step": 389
+    },
+    {
+      "acc_char": 0.3014,
+      "acc_token": 0.6679,
+      "epoch": 0.01,
+      "loss_char": 0.7908,
+      "loss_token": 1.7331,
+      "lr": "2.00e-07",
+      "norm": 6.0986,
+      "step": 390
+    },
+    {
+      "acc_char": 0.2952,
+      "acc_token": 0.6688,
+      "epoch": 0.01,
+      "loss_char": 0.7422,
+      "loss_token": 1.6141,
+      "lr": "2.00e-07",
+      "norm": 5.766,
+      "step": 391
+    },
+    {
+      "acc_char": 0.2958,
+      "acc_token": 0.6717,
+      "epoch": 0.01,
+      "loss_char": 0.7446,
+      "loss_token": 1.6848,
+      "lr": "2.00e-07",
+      "norm": 5.7914,
+      "step": 392
+    },
+    {
+      "acc_char": 0.2967,
+      "acc_token": 0.6725,
+      "epoch": 0.01,
+      "loss_char": 0.7515,
+      "loss_token": 1.6382,
+      "lr": "2.00e-07",
+      "norm": 5.5648,
+      "step": 393
+    },
+    {
+      "acc_char": 0.3051,
+      "acc_token": 0.6734,
+      "epoch": 0.01,
+      "loss_char": 0.7527,
+      "loss_token": 1.6441,
+      "lr": "2.00e-07",
+      "norm": 5.8132,
+      "step": 394
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6883,
+      "epoch": 0.01,
+      "loss_char": 0.6923,
+      "loss_token": 1.6338,
+      "lr": "2.00e-07",
+      "norm": 5.6723,
+      "step": 395
+    },
+    {
+      "acc_char": 0.3039,
+      "acc_token": 0.6824,
+      "epoch": 0.01,
+      "loss_char": 0.7433,
+      "loss_token": 1.6936,
+      "lr": "2.00e-07",
+      "norm": 5.8343,
+      "step": 396
+    },
+    {
+      "acc_char": 0.3058,
+      "acc_token": 0.6873,
+      "epoch": 0.01,
+      "loss_char": 0.7231,
+      "loss_token": 1.6963,
+      "lr": "2.00e-07",
+      "norm": 6.3947,
+      "step": 397
+    },
+    {
+      "acc_char": 0.2999,
+      "acc_token": 0.6678,
+      "epoch": 0.01,
+      "loss_char": 0.7728,
+      "loss_token": 1.5994,
+      "lr": "2.00e-07",
+      "norm": 5.501,
+      "step": 398
+    },
+    {
+      "acc_char": 0.3092,
+      "acc_token": 0.7044,
+      "epoch": 0.01,
+      "loss_char": 0.6667,
+      "loss_token": 1.6234,
+      "lr": "2.00e-07",
+      "norm": 5.8259,
+      "step": 399
+    },
+    {
+      "acc_char": 0.3093,
+      "acc_token": 0.6906,
+      "epoch": 0.01,
+      "loss_char": 0.7331,
+      "loss_token": 1.604,
+      "lr": "2.00e-07",
+      "norm": 5.7938,
+      "step": 400
+    },
+    {
+      "acc_char": 0.2956,
+      "acc_token": 0.6737,
+      "epoch": 0.01,
+      "loss_char": 0.7542,
+      "loss_token": 1.6651,
+      "lr": "2.00e-07",
+      "norm": 5.6362,
+      "step": 401
+    },
+    {
+      "acc_char": 0.2992,
+      "acc_token": 0.6798,
+      "epoch": 0.01,
+      "loss_char": 0.757,
+      "loss_token": 1.6434,
+      "lr": "2.00e-07",
+      "norm": 5.7258,
+      "step": 402
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6818,
+      "epoch": 0.01,
+      "loss_char": 0.7265,
+      "loss_token": 1.6712,
+      "lr": "2.00e-07",
+      "norm": 5.5462,
+      "step": 403
+    },
+    {
+      "acc_char": 0.3177,
+      "acc_token": 0.7261,
+      "epoch": 0.01,
+      "loss_char": 0.6203,
+      "loss_token": 1.4859,
+      "lr": "2.00e-07",
+      "norm": 5.4174,
+      "step": 404
+    },
+    {
+      "acc_char": 0.3025,
+      "acc_token": 0.6801,
+      "epoch": 0.01,
+      "loss_char": 0.7286,
+      "loss_token": 1.6256,
+      "lr": "2.00e-07",
+      "norm": 5.3541,
+      "step": 405
+    },
+    {
+      "acc_char": 0.3148,
+      "acc_token": 0.7173,
+      "epoch": 0.01,
+      "loss_char": 0.6452,
+      "loss_token": 1.5515,
+      "lr": "2.00e-07",
+      "norm": 5.4521,
+      "step": 406
+    },
+    {
+      "acc_char": 0.3041,
+      "acc_token": 0.6856,
+      "epoch": 0.01,
+      "loss_char": 0.7246,
+      "loss_token": 1.6415,
+      "lr": "2.00e-07",
+      "norm": 5.712,
+      "step": 407
+    },
+    {
+      "acc_char": 0.3038,
+      "acc_token": 0.693,
+      "epoch": 0.01,
+      "loss_char": 0.7085,
+      "loss_token": 1.6706,
+      "lr": "2.00e-07",
+      "norm": 5.8686,
+      "step": 408
+    },
+    {
+      "acc_char": 0.281,
+      "acc_token": 0.6554,
+      "epoch": 0.01,
+      "loss_char": 0.7944,
+      "loss_token": 1.728,
+      "lr": "2.00e-07",
+      "norm": 5.8449,
+      "step": 409
+    },
+    {
+      "acc_char": 0.3084,
+      "acc_token": 0.6939,
+      "epoch": 0.01,
+      "loss_char": 0.6905,
+      "loss_token": 1.6071,
+      "lr": "2.00e-07",
+      "norm": 5.521,
+      "step": 410
+    },
+    {
+      "acc_char": 0.3061,
+      "acc_token": 0.6905,
+      "epoch": 0.01,
+      "loss_char": 0.7053,
+      "loss_token": 1.6249,
+      "lr": "2.00e-07",
+      "norm": 5.6452,
+      "step": 411
+    },
+    {
+      "acc_char": 0.3026,
+      "acc_token": 0.6691,
+      "epoch": 0.01,
+      "loss_char": 0.7783,
+      "loss_token": 1.66,
+      "lr": "2.00e-07",
+      "norm": 5.9179,
+      "step": 412
+    },
+    {
+      "acc_char": 0.309,
+      "acc_token": 0.705,
+      "epoch": 0.01,
+      "loss_char": 0.6664,
+      "loss_token": 1.5434,
+      "lr": "2.00e-07",
+      "norm": 5.302,
+      "step": 413
+    },
+    {
+      "acc_char": 0.3089,
+      "acc_token": 0.6967,
+      "epoch": 0.01,
+      "loss_char": 0.6844,
+      "loss_token": 1.5804,
+      "lr": "2.00e-07",
+      "norm": 5.8454,
+      "step": 414
+    },
+    {
+      "acc_char": 0.2907,
+      "acc_token": 0.6686,
+      "epoch": 0.01,
+      "loss_char": 0.7367,
+      "loss_token": 1.6778,
+      "lr": "2.00e-07",
+      "norm": 5.7327,
+      "step": 415
+    },
+    {
+      "acc_char": 0.2921,
+      "acc_token": 0.6253,
+      "epoch": 0.01,
+      "loss_char": 0.8878,
+      "loss_token": 1.7222,
+      "lr": "2.00e-07",
+      "norm": 5.5795,
+      "step": 416
+    },
+    {
+      "acc_char": 0.2962,
+      "acc_token": 0.6945,
+      "epoch": 0.01,
+      "loss_char": 0.6684,
+      "loss_token": 1.5484,
+      "lr": "2.00e-07",
+      "norm": 5.6565,
+      "step": 417
+    },
+    {
+      "acc_char": 0.2975,
+      "acc_token": 0.6701,
+      "epoch": 0.01,
+      "loss_char": 0.7654,
+      "loss_token": 1.6914,
+      "lr": "2.00e-07",
+      "norm": 5.5792,
+      "step": 418
+    },
+    {
+      "acc_char": 0.3157,
+      "acc_token": 0.7062,
+      "epoch": 0.01,
+      "loss_char": 0.6615,
+      "loss_token": 1.6023,
+      "lr": "2.00e-07",
+      "norm": 5.8135,
+      "step": 419
+    },
+    {
+      "acc_char": 0.3077,
+      "acc_token": 0.693,
+      "epoch": 0.01,
+      "loss_char": 0.6872,
+      "loss_token": 1.5777,
+      "lr": "2.00e-07",
+      "norm": 5.8938,
+      "step": 420
+    },
+    {
+      "acc_char": 0.2868,
+      "acc_token": 0.6238,
+      "epoch": 0.01,
+      "loss_char": 0.8702,
+      "loss_token": 1.664,
+      "lr": "2.00e-07",
+      "norm": 5.3468,
+      "step": 421
+    },
+    {
+      "acc_char": 0.3136,
+      "acc_token": 0.7069,
+      "epoch": 0.01,
+      "loss_char": 0.6709,
+      "loss_token": 1.5435,
+      "lr": "2.00e-07",
+      "norm": 5.5504,
+      "step": 422
+    },
+    {
+      "acc_char": 0.299,
+      "acc_token": 0.6801,
+      "epoch": 0.01,
+      "loss_char": 0.7479,
+      "loss_token": 1.6409,
+      "lr": "2.00e-07",
+      "norm": 6.0676,
+      "step": 423
+    },
+    {
+      "acc_char": 0.3009,
+      "acc_token": 0.6743,
+      "epoch": 0.01,
+      "loss_char": 0.7493,
+      "loss_token": 1.5541,
+      "lr": "2.00e-07",
+      "norm": 5.6725,
+      "step": 424
+    },
+    {
+      "acc_char": 0.3095,
+      "acc_token": 0.7043,
+      "epoch": 0.01,
+      "loss_char": 0.6468,
+      "loss_token": 1.5845,
+      "lr": "2.00e-07",
+      "norm": 5.718,
+      "step": 425
+    },
+    {
+      "acc_char": 0.3033,
+      "acc_token": 0.6742,
+      "epoch": 0.01,
+      "loss_char": 0.7576,
+      "loss_token": 1.6773,
+      "lr": "2.00e-07",
+      "norm": 5.778,
+      "step": 426
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.674,
+      "epoch": 0.01,
+      "loss_char": 0.7737,
+      "loss_token": 1.7231,
+      "lr": "2.00e-07",
+      "norm": 5.9369,
+      "step": 427
+    },
+    {
+      "acc_char": 0.3146,
+      "acc_token": 0.7153,
+      "epoch": 0.01,
+      "loss_char": 0.6293,
+      "loss_token": 1.5328,
+      "lr": "2.00e-07",
+      "norm": 5.5686,
+      "step": 428
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.6914,
+      "epoch": 0.01,
+      "loss_char": 0.7156,
+      "loss_token": 1.5677,
+      "lr": "2.00e-07",
+      "norm": 5.7421,
+      "step": 429
+    },
+    {
+      "acc_char": 0.2992,
+      "acc_token": 0.6773,
+      "epoch": 0.01,
+      "loss_char": 0.7398,
+      "loss_token": 1.7212,
+      "lr": "2.00e-07",
+      "norm": 5.7725,
+      "step": 430
+    },
+    {
+      "acc_char": 0.2965,
+      "acc_token": 0.669,
+      "epoch": 0.01,
+      "loss_char": 0.7813,
+      "loss_token": 1.7132,
+      "lr": "2.00e-07",
+      "norm": 5.7314,
+      "step": 431
+    },
+    {
+      "acc_char": 0.3065,
+      "acc_token": 0.6996,
+      "epoch": 0.01,
+      "loss_char": 0.6745,
+      "loss_token": 1.5433,
+      "lr": "2.00e-07",
+      "norm": 5.6082,
+      "step": 432
+    },
+    {
+      "acc_char": 0.3203,
+      "acc_token": 0.7247,
+      "epoch": 0.01,
+      "loss_char": 0.6148,
+      "loss_token": 1.4884,
+      "lr": "2.00e-07",
+      "norm": 5.4403,
+      "step": 433
+    },
+    {
+      "acc_char": 0.3058,
+      "acc_token": 0.699,
+      "epoch": 0.01,
+      "loss_char": 0.6423,
+      "loss_token": 1.5294,
+      "lr": "2.00e-07",
+      "norm": 5.5608,
+      "step": 434
+    },
+    {
+      "acc_char": 0.2914,
+      "acc_token": 0.6616,
+      "epoch": 0.01,
+      "loss_char": 0.7899,
+      "loss_token": 1.7818,
+      "lr": "2.00e-07",
+      "norm": 6.3515,
+      "step": 435
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6781,
+      "epoch": 0.01,
+      "loss_char": 0.7233,
+      "loss_token": 1.643,
+      "lr": "2.00e-07",
+      "norm": 5.7809,
+      "step": 436
+    },
+    {
+      "acc_char": 0.3071,
+      "acc_token": 0.6873,
+      "epoch": 0.01,
+      "loss_char": 0.7219,
+      "loss_token": 1.6312,
+      "lr": "2.00e-07",
+      "norm": 5.8777,
+      "step": 437
+    },
+    {
+      "acc_char": 0.293,
+      "acc_token": 0.6592,
+      "epoch": 0.01,
+      "loss_char": 0.7718,
+      "loss_token": 1.7166,
+      "lr": "2.00e-07",
+      "norm": 6.0267,
+      "step": 438
+    },
+    {
+      "acc_char": 0.2957,
+      "acc_token": 0.6753,
+      "epoch": 0.01,
+      "loss_char": 0.75,
+      "loss_token": 1.6695,
+      "lr": "2.00e-07",
+      "norm": 5.7281,
+      "step": 439
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6798,
+      "epoch": 0.01,
+      "loss_char": 0.7365,
+      "loss_token": 1.6402,
+      "lr": "2.00e-07",
+      "norm": 5.9977,
+      "step": 440
+    },
+    {
+      "acc_char": 0.296,
+      "acc_token": 0.6709,
+      "epoch": 0.01,
+      "loss_char": 0.7499,
+      "loss_token": 1.6319,
+      "lr": "2.00e-07",
+      "norm": 5.8092,
+      "step": 441
+    },
+    {
+      "acc_char": 0.303,
+      "acc_token": 0.6867,
+      "epoch": 0.01,
+      "loss_char": 0.696,
+      "loss_token": 1.535,
+      "lr": "2.00e-07",
+      "norm": 5.5169,
+      "step": 442
+    },
+    {
+      "acc_char": 0.3017,
+      "acc_token": 0.685,
+      "epoch": 0.01,
+      "loss_char": 0.7092,
+      "loss_token": 1.6507,
+      "lr": "2.00e-07",
+      "norm": 6.0394,
+      "step": 443
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6729,
+      "epoch": 0.01,
+      "loss_char": 0.7755,
+      "loss_token": 1.6699,
+      "lr": "2.00e-07",
+      "norm": 6.0785,
+      "step": 444
+    },
+    {
+      "acc_char": 0.2947,
+      "acc_token": 0.6726,
+      "epoch": 0.01,
+      "loss_char": 0.745,
+      "loss_token": 1.665,
+      "lr": "2.00e-07",
+      "norm": 5.4631,
+      "step": 445
+    },
+    {
+      "acc_char": 0.2996,
+      "acc_token": 0.6881,
+      "epoch": 0.01,
+      "loss_char": 0.6957,
+      "loss_token": 1.5646,
+      "lr": "2.00e-07",
+      "norm": 6.0473,
+      "step": 446
+    },
+    {
+      "acc_char": 0.3009,
+      "acc_token": 0.6825,
+      "epoch": 0.01,
+      "loss_char": 0.7218,
+      "loss_token": 1.6613,
+      "lr": "2.00e-07",
+      "norm": 5.7581,
+      "step": 447
+    },
+    {
+      "acc_char": 0.3038,
+      "acc_token": 0.6971,
+      "epoch": 0.01,
+      "loss_char": 0.6581,
+      "loss_token": 1.5529,
+      "lr": "2.00e-07",
+      "norm": 5.6224,
+      "step": 448
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6864,
+      "epoch": 0.01,
+      "loss_char": 0.7059,
+      "loss_token": 1.6413,
+      "lr": "2.00e-07",
+      "norm": 5.9119,
+      "step": 449
+    },
+    {
+      "acc_char": 0.3022,
+      "acc_token": 0.692,
+      "epoch": 0.01,
+      "loss_char": 0.6988,
+      "loss_token": 1.615,
+      "lr": "2.00e-07",
+      "norm": 5.6076,
+      "step": 450
+    },
+    {
+      "acc_char": 0.3073,
+      "acc_token": 0.6971,
+      "epoch": 0.01,
+      "loss_char": 0.6754,
+      "loss_token": 1.6428,
+      "lr": "2.00e-07",
+      "norm": 5.8904,
+      "step": 451
+    },
+    {
+      "acc_char": 0.3064,
+      "acc_token": 0.6801,
+      "epoch": 0.01,
+      "loss_char": 0.7328,
+      "loss_token": 1.631,
+      "lr": "2.00e-07",
+      "norm": 5.8551,
+      "step": 452
+    },
+    {
+      "acc_char": 0.3071,
+      "acc_token": 0.6928,
+      "epoch": 0.01,
+      "loss_char": 0.6906,
+      "loss_token": 1.5022,
+      "lr": "2.00e-07",
+      "norm": 5.4729,
+      "step": 453
+    },
+    {
+      "acc_char": 0.3208,
+      "acc_token": 0.7102,
+      "epoch": 0.01,
+      "loss_char": 0.6756,
+      "loss_token": 1.5461,
+      "lr": "2.00e-07",
+      "norm": 5.6754,
+      "step": 454
+    },
+    {
+      "acc_char": 0.3092,
+      "acc_token": 0.6937,
+      "epoch": 0.01,
+      "loss_char": 0.7156,
+      "loss_token": 1.6372,
+      "lr": "2.00e-07",
+      "norm": 6.0278,
+      "step": 455
+    },
+    {
+      "acc_char": 0.3199,
+      "acc_token": 0.7228,
+      "epoch": 0.01,
+      "loss_char": 0.6196,
+      "loss_token": 1.5193,
+      "lr": "2.00e-07",
+      "norm": 5.64,
+      "step": 456
+    },
+    {
+      "acc_char": 0.2966,
+      "acc_token": 0.6698,
+      "epoch": 0.01,
+      "loss_char": 0.7384,
+      "loss_token": 1.6731,
+      "lr": "2.00e-07",
+      "norm": 5.6788,
+      "step": 457
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6484,
+      "epoch": 0.01,
+      "loss_char": 0.8133,
+      "loss_token": 1.6824,
+      "lr": "2.00e-07",
+      "norm": 5.3832,
+      "step": 458
+    },
+    {
+      "acc_char": 0.2934,
+      "acc_token": 0.6721,
+      "epoch": 0.01,
+      "loss_char": 0.7316,
+      "loss_token": 1.729,
+      "lr": "2.00e-07",
+      "norm": 5.854,
+      "step": 459
+    },
+    {
+      "acc_char": 0.2848,
+      "acc_token": 0.619,
+      "epoch": 0.01,
+      "loss_char": 0.8858,
+      "loss_token": 1.7516,
+      "lr": "2.00e-07",
+      "norm": 5.5137,
+      "step": 460
+    },
+    {
+      "acc_char": 0.2997,
+      "acc_token": 0.6674,
+      "epoch": 0.01,
+      "loss_char": 0.7782,
+      "loss_token": 1.6639,
+      "lr": "2.00e-07",
+      "norm": 5.8666,
+      "step": 461
+    },
+    {
+      "acc_char": 0.3221,
+      "acc_token": 0.7197,
+      "epoch": 0.01,
+      "loss_char": 0.6214,
+      "loss_token": 1.4808,
+      "lr": "2.00e-07",
+      "norm": 5.3481,
+      "step": 462
+    },
+    {
+      "acc_char": 0.3061,
+      "acc_token": 0.6853,
+      "epoch": 0.01,
+      "loss_char": 0.7414,
+      "loss_token": 1.57,
+      "lr": "2.00e-07",
+      "norm": 5.7585,
+      "step": 463
+    },
+    {
+      "acc_char": 0.314,
+      "acc_token": 0.7022,
+      "epoch": 0.01,
+      "loss_char": 0.6737,
+      "loss_token": 1.5346,
+      "lr": "2.00e-07",
+      "norm": 5.5364,
+      "step": 464
+    },
+    {
+      "acc_char": 0.3145,
+      "acc_token": 0.7132,
+      "epoch": 0.01,
+      "loss_char": 0.6335,
+      "loss_token": 1.471,
+      "lr": "2.00e-07",
+      "norm": 5.0732,
+      "step": 465
+    },
+    {
+      "acc_char": 0.31,
+      "acc_token": 0.6899,
+      "epoch": 0.01,
+      "loss_char": 0.7093,
+      "loss_token": 1.6284,
+      "lr": "2.00e-07",
+      "norm": 5.6945,
+      "step": 466
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6829,
+      "epoch": 0.01,
+      "loss_char": 0.7209,
+      "loss_token": 1.6079,
+      "lr": "2.00e-07",
+      "norm": 5.8328,
+      "step": 467
+    },
+    {
+      "acc_char": 0.2971,
+      "acc_token": 0.6809,
+      "epoch": 0.01,
+      "loss_char": 0.7248,
+      "loss_token": 1.6548,
+      "lr": "2.00e-07",
+      "norm": 5.8493,
+      "step": 468
+    },
+    {
+      "acc_char": 0.309,
+      "acc_token": 0.7055,
+      "epoch": 0.01,
+      "loss_char": 0.6652,
+      "loss_token": 1.552,
+      "lr": "2.00e-07",
+      "norm": 5.7961,
+      "step": 469
+    },
+    {
+      "acc_char": 0.2997,
+      "acc_token": 0.6878,
+      "epoch": 0.01,
+      "loss_char": 0.6956,
+      "loss_token": 1.6944,
+      "lr": "2.00e-07",
+      "norm": 5.8907,
+      "step": 470
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6759,
+      "epoch": 0.01,
+      "loss_char": 0.7354,
+      "loss_token": 1.5767,
+      "lr": "2.00e-07",
+      "norm": 5.6017,
+      "step": 471
+    },
+    {
+      "acc_char": 0.2985,
+      "acc_token": 0.6779,
+      "epoch": 0.01,
+      "loss_char": 0.7313,
+      "loss_token": 1.6971,
+      "lr": "2.00e-07",
+      "norm": 5.7723,
+      "step": 472
+    },
+    {
+      "acc_char": 0.2867,
+      "acc_token": 0.6601,
+      "epoch": 0.01,
+      "loss_char": 0.7551,
+      "loss_token": 1.709,
+      "lr": "2.00e-07",
+      "norm": 5.6993,
+      "step": 473
+    },
+    {
+      "acc_char": 0.279,
+      "acc_token": 0.6141,
+      "epoch": 0.01,
+      "loss_char": 0.9209,
+      "loss_token": 1.8436,
+      "lr": "2.00e-07",
+      "norm": 5.3151,
+      "step": 474
+    },
+    {
+      "acc_char": 0.3006,
+      "acc_token": 0.6697,
+      "epoch": 0.01,
+      "loss_char": 0.7635,
+      "loss_token": 1.6591,
+      "lr": "2.00e-07",
+      "norm": 5.8138,
+      "step": 475
+    },
+    {
+      "acc_char": 0.3053,
+      "acc_token": 0.6845,
+      "epoch": 0.01,
+      "loss_char": 0.7014,
+      "loss_token": 1.5667,
+      "lr": "2.00e-07",
+      "norm": 5.6497,
+      "step": 476
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.6823,
+      "epoch": 0.01,
+      "loss_char": 0.7325,
+      "loss_token": 1.6252,
+      "lr": "2.00e-07",
+      "norm": 5.7174,
+      "step": 477
+    },
+    {
+      "acc_char": 0.3232,
+      "acc_token": 0.7062,
+      "epoch": 0.01,
+      "loss_char": 0.7192,
+      "loss_token": 1.6411,
+      "lr": "2.00e-07",
+      "norm": 5.4237,
+      "step": 478
+    },
+    {
+      "acc_char": 0.3098,
+      "acc_token": 0.6917,
+      "epoch": 0.01,
+      "loss_char": 0.708,
+      "loss_token": 1.6068,
+      "lr": "2.00e-07",
+      "norm": 5.5393,
+      "step": 479
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6798,
+      "epoch": 0.01,
+      "loss_char": 0.6922,
+      "loss_token": 1.4888,
+      "lr": "2.00e-07",
+      "norm": 5.7247,
+      "step": 480
+    },
+    {
+      "acc_char": 0.3042,
+      "acc_token": 0.6869,
+      "epoch": 0.01,
+      "loss_char": 0.7076,
+      "loss_token": 1.6158,
+      "lr": "2.00e-07",
+      "norm": 5.9854,
+      "step": 481
+    },
+    {
+      "acc_char": 0.3018,
+      "acc_token": 0.6875,
+      "epoch": 0.01,
+      "loss_char": 0.6984,
+      "loss_token": 1.6211,
+      "lr": "2.00e-07",
+      "norm": 5.8652,
+      "step": 482
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.68,
+      "epoch": 0.01,
+      "loss_char": 0.7306,
+      "loss_token": 1.6727,
+      "lr": "2.00e-07",
+      "norm": 5.6525,
+      "step": 483
+    },
+    {
+      "acc_char": 0.3038,
+      "acc_token": 0.6767,
+      "epoch": 0.01,
+      "loss_char": 0.7567,
+      "loss_token": 1.6292,
+      "lr": "2.00e-07",
+      "norm": 5.5359,
+      "step": 484
+    },
+    {
+      "acc_char": 0.3084,
+      "acc_token": 0.6953,
+      "epoch": 0.01,
+      "loss_char": 0.6971,
+      "loss_token": 1.5871,
+      "lr": "2.00e-07",
+      "norm": 5.9003,
+      "step": 485
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6804,
+      "epoch": 0.01,
+      "loss_char": 0.7237,
+      "loss_token": 1.6143,
+      "lr": "2.00e-07",
+      "norm": 5.6656,
+      "step": 486
+    },
+    {
+      "acc_char": 0.2974,
+      "acc_token": 0.6946,
+      "epoch": 0.01,
+      "loss_char": 0.6534,
+      "loss_token": 1.6319,
+      "lr": "2.00e-07",
+      "norm": 5.7585,
+      "step": 487
+    },
+    {
+      "acc_char": 0.2946,
+      "acc_token": 0.6734,
+      "epoch": 0.01,
+      "loss_char": 0.7431,
+      "loss_token": 1.6914,
+      "lr": "2.00e-07",
+      "norm": 5.6418,
+      "step": 488
+    },
+    {
+      "acc_char": 0.3119,
+      "acc_token": 0.7011,
+      "epoch": 0.01,
+      "loss_char": 0.6807,
+      "loss_token": 1.6073,
+      "lr": "2.00e-07",
+      "norm": 5.6436,
+      "step": 489
+    },
+    {
+      "acc_char": 0.2839,
+      "acc_token": 0.6155,
+      "epoch": 0.01,
+      "loss_char": 0.9091,
+      "loss_token": 1.8123,
+      "lr": "2.00e-07",
+      "norm": 5.6981,
+      "step": 490
+    },
+    {
+      "acc_char": 0.2984,
+      "acc_token": 0.6703,
+      "epoch": 0.01,
+      "loss_char": 0.7578,
+      "loss_token": 1.5599,
+      "lr": "2.00e-07",
+      "norm": 5.6947,
+      "step": 491
+    },
+    {
+      "acc_char": 0.3065,
+      "acc_token": 0.6484,
+      "epoch": 0.01,
+      "loss_char": 0.8497,
+      "loss_token": 1.6805,
+      "lr": "2.00e-07",
+      "norm": 5.4046,
+      "step": 492
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.679,
+      "epoch": 0.01,
+      "loss_char": 0.7339,
+      "loss_token": 1.6077,
+      "lr": "2.00e-07",
+      "norm": 5.4286,
+      "step": 493
+    },
+    {
+      "acc_char": 0.2944,
+      "acc_token": 0.6653,
+      "epoch": 0.01,
+      "loss_char": 0.7696,
+      "loss_token": 1.7296,
+      "lr": "2.00e-07",
+      "norm": 5.8803,
+      "step": 494
+    },
+    {
+      "acc_char": 0.3013,
+      "acc_token": 0.6859,
+      "epoch": 0.01,
+      "loss_char": 0.7333,
+      "loss_token": 1.6756,
+      "lr": "2.00e-07",
+      "norm": 5.6646,
+      "step": 495
+    },
+    {
+      "acc_char": 0.2916,
+      "acc_token": 0.6663,
+      "epoch": 0.01,
+      "loss_char": 0.7465,
+      "loss_token": 1.7513,
+      "lr": "2.00e-07",
+      "norm": 5.9079,
+      "step": 496
+    },
+    {
+      "acc_char": 0.3041,
+      "acc_token": 0.6897,
+      "epoch": 0.01,
+      "loss_char": 0.7028,
+      "loss_token": 1.5441,
+      "lr": "2.00e-07",
+      "norm": 5.4711,
+      "step": 497
+    },
+    {
+      "acc_char": 0.298,
+      "acc_token": 0.6945,
+      "epoch": 0.01,
+      "loss_char": 0.6889,
+      "loss_token": 1.5937,
+      "lr": "2.00e-07",
+      "norm": 5.9388,
+      "step": 498
+    },
+    {
+      "acc_char": 0.3092,
+      "acc_token": 0.6973,
+      "epoch": 0.01,
+      "loss_char": 0.6963,
+      "loss_token": 1.5762,
+      "lr": "2.00e-07",
+      "norm": 5.4772,
+      "step": 499
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.6663,
+      "epoch": 0.01,
+      "loss_char": 0.7789,
+      "loss_token": 1.6915,
+      "lr": "2.00e-07",
+      "norm": 5.8044,
+      "step": 500
+    },
+    {
+      "acc_char": 0.3078,
+      "acc_token": 0.6932,
+      "epoch": 0.01,
+      "loss_char": 0.7231,
+      "loss_token": 1.5793,
+      "lr": "2.00e-07",
+      "norm": 5.6492,
+      "step": 501
+    },
+    {
+      "acc_char": 0.2957,
+      "acc_token": 0.6777,
+      "epoch": 0.01,
+      "loss_char": 0.7369,
+      "loss_token": 1.6032,
+      "lr": "2.00e-07",
+      "norm": 5.6355,
+      "step": 502
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6943,
+      "epoch": 0.01,
+      "loss_char": 0.6721,
+      "loss_token": 1.6407,
+      "lr": "2.00e-07",
+      "norm": 5.3251,
+      "step": 503
+    },
+    {
+      "acc_char": 0.3039,
+      "acc_token": 0.6851,
+      "epoch": 0.01,
+      "loss_char": 0.7497,
+      "loss_token": 1.67,
+      "lr": "2.00e-07",
+      "norm": 5.6149,
+      "step": 504
+    },
+    {
+      "acc_char": 0.3165,
+      "acc_token": 0.7085,
+      "epoch": 0.01,
+      "loss_char": 0.6598,
+      "loss_token": 1.5165,
+      "lr": "2.00e-07",
+      "norm": 5.5233,
+      "step": 505
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.6898,
+      "epoch": 0.01,
+      "loss_char": 0.6862,
+      "loss_token": 1.6602,
+      "lr": "2.00e-07",
+      "norm": 5.7605,
+      "step": 506
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6698,
+      "epoch": 0.01,
+      "loss_char": 0.7666,
+      "loss_token": 1.7187,
+      "lr": "2.00e-07",
+      "norm": 6.0065,
+      "step": 507
+    },
+    {
+      "acc_char": 0.2914,
+      "acc_token": 0.6369,
+      "epoch": 0.01,
+      "loss_char": 0.8238,
+      "loss_token": 1.7726,
+      "lr": "2.00e-07",
+      "norm": 5.4933,
+      "step": 508
+    },
+    {
+      "acc_char": 0.3007,
+      "acc_token": 0.6841,
+      "epoch": 0.01,
+      "loss_char": 0.7153,
+      "loss_token": 1.6288,
+      "lr": "2.00e-07",
+      "norm": 5.4807,
+      "step": 509
+    },
+    {
+      "acc_char": 0.2938,
+      "acc_token": 0.6602,
+      "epoch": 0.01,
+      "loss_char": 0.7827,
+      "loss_token": 1.7307,
+      "lr": "2.00e-07",
+      "norm": 5.9197,
+      "step": 510
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.6911,
+      "epoch": 0.01,
+      "loss_char": 0.6722,
+      "loss_token": 1.5813,
+      "lr": "2.00e-07",
+      "norm": 5.7085,
+      "step": 511
+    },
+    {
+      "acc_char": 0.3012,
+      "acc_token": 0.6763,
+      "epoch": 0.01,
+      "loss_char": 0.7463,
+      "loss_token": 1.6278,
+      "lr": "2.00e-07",
+      "norm": 5.6667,
+      "step": 512
+    },
+    {
+      "acc_char": 0.3061,
+      "acc_token": 0.682,
+      "epoch": 0.01,
+      "loss_char": 0.7469,
+      "loss_token": 1.6617,
+      "lr": "2.00e-07",
+      "norm": 5.6931,
+      "step": 513
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6861,
+      "epoch": 0.01,
+      "loss_char": 0.7019,
+      "loss_token": 1.4865,
+      "lr": "2.00e-07",
+      "norm": 5.4473,
+      "step": 514
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6729,
+      "epoch": 0.01,
+      "loss_char": 0.7615,
+      "loss_token": 1.6468,
+      "lr": "2.00e-07",
+      "norm": 5.6802,
+      "step": 515
+    },
+    {
+      "acc_char": 0.2938,
+      "acc_token": 0.6499,
+      "epoch": 0.01,
+      "loss_char": 0.8332,
+      "loss_token": 1.8228,
+      "lr": "2.00e-07",
+      "norm": 6.4352,
+      "step": 516
+    },
+    {
+      "acc_char": 0.3015,
+      "acc_token": 0.675,
+      "epoch": 0.01,
+      "loss_char": 0.7496,
+      "loss_token": 1.6712,
+      "lr": "2.00e-07",
+      "norm": 6.3257,
+      "step": 517
+    },
+    {
+      "acc_char": 0.2868,
+      "acc_token": 0.6577,
+      "epoch": 0.01,
+      "loss_char": 0.8048,
+      "loss_token": 1.7543,
+      "lr": "2.00e-07",
+      "norm": 5.8648,
+      "step": 518
+    },
+    {
+      "acc_char": 0.3102,
+      "acc_token": 0.6952,
+      "epoch": 0.01,
+      "loss_char": 0.7037,
+      "loss_token": 1.6965,
+      "lr": "2.00e-07",
+      "norm": 6.092,
+      "step": 519
+    },
+    {
+      "acc_char": 0.3043,
+      "acc_token": 0.6834,
+      "epoch": 0.01,
+      "loss_char": 0.7227,
+      "loss_token": 1.6312,
+      "lr": "2.00e-07",
+      "norm": 5.761,
+      "step": 520
+    },
+    {
+      "acc_char": 0.2938,
+      "acc_token": 0.6543,
+      "epoch": 0.01,
+      "loss_char": 0.8214,
+      "loss_token": 1.7083,
+      "lr": "2.00e-07",
+      "norm": 5.9636,
+      "step": 521
+    },
+    {
+      "acc_char": 0.3135,
+      "acc_token": 0.7087,
+      "epoch": 0.01,
+      "loss_char": 0.6632,
+      "loss_token": 1.6053,
+      "lr": "2.00e-07",
+      "norm": 5.7167,
+      "step": 522
+    },
+    {
+      "acc_char": 0.309,
+      "acc_token": 0.6883,
+      "epoch": 0.01,
+      "loss_char": 0.7156,
+      "loss_token": 1.6225,
+      "lr": "2.00e-07",
+      "norm": 5.7743,
+      "step": 523
+    },
+    {
+      "acc_char": 0.2965,
+      "acc_token": 0.6721,
+      "epoch": 0.01,
+      "loss_char": 0.7391,
+      "loss_token": 1.7042,
+      "lr": "2.00e-07",
+      "norm": 5.8933,
+      "step": 524
+    },
+    {
+      "acc_char": 0.3038,
+      "acc_token": 0.6833,
+      "epoch": 0.01,
+      "loss_char": 0.7208,
+      "loss_token": 1.6832,
+      "lr": "2.00e-07",
+      "norm": 5.7907,
+      "step": 525
+    },
+    {
+      "acc_char": 0.3054,
+      "acc_token": 0.6899,
+      "epoch": 0.01,
+      "loss_char": 0.7107,
+      "loss_token": 1.5754,
+      "lr": "2.00e-07",
+      "norm": 5.5543,
+      "step": 526
+    },
+    {
+      "acc_char": 0.3021,
+      "acc_token": 0.6727,
+      "epoch": 0.01,
+      "loss_char": 0.7716,
+      "loss_token": 1.6425,
+      "lr": "2.00e-07",
+      "norm": 6.5612,
+      "step": 527
+    },
+    {
+      "acc_char": 0.2987,
+      "acc_token": 0.6779,
+      "epoch": 0.01,
+      "loss_char": 0.7249,
+      "loss_token": 1.5985,
+      "lr": "2.00e-07",
+      "norm": 5.6789,
+      "step": 528
+    },
+    {
+      "acc_char": 0.2996,
+      "acc_token": 0.6786,
+      "epoch": 0.01,
+      "loss_char": 0.7245,
+      "loss_token": 1.6649,
+      "lr": "2.00e-07",
+      "norm": 5.9216,
+      "step": 529
+    },
+    {
+      "acc_char": 0.3084,
+      "acc_token": 0.6999,
+      "epoch": 0.01,
+      "loss_char": 0.6842,
+      "loss_token": 1.5994,
+      "lr": "2.00e-07",
+      "norm": 5.4386,
+      "step": 530
+    },
+    {
+      "acc_char": 0.3033,
+      "acc_token": 0.6736,
+      "epoch": 0.01,
+      "loss_char": 0.7615,
+      "loss_token": 1.5393,
+      "lr": "2.00e-07",
+      "norm": 5.4897,
+      "step": 531
+    },
+    {
+      "acc_char": 0.2958,
+      "acc_token": 0.6759,
+      "epoch": 0.01,
+      "loss_char": 0.7476,
+      "loss_token": 1.5885,
+      "lr": "2.00e-07",
+      "norm": 5.529,
+      "step": 532
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.6895,
+      "epoch": 0.01,
+      "loss_char": 0.6842,
+      "loss_token": 1.5783,
+      "lr": "2.00e-07",
+      "norm": 5.5189,
+      "step": 533
+    },
+    {
+      "acc_char": 0.2983,
+      "acc_token": 0.6761,
+      "epoch": 0.01,
+      "loss_char": 0.7425,
+      "loss_token": 1.5862,
+      "lr": "2.00e-07",
+      "norm": 5.5887,
+      "step": 534
+    },
+    {
+      "acc_char": 0.3002,
+      "acc_token": 0.692,
+      "epoch": 0.01,
+      "loss_char": 0.6575,
+      "loss_token": 1.5723,
+      "lr": "2.00e-07",
+      "norm": 5.3558,
+      "step": 535
+    },
+    {
+      "acc_char": 0.298,
+      "acc_token": 0.676,
+      "epoch": 0.01,
+      "loss_char": 0.7515,
+      "loss_token": 1.6691,
+      "lr": "2.00e-07",
+      "norm": 6.0115,
+      "step": 536
+    },
+    {
+      "acc_char": 0.2924,
+      "acc_token": 0.674,
+      "epoch": 0.01,
+      "loss_char": 0.7274,
+      "loss_token": 1.611,
+      "lr": "2.00e-07",
+      "norm": 5.4644,
+      "step": 537
+    },
+    {
+      "acc_char": 0.3046,
+      "acc_token": 0.6955,
+      "epoch": 0.01,
+      "loss_char": 0.6839,
+      "loss_token": 1.6216,
+      "lr": "2.00e-07",
+      "norm": 5.4058,
+      "step": 538
+    },
+    {
+      "acc_char": 0.3,
+      "acc_token": 0.6712,
+      "epoch": 0.01,
+      "loss_char": 0.7657,
+      "loss_token": 1.745,
+      "lr": "2.00e-07",
+      "norm": 5.8636,
+      "step": 539
+    },
+    {
+      "acc_char": 0.3209,
+      "acc_token": 0.7264,
+      "epoch": 0.01,
+      "loss_char": 0.6182,
+      "loss_token": 1.5795,
+      "lr": "2.00e-07",
+      "norm": 5.4199,
+      "step": 540
+    },
+    {
+      "acc_char": 0.291,
+      "acc_token": 0.6599,
+      "epoch": 0.01,
+      "loss_char": 0.7919,
+      "loss_token": 1.7544,
+      "lr": "2.00e-07",
+      "norm": 5.7709,
+      "step": 541
+    },
+    {
+      "acc_char": 0.3051,
+      "acc_token": 0.6966,
+      "epoch": 0.01,
+      "loss_char": 0.6722,
+      "loss_token": 1.5357,
+      "lr": "2.00e-07",
+      "norm": 5.7007,
+      "step": 542
+    },
+    {
+      "acc_char": 0.3207,
+      "acc_token": 0.6919,
+      "epoch": 0.01,
+      "loss_char": 0.6804,
+      "loss_token": 1.5393,
+      "lr": "2.00e-07",
+      "norm": 6.2127,
+      "step": 543
+    },
+    {
+      "acc_char": 0.316,
+      "acc_token": 0.705,
+      "epoch": 0.01,
+      "loss_char": 0.6765,
+      "loss_token": 1.5082,
+      "lr": "2.00e-07",
+      "norm": 5.3352,
+      "step": 544
+    },
+    {
+      "acc_char": 0.3092,
+      "acc_token": 0.6964,
+      "epoch": 0.01,
+      "loss_char": 0.6843,
+      "loss_token": 1.5882,
+      "lr": "2.00e-07",
+      "norm": 5.6973,
+      "step": 545
+    },
+    {
+      "acc_char": 0.2907,
+      "acc_token": 0.6619,
+      "epoch": 0.01,
+      "loss_char": 0.7776,
+      "loss_token": 1.7012,
+      "lr": "2.00e-07",
+      "norm": 5.6625,
+      "step": 546
+    },
+    {
+      "acc_char": 0.3011,
+      "acc_token": 0.6715,
+      "epoch": 0.01,
+      "loss_char": 0.7724,
+      "loss_token": 1.7237,
+      "lr": "2.00e-07",
+      "norm": 5.8425,
+      "step": 547
+    },
+    {
+      "acc_char": 0.292,
+      "acc_token": 0.6715,
+      "epoch": 0.01,
+      "loss_char": 0.7631,
+      "loss_token": 1.6806,
+      "lr": "2.00e-07",
+      "norm": 5.5621,
+      "step": 548
+    },
+    {
+      "acc_char": 0.3137,
+      "acc_token": 0.7052,
+      "epoch": 0.01,
+      "loss_char": 0.6651,
+      "loss_token": 1.5382,
+      "lr": "2.00e-07",
+      "norm": 5.3519,
+      "step": 549
+    },
+    {
+      "acc_char": 0.3066,
+      "acc_token": 0.6953,
+      "epoch": 0.01,
+      "loss_char": 0.684,
+      "loss_token": 1.6192,
+      "lr": "2.00e-07",
+      "norm": 5.5761,
+      "step": 550
+    },
+    {
+      "acc_char": 0.3174,
+      "acc_token": 0.7004,
+      "epoch": 0.01,
+      "loss_char": 0.6984,
+      "loss_token": 1.5912,
+      "lr": "2.00e-07",
+      "norm": 5.7781,
+      "step": 551
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6731,
+      "epoch": 0.01,
+      "loss_char": 0.7512,
+      "loss_token": 1.6747,
+      "lr": "2.00e-07",
+      "norm": 5.8476,
+      "step": 552
+    },
+    {
+      "acc_char": 0.2734,
+      "acc_token": 0.6097,
+      "epoch": 0.01,
+      "loss_char": 0.8934,
+      "loss_token": 1.8362,
+      "lr": "2.00e-07",
+      "norm": 5.7851,
+      "step": 553
+    },
+    {
+      "acc_char": 0.3226,
+      "acc_token": 0.7002,
+      "epoch": 0.01,
+      "loss_char": 0.6766,
+      "loss_token": 1.5276,
+      "lr": "2.00e-07",
+      "norm": 5.8662,
+      "step": 554
+    },
+    {
+      "acc_char": 0.3191,
+      "acc_token": 0.7262,
+      "epoch": 0.01,
+      "loss_char": 0.6071,
+      "loss_token": 1.4902,
+      "lr": "2.00e-07",
+      "norm": 5.7735,
+      "step": 555
+    },
+    {
+      "acc_char": 0.2984,
+      "acc_token": 0.6792,
+      "epoch": 0.01,
+      "loss_char": 0.7263,
+      "loss_token": 1.6368,
+      "lr": "2.00e-07",
+      "norm": 5.5514,
+      "step": 556
+    },
+    {
+      "acc_char": 0.3112,
+      "acc_token": 0.6938,
+      "epoch": 0.01,
+      "loss_char": 0.7188,
+      "loss_token": 1.5125,
+      "lr": "2.00e-07",
+      "norm": 5.6774,
+      "step": 557
+    },
+    {
+      "acc_char": 0.3056,
+      "acc_token": 0.6932,
+      "epoch": 0.01,
+      "loss_char": 0.683,
+      "loss_token": 1.5882,
+      "lr": "2.00e-07",
+      "norm": 5.5219,
+      "step": 558
+    },
+    {
+      "acc_char": 0.3024,
+      "acc_token": 0.6932,
+      "epoch": 0.01,
+      "loss_char": 0.7094,
+      "loss_token": 1.6385,
+      "lr": "2.00e-07",
+      "norm": 5.9762,
+      "step": 559
+    },
+    {
+      "acc_char": 0.2901,
+      "acc_token": 0.664,
+      "epoch": 0.01,
+      "loss_char": 0.7806,
+      "loss_token": 1.7459,
+      "lr": "2.00e-07",
+      "norm": 6.0142,
+      "step": 560
+    },
+    {
+      "acc_char": 0.3097,
+      "acc_token": 0.6902,
+      "epoch": 0.01,
+      "loss_char": 0.7173,
+      "loss_token": 1.6327,
+      "lr": "2.00e-07",
+      "norm": 6.0807,
+      "step": 561
+    },
+    {
+      "acc_char": 0.3026,
+      "acc_token": 0.6837,
+      "epoch": 0.01,
+      "loss_char": 0.7394,
+      "loss_token": 1.6375,
+      "lr": "2.00e-07",
+      "norm": 5.6845,
+      "step": 562
+    },
+    {
+      "acc_char": 0.296,
+      "acc_token": 0.676,
+      "epoch": 0.01,
+      "loss_char": 0.7309,
+      "loss_token": 1.6614,
+      "lr": "2.00e-07",
+      "norm": 5.6079,
+      "step": 563
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.6811,
+      "epoch": 0.01,
+      "loss_char": 0.7016,
+      "loss_token": 1.5666,
+      "lr": "2.00e-07",
+      "norm": 5.4709,
+      "step": 564
+    },
+    {
+      "acc_char": 0.3028,
+      "acc_token": 0.6891,
+      "epoch": 0.01,
+      "loss_char": 0.7079,
+      "loss_token": 1.6339,
+      "lr": "2.00e-07",
+      "norm": 5.6625,
+      "step": 565
+    },
+    {
+      "acc_char": 0.2735,
+      "acc_token": 0.6008,
+      "epoch": 0.01,
+      "loss_char": 0.9197,
+      "loss_token": 1.8075,
+      "lr": "2.00e-07",
+      "norm": 5.5777,
+      "step": 566
+    },
+    {
+      "acc_char": 0.3075,
+      "acc_token": 0.6919,
+      "epoch": 0.01,
+      "loss_char": 0.7049,
+      "loss_token": 1.5574,
+      "lr": "2.00e-07",
+      "norm": 5.8756,
+      "step": 567
+    },
+    {
+      "acc_char": 0.3068,
+      "acc_token": 0.6863,
+      "epoch": 0.01,
+      "loss_char": 0.7152,
+      "loss_token": 1.7041,
+      "lr": "2.00e-07",
+      "norm": 5.9522,
+      "step": 568
+    },
+    {
+      "acc_char": 0.2755,
+      "acc_token": 0.6058,
+      "epoch": 0.01,
+      "loss_char": 0.9253,
+      "loss_token": 1.8017,
+      "lr": "2.00e-07",
+      "norm": 5.3494,
+      "step": 569
+    },
+    {
+      "acc_char": 0.295,
+      "acc_token": 0.6807,
+      "epoch": 0.01,
+      "loss_char": 0.709,
+      "loss_token": 1.6612,
+      "lr": "2.00e-07",
+      "norm": 5.6727,
+      "step": 570
+    },
+    {
+      "acc_char": 0.3051,
+      "acc_token": 0.6734,
+      "epoch": 0.01,
+      "loss_char": 0.7934,
+      "loss_token": 1.6827,
+      "lr": "2.00e-07",
+      "norm": 5.8125,
+      "step": 571
+    },
+    {
+      "acc_char": 0.2914,
+      "acc_token": 0.6652,
+      "epoch": 0.01,
+      "loss_char": 0.7623,
+      "loss_token": 1.6601,
+      "lr": "2.00e-07",
+      "norm": 5.7307,
+      "step": 572
+    },
+    {
+      "acc_char": 0.2923,
+      "acc_token": 0.674,
+      "epoch": 0.01,
+      "loss_char": 0.7213,
+      "loss_token": 1.7202,
+      "lr": "2.00e-07",
+      "norm": 6.0228,
+      "step": 573
+    },
+    {
+      "acc_char": 0.3059,
+      "acc_token": 0.6905,
+      "epoch": 0.01,
+      "loss_char": 0.7063,
+      "loss_token": 1.6059,
+      "lr": "2.00e-07",
+      "norm": 5.7055,
+      "step": 574
+    },
+    {
+      "acc_char": 0.2859,
+      "acc_token": 0.6212,
+      "epoch": 0.01,
+      "loss_char": 0.8812,
+      "loss_token": 1.7594,
+      "lr": "2.00e-07",
+      "norm": 5.642,
+      "step": 575
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6935,
+      "epoch": 0.01,
+      "loss_char": 0.6684,
+      "loss_token": 1.5689,
+      "lr": "2.00e-07",
+      "norm": 5.6313,
+      "step": 576
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.7125,
+      "epoch": 0.01,
+      "loss_char": 0.6309,
+      "loss_token": 1.593,
+      "lr": "2.00e-07",
+      "norm": 5.3535,
+      "step": 577
+    },
+    {
+      "acc_char": 0.299,
+      "acc_token": 0.6642,
+      "epoch": 0.01,
+      "loss_char": 0.7921,
+      "loss_token": 1.662,
+      "lr": "2.00e-07",
+      "norm": 5.6795,
+      "step": 578
+    },
+    {
+      "acc_char": 0.3058,
+      "acc_token": 0.6969,
+      "epoch": 0.01,
+      "loss_char": 0.6817,
+      "loss_token": 1.6438,
+      "lr": "2.00e-07",
+      "norm": 5.6116,
+      "step": 579
+    },
+    {
+      "acc_char": 0.292,
+      "acc_token": 0.6723,
+      "epoch": 0.01,
+      "loss_char": 0.7104,
+      "loss_token": 1.6338,
+      "lr": "2.00e-07",
+      "norm": 5.7615,
+      "step": 580
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.6732,
+      "epoch": 0.01,
+      "loss_char": 0.7542,
+      "loss_token": 1.6422,
+      "lr": "2.00e-07",
+      "norm": 5.9272,
+      "step": 581
+    },
+    {
+      "acc_char": 0.2948,
+      "acc_token": 0.6641,
+      "epoch": 0.01,
+      "loss_char": 0.781,
+      "loss_token": 1.6998,
+      "lr": "2.00e-07",
+      "norm": 5.6837,
+      "step": 582
+    },
+    {
+      "acc_char": 0.3182,
+      "acc_token": 0.7073,
+      "epoch": 0.01,
+      "loss_char": 0.6854,
+      "loss_token": 1.5821,
+      "lr": "2.00e-07",
+      "norm": 5.727,
+      "step": 583
+    },
+    {
+      "acc_char": 0.2881,
+      "acc_token": 0.6246,
+      "epoch": 0.01,
+      "loss_char": 0.8612,
+      "loss_token": 1.7089,
+      "lr": "2.00e-07",
+      "norm": 5.4477,
+      "step": 584
+    },
+    {
+      "acc_char": 0.3289,
+      "acc_token": 0.6926,
+      "epoch": 0.01,
+      "loss_char": 0.7151,
+      "loss_token": 1.4939,
+      "lr": "2.00e-07",
+      "norm": 5.2305,
+      "step": 585
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6918,
+      "epoch": 0.01,
+      "loss_char": 0.6788,
+      "loss_token": 1.5943,
+      "lr": "2.00e-07",
+      "norm": 5.6058,
+      "step": 586
+    },
+    {
+      "acc_char": 0.2987,
+      "acc_token": 0.6822,
+      "epoch": 0.01,
+      "loss_char": 0.7151,
+      "loss_token": 1.67,
+      "lr": "2.00e-07",
+      "norm": 5.5937,
+      "step": 587
+    },
+    {
+      "acc_char": 0.3001,
+      "acc_token": 0.6367,
+      "epoch": 0.01,
+      "loss_char": 0.8631,
+      "loss_token": 1.7495,
+      "lr": "2.00e-07",
+      "norm": 8.798,
+      "step": 588
+    },
+    {
+      "acc_char": 0.2933,
+      "acc_token": 0.6781,
+      "epoch": 0.01,
+      "loss_char": 0.7214,
+      "loss_token": 1.6052,
+      "lr": "2.00e-07",
+      "norm": 5.4492,
+      "step": 589
+    },
+    {
+      "acc_char": 0.3158,
+      "acc_token": 0.6998,
+      "epoch": 0.01,
+      "loss_char": 0.7083,
+      "loss_token": 1.5872,
+      "lr": "2.00e-07",
+      "norm": 5.6008,
+      "step": 590
+    },
+    {
+      "acc_char": 0.3111,
+      "acc_token": 0.6977,
+      "epoch": 0.01,
+      "loss_char": 0.6923,
+      "loss_token": 1.6226,
+      "lr": "2.00e-07",
+      "norm": 5.7109,
+      "step": 591
+    },
+    {
+      "acc_char": 0.2877,
+      "acc_token": 0.6725,
+      "epoch": 0.01,
+      "loss_char": 0.7402,
+      "loss_token": 1.6398,
+      "lr": "2.00e-07",
+      "norm": 5.629,
+      "step": 592
+    },
+    {
+      "acc_char": 0.2991,
+      "acc_token": 0.6806,
+      "epoch": 0.01,
+      "loss_char": 0.7446,
+      "loss_token": 1.6434,
+      "lr": "2.00e-07",
+      "norm": 6.1502,
+      "step": 593
+    },
+    {
+      "acc_char": 0.31,
+      "acc_token": 0.6981,
+      "epoch": 0.01,
+      "loss_char": 0.6608,
+      "loss_token": 1.5804,
+      "lr": "2.00e-07",
+      "norm": 5.528,
+      "step": 594
+    },
+    {
+      "acc_char": 0.3065,
+      "acc_token": 0.6986,
+      "epoch": 0.01,
+      "loss_char": 0.6774,
+      "loss_token": 1.5434,
+      "lr": "2.00e-07",
+      "norm": 5.4188,
+      "step": 595
+    },
+    {
+      "acc_char": 0.2961,
+      "acc_token": 0.6391,
+      "epoch": 0.01,
+      "loss_char": 0.851,
+      "loss_token": 1.6733,
+      "lr": "2.00e-07",
+      "norm": 5.2518,
+      "step": 596
+    },
+    {
+      "acc_char": 0.2991,
+      "acc_token": 0.6732,
+      "epoch": 0.01,
+      "loss_char": 0.7501,
+      "loss_token": 1.6025,
+      "lr": "2.00e-07",
+      "norm": 5.7633,
+      "step": 597
+    },
+    {
+      "acc_char": 0.3057,
+      "acc_token": 0.6638,
+      "epoch": 0.01,
+      "loss_char": 0.7828,
+      "loss_token": 1.6712,
+      "lr": "2.00e-07",
+      "norm": 5.4425,
+      "step": 598
+    },
+    {
+      "acc_char": 0.3001,
+      "acc_token": 0.6873,
+      "epoch": 0.01,
+      "loss_char": 0.6919,
+      "loss_token": 1.6011,
+      "lr": "2.00e-07",
+      "norm": 5.5889,
+      "step": 599
+    },
+    {
+      "acc_char": 0.3081,
+      "acc_token": 0.6991,
+      "epoch": 0.01,
+      "loss_char": 0.6603,
+      "loss_token": 1.5127,
+      "lr": "2.00e-07",
+      "norm": 5.6657,
+      "step": 600
+    },
+    {
+      "acc_char": 0.3022,
+      "acc_token": 0.6785,
+      "epoch": 0.01,
+      "loss_char": 0.7635,
+      "loss_token": 1.6259,
+      "lr": "2.00e-07",
+      "norm": 5.6799,
+      "step": 601
+    },
+    {
+      "acc_char": 0.3026,
+      "acc_token": 0.6963,
+      "epoch": 0.01,
+      "loss_char": 0.6647,
+      "loss_token": 1.5104,
+      "lr": "2.00e-07",
+      "norm": 5.58,
+      "step": 602
+    },
+    {
+      "acc_char": 0.3058,
+      "acc_token": 0.6903,
+      "epoch": 0.01,
+      "loss_char": 0.7082,
+      "loss_token": 1.513,
+      "lr": "2.00e-07",
+      "norm": 5.5287,
+      "step": 603
+    },
+    {
+      "acc_char": 0.2897,
+      "acc_token": 0.6295,
+      "epoch": 0.01,
+      "loss_char": 0.8635,
+      "loss_token": 1.744,
+      "lr": "2.00e-07",
+      "norm": 5.838,
+      "step": 604
+    },
+    {
+      "acc_char": 0.3073,
+      "acc_token": 0.7038,
+      "epoch": 0.01,
+      "loss_char": 0.6604,
+      "loss_token": 1.5419,
+      "lr": "2.00e-07",
+      "norm": 5.6741,
+      "step": 605
+    },
+    {
+      "acc_char": 0.2967,
+      "acc_token": 0.6721,
+      "epoch": 0.01,
+      "loss_char": 0.7576,
+      "loss_token": 1.664,
+      "lr": "2.00e-07",
+      "norm": 6.1815,
+      "step": 606
+    },
+    {
+      "acc_char": 0.2971,
+      "acc_token": 0.6848,
+      "epoch": 0.01,
+      "loss_char": 0.7164,
+      "loss_token": 1.7081,
+      "lr": "2.00e-07",
+      "norm": 5.8114,
+      "step": 607
+    },
+    {
+      "acc_char": 0.2944,
+      "acc_token": 0.6658,
+      "epoch": 0.01,
+      "loss_char": 0.7991,
+      "loss_token": 1.7166,
+      "lr": "2.00e-07",
+      "norm": 5.7798,
+      "step": 608
+    },
+    {
+      "acc_char": 0.2876,
+      "acc_token": 0.6315,
+      "epoch": 0.01,
+      "loss_char": 0.8615,
+      "loss_token": 1.756,
+      "lr": "2.00e-07",
+      "norm": 5.2967,
+      "step": 609
+    },
+    {
+      "acc_char": 0.3101,
+      "acc_token": 0.7015,
+      "epoch": 0.01,
+      "loss_char": 0.6474,
+      "loss_token": 1.5666,
+      "lr": "2.00e-07",
+      "norm": 5.5296,
+      "step": 610
+    },
+    {
+      "acc_char": 0.3071,
+      "acc_token": 0.7117,
+      "epoch": 0.01,
+      "loss_char": 0.609,
+      "loss_token": 1.3597,
+      "lr": "2.00e-07",
+      "norm": 5.9792,
+      "step": 611
+    },
+    {
+      "acc_char": 0.3121,
+      "acc_token": 0.7013,
+      "epoch": 0.01,
+      "loss_char": 0.6796,
+      "loss_token": 1.5647,
+      "lr": "2.00e-07",
+      "norm": 5.7678,
+      "step": 612
+    },
+    {
+      "acc_char": 0.3108,
+      "acc_token": 0.7027,
+      "epoch": 0.01,
+      "loss_char": 0.6619,
+      "loss_token": 1.5006,
+      "lr": "2.00e-07",
+      "norm": 5.5738,
+      "step": 613
+    },
+    {
+      "acc_char": 0.3075,
+      "acc_token": 0.6938,
+      "epoch": 0.01,
+      "loss_char": 0.6819,
+      "loss_token": 1.5156,
+      "lr": "2.00e-07",
+      "norm": 5.315,
+      "step": 614
+    },
+    {
+      "acc_char": 0.3043,
+      "acc_token": 0.6782,
+      "epoch": 0.01,
+      "loss_char": 0.744,
+      "loss_token": 1.6767,
+      "lr": "2.00e-07",
+      "norm": 6.1106,
+      "step": 615
+    },
+    {
+      "acc_char": 0.2944,
+      "acc_token": 0.6681,
+      "epoch": 0.01,
+      "loss_char": 0.8069,
+      "loss_token": 1.7487,
+      "lr": "2.00e-07",
+      "norm": 6.015,
+      "step": 616
+    },
+    {
+      "acc_char": 0.3186,
+      "acc_token": 0.7171,
+      "epoch": 0.01,
+      "loss_char": 0.642,
+      "loss_token": 1.5414,
+      "lr": "2.00e-07",
+      "norm": 5.3756,
+      "step": 617
+    },
+    {
+      "acc_char": 0.3015,
+      "acc_token": 0.679,
+      "epoch": 0.01,
+      "loss_char": 0.7143,
+      "loss_token": 1.6086,
+      "lr": "2.00e-07",
+      "norm": 5.9773,
+      "step": 618
+    },
+    {
+      "acc_char": 0.3277,
+      "acc_token": 0.7303,
+      "epoch": 0.01,
+      "loss_char": 0.6219,
+      "loss_token": 1.5241,
+      "lr": "2.00e-07",
+      "norm": 5.9241,
+      "step": 619
+    },
+    {
+      "acc_char": 0.3184,
+      "acc_token": 0.7167,
+      "epoch": 0.01,
+      "loss_char": 0.6294,
+      "loss_token": 1.525,
+      "lr": "2.00e-07",
+      "norm": 5.6258,
+      "step": 620
+    },
+    {
+      "acc_char": 0.2997,
+      "acc_token": 0.6806,
+      "epoch": 0.01,
+      "loss_char": 0.7262,
+      "loss_token": 1.6419,
+      "lr": "2.00e-07",
+      "norm": 5.6224,
+      "step": 621
+    },
+    {
+      "acc_char": 0.2978,
+      "acc_token": 0.6767,
+      "epoch": 0.01,
+      "loss_char": 0.7415,
+      "loss_token": 1.6946,
+      "lr": "2.00e-07",
+      "norm": 5.9768,
+      "step": 622
+    },
+    {
+      "acc_char": 0.2806,
+      "acc_token": 0.6274,
+      "epoch": 0.01,
+      "loss_char": 0.8257,
+      "loss_token": 1.6916,
+      "lr": "2.00e-07",
+      "norm": 5.191,
+      "step": 623
+    },
+    {
+      "acc_char": 0.3144,
+      "acc_token": 0.7105,
+      "epoch": 0.01,
+      "loss_char": 0.6586,
+      "loss_token": 1.5043,
+      "lr": "2.00e-07",
+      "norm": 5.4252,
+      "step": 624
+    },
+    {
+      "acc_char": 0.2839,
+      "acc_token": 0.6208,
+      "epoch": 0.01,
+      "loss_char": 0.8839,
+      "loss_token": 1.7884,
+      "lr": "2.00e-07",
+      "norm": 5.5514,
+      "step": 625
+    },
+    {
+      "acc_char": 0.3133,
+      "acc_token": 0.6975,
+      "epoch": 0.01,
+      "loss_char": 0.689,
+      "loss_token": 1.6005,
+      "lr": "2.00e-07",
+      "norm": 5.5656,
+      "step": 626
+    },
+    {
+      "acc_char": 0.2983,
+      "acc_token": 0.6831,
+      "epoch": 0.01,
+      "loss_char": 0.7115,
+      "loss_token": 1.5559,
+      "lr": "2.00e-07",
+      "norm": 5.4372,
+      "step": 627
+    },
+    {
+      "acc_char": 0.2926,
+      "acc_token": 0.6629,
+      "epoch": 0.01,
+      "loss_char": 0.7945,
+      "loss_token": 1.6611,
+      "lr": "2.00e-07",
+      "norm": 5.6938,
+      "step": 628
+    },
+    {
+      "acc_char": 0.2975,
+      "acc_token": 0.6745,
+      "epoch": 0.01,
+      "loss_char": 0.7395,
+      "loss_token": 1.6412,
+      "lr": "2.00e-07",
+      "norm": 5.7502,
+      "step": 629
+    },
+    {
+      "acc_char": 0.3138,
+      "acc_token": 0.7,
+      "epoch": 0.01,
+      "loss_char": 0.6856,
+      "loss_token": 1.6453,
+      "lr": "2.00e-07",
+      "norm": 5.8189,
+      "step": 630
+    },
+    {
+      "acc_char": 0.3096,
+      "acc_token": 0.6971,
+      "epoch": 0.01,
+      "loss_char": 0.7,
+      "loss_token": 1.576,
+      "lr": "2.00e-07",
+      "norm": 5.8166,
+      "step": 631
+    },
+    {
+      "acc_char": 0.3025,
+      "acc_token": 0.6913,
+      "epoch": 0.01,
+      "loss_char": 0.6716,
+      "loss_token": 1.623,
+      "lr": "2.00e-07",
+      "norm": 5.5634,
+      "step": 632
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6753,
+      "epoch": 0.01,
+      "loss_char": 0.7604,
+      "loss_token": 1.5291,
+      "lr": "2.00e-07",
+      "norm": 5.5975,
+      "step": 633
+    },
+    {
+      "acc_char": 0.3103,
+      "acc_token": 0.6774,
+      "epoch": 0.01,
+      "loss_char": 0.7568,
+      "loss_token": 1.6828,
+      "lr": "2.00e-07",
+      "norm": 6.1007,
+      "step": 634
+    },
+    {
+      "acc_char": 0.2999,
+      "acc_token": 0.674,
+      "epoch": 0.01,
+      "loss_char": 0.7689,
+      "loss_token": 1.6685,
+      "lr": "2.00e-07",
+      "norm": 5.8519,
+      "step": 635
+    },
+    {
+      "acc_char": 0.2997,
+      "acc_token": 0.6732,
+      "epoch": 0.01,
+      "loss_char": 0.7638,
+      "loss_token": 1.648,
+      "lr": "2.00e-07",
+      "norm": 6.0581,
+      "step": 636
+    },
+    {
+      "acc_char": 0.2949,
+      "acc_token": 0.659,
+      "epoch": 0.01,
+      "loss_char": 0.8081,
+      "loss_token": 1.7417,
+      "lr": "2.00e-07",
+      "norm": 5.8723,
+      "step": 637
+    },
+    {
+      "acc_char": 0.296,
+      "acc_token": 0.6734,
+      "epoch": 0.01,
+      "loss_char": 0.7464,
+      "loss_token": 1.6751,
+      "lr": "2.00e-07",
+      "norm": 5.7903,
+      "step": 638
+    },
+    {
+      "acc_char": 0.2949,
+      "acc_token": 0.6709,
+      "epoch": 0.01,
+      "loss_char": 0.7442,
+      "loss_token": 1.6243,
+      "lr": "2.00e-07",
+      "norm": 6.1123,
+      "step": 639
+    },
+    {
+      "acc_char": 0.3115,
+      "acc_token": 0.695,
+      "epoch": 0.01,
+      "loss_char": 0.6904,
+      "loss_token": 1.6397,
+      "lr": "2.00e-07",
+      "norm": 5.9206,
+      "step": 640
+    },
+    {
+      "acc_char": 0.3042,
+      "acc_token": 0.682,
+      "epoch": 0.01,
+      "loss_char": 0.7441,
+      "loss_token": 1.658,
+      "lr": "2.00e-07",
+      "norm": 6.0555,
+      "step": 641
+    },
+    {
+      "acc_char": 0.2977,
+      "acc_token": 0.6757,
+      "epoch": 0.01,
+      "loss_char": 0.7398,
+      "loss_token": 1.6909,
+      "lr": "2.00e-07",
+      "norm": 5.7038,
+      "step": 642
+    },
+    {
+      "acc_char": 0.2856,
+      "acc_token": 0.6136,
+      "epoch": 0.01,
+      "loss_char": 0.9157,
+      "loss_token": 1.7935,
+      "lr": "2.00e-07",
+      "norm": 5.842,
+      "step": 643
+    },
+    {
+      "acc_char": 0.3013,
+      "acc_token": 0.6897,
+      "epoch": 0.01,
+      "loss_char": 0.6881,
+      "loss_token": 1.6058,
+      "lr": "2.00e-07",
+      "norm": 5.7919,
+      "step": 644
+    },
+    {
+      "acc_char": 0.2962,
+      "acc_token": 0.6715,
+      "epoch": 0.01,
+      "loss_char": 0.7371,
+      "loss_token": 1.6975,
+      "lr": "2.00e-07",
+      "norm": 5.8936,
+      "step": 645
+    },
+    {
+      "acc_char": 0.302,
+      "acc_token": 0.6913,
+      "epoch": 0.01,
+      "loss_char": 0.6837,
+      "loss_token": 1.6074,
+      "lr": "2.00e-07",
+      "norm": 5.7401,
+      "step": 646
+    },
+    {
+      "acc_char": 0.3001,
+      "acc_token": 0.6957,
+      "epoch": 0.01,
+      "loss_char": 0.6536,
+      "loss_token": 1.4661,
+      "lr": "2.00e-07",
+      "norm": 5.4626,
+      "step": 647
+    },
+    {
+      "acc_char": 0.2965,
+      "acc_token": 0.6724,
+      "epoch": 0.01,
+      "loss_char": 0.7533,
+      "loss_token": 1.6964,
+      "lr": "2.00e-07",
+      "norm": 5.8447,
+      "step": 648
+    },
+    {
+      "acc_char": 0.3172,
+      "acc_token": 0.7063,
+      "epoch": 0.01,
+      "loss_char": 0.6653,
+      "loss_token": 1.569,
+      "lr": "2.00e-07",
+      "norm": 5.7856,
+      "step": 649
+    },
+    {
+      "acc_char": 0.3075,
+      "acc_token": 0.6877,
+      "epoch": 0.01,
+      "loss_char": 0.7405,
+      "loss_token": 1.6241,
+      "lr": "2.00e-07",
+      "norm": 5.8337,
+      "step": 650
+    },
+    {
+      "acc_char": 0.3189,
+      "acc_token": 0.7198,
+      "epoch": 0.01,
+      "loss_char": 0.6324,
+      "loss_token": 1.5076,
+      "lr": "2.00e-07",
+      "norm": 5.5855,
+      "step": 651
+    },
+    {
+      "acc_char": 0.2925,
+      "acc_token": 0.6688,
+      "epoch": 0.01,
+      "loss_char": 0.7463,
+      "loss_token": 1.6756,
+      "lr": "2.00e-07",
+      "norm": 5.5517,
+      "step": 652
+    },
+    {
+      "acc_char": 0.2955,
+      "acc_token": 0.6403,
+      "epoch": 0.01,
+      "loss_char": 0.8404,
+      "loss_token": 1.7651,
+      "lr": "2.00e-07",
+      "norm": 5.3183,
+      "step": 653
+    },
+    {
+      "acc_char": 0.3084,
+      "acc_token": 0.6892,
+      "epoch": 0.01,
+      "loss_char": 0.7223,
+      "loss_token": 1.6726,
+      "lr": "2.00e-07",
+      "norm": 5.9224,
+      "step": 654
+    },
+    {
+      "acc_char": 0.3103,
+      "acc_token": 0.6965,
+      "epoch": 0.01,
+      "loss_char": 0.687,
+      "loss_token": 1.5716,
+      "lr": "2.00e-07",
+      "norm": 5.8924,
+      "step": 655
+    },
+    {
+      "acc_char": 0.2892,
+      "acc_token": 0.6546,
+      "epoch": 0.01,
+      "loss_char": 0.7878,
+      "loss_token": 1.6663,
+      "lr": "2.00e-07",
+      "norm": 5.677,
+      "step": 656
+    },
+    {
+      "acc_char": 0.3089,
+      "acc_token": 0.7057,
+      "epoch": 0.01,
+      "loss_char": 0.6499,
+      "loss_token": 1.5976,
+      "lr": "2.00e-07",
+      "norm": 5.5667,
+      "step": 657
+    },
+    {
+      "acc_char": 0.3024,
+      "acc_token": 0.6709,
+      "epoch": 0.01,
+      "loss_char": 0.7571,
+      "loss_token": 1.6129,
+      "lr": "2.00e-07",
+      "norm": 5.9277,
+      "step": 658
+    },
+    {
+      "acc_char": 0.3105,
+      "acc_token": 0.7038,
+      "epoch": 0.01,
+      "loss_char": 0.673,
+      "loss_token": 1.6015,
+      "lr": "2.00e-07",
+      "norm": 5.6025,
+      "step": 659
+    },
+    {
+      "acc_char": 0.2913,
+      "acc_token": 0.6678,
+      "epoch": 0.01,
+      "loss_char": 0.7561,
+      "loss_token": 1.6811,
+      "lr": "2.00e-07",
+      "norm": 6.2516,
+      "step": 660
+    },
+    {
+      "acc_char": 0.3001,
+      "acc_token": 0.6993,
+      "epoch": 0.01,
+      "loss_char": 0.6578,
+      "loss_token": 1.5482,
+      "lr": "2.00e-07",
+      "norm": 5.6221,
+      "step": 661
+    },
+    {
+      "acc_char": 0.3069,
+      "acc_token": 0.6845,
+      "epoch": 0.01,
+      "loss_char": 0.7393,
+      "loss_token": 1.6031,
+      "lr": "2.00e-07",
+      "norm": 5.45,
+      "step": 662
+    },
+    {
+      "acc_char": 0.3284,
+      "acc_token": 0.7275,
+      "epoch": 0.01,
+      "loss_char": 0.6488,
+      "loss_token": 1.5209,
+      "lr": "2.00e-07",
+      "norm": 5.5336,
+      "step": 663
+    },
+    {
+      "acc_char": 0.3029,
+      "acc_token": 0.6869,
+      "epoch": 0.01,
+      "loss_char": 0.7221,
+      "loss_token": 1.6319,
+      "lr": "2.00e-07",
+      "norm": 5.9041,
+      "step": 664
+    },
+    {
+      "acc_char": 0.2964,
+      "acc_token": 0.671,
+      "epoch": 0.01,
+      "loss_char": 0.7496,
+      "loss_token": 1.5912,
+      "lr": "2.00e-07",
+      "norm": 5.6728,
+      "step": 665
+    },
+    {
+      "acc_char": 0.3056,
+      "acc_token": 0.6947,
+      "epoch": 0.01,
+      "loss_char": 0.6923,
+      "loss_token": 1.5917,
+      "lr": "2.00e-07",
+      "norm": 5.6104,
+      "step": 666
+    },
+    {
+      "acc_char": 0.301,
+      "acc_token": 0.6766,
+      "epoch": 0.01,
+      "loss_char": 0.7448,
+      "loss_token": 1.5958,
+      "lr": "2.00e-07",
+      "norm": 5.7438,
+      "step": 667
+    },
+    {
+      "acc_char": 0.3047,
+      "acc_token": 0.684,
+      "epoch": 0.01,
+      "loss_char": 0.7429,
+      "loss_token": 1.715,
+      "lr": "2.00e-07",
+      "norm": 5.9348,
+      "step": 668
+    },
+    {
+      "acc_char": 0.2961,
+      "acc_token": 0.6706,
+      "epoch": 0.01,
+      "loss_char": 0.7799,
+      "loss_token": 1.6612,
+      "lr": "2.00e-07",
+      "norm": 5.7273,
+      "step": 669
+    },
+    {
+      "acc_char": 0.3023,
+      "acc_token": 0.6895,
+      "epoch": 0.01,
+      "loss_char": 0.6908,
+      "loss_token": 1.5717,
+      "lr": "2.00e-07",
+      "norm": 5.6183,
+      "step": 670
+    },
+    {
+      "acc_char": 0.3234,
+      "acc_token": 0.7105,
+      "epoch": 0.01,
+      "loss_char": 0.6642,
+      "loss_token": 1.5869,
+      "lr": "2.00e-07",
+      "norm": 5.6473,
+      "step": 671
+    },
+    {
+      "acc_char": 0.3022,
+      "acc_token": 0.694,
+      "epoch": 0.01,
+      "loss_char": 0.6711,
+      "loss_token": 1.5884,
+      "lr": "2.00e-07",
+      "norm": 5.3875,
+      "step": 672
+    },
+    {
+      "acc_char": 0.3039,
+      "acc_token": 0.6902,
+      "epoch": 0.01,
+      "loss_char": 0.7018,
+      "loss_token": 1.5795,
+      "lr": "2.00e-07",
+      "norm": 5.7039,
+      "step": 673
+    },
+    {
+      "acc_char": 0.3078,
+      "acc_token": 0.6989,
+      "epoch": 0.01,
+      "loss_char": 0.6835,
+      "loss_token": 1.5994,
+      "lr": "2.00e-07",
+      "norm": 5.6163,
+      "step": 674
+    },
+    {
+      "acc_char": 0.3056,
+      "acc_token": 0.6877,
+      "epoch": 0.01,
+      "loss_char": 0.7003,
+      "loss_token": 1.581,
+      "lr": "2.00e-07",
+      "norm": 5.638,
+      "step": 675
+    },
+    {
+      "acc_char": 0.2879,
+      "acc_token": 0.661,
+      "epoch": 0.01,
+      "loss_char": 0.7901,
+      "loss_token": 1.7105,
+      "lr": "2.00e-07",
+      "norm": 5.9664,
+      "step": 676
+    },
+    {
+      "acc_char": 0.2932,
+      "acc_token": 0.6609,
+      "epoch": 0.01,
+      "loss_char": 0.8064,
+      "loss_token": 1.7189,
+      "lr": "2.00e-07",
+      "norm": 5.7921,
+      "step": 677
+    },
+    {
+      "acc_char": 0.3141,
+      "acc_token": 0.6946,
+      "epoch": 0.01,
+      "loss_char": 0.6943,
+      "loss_token": 1.5659,
+      "lr": "2.00e-07",
+      "norm": 5.6831,
+      "step": 678
+    },
+    {
+      "acc_char": 0.3159,
+      "acc_token": 0.7043,
+      "epoch": 0.01,
+      "loss_char": 0.6686,
+      "loss_token": 1.5816,
+      "lr": "2.00e-07",
+      "norm": 5.5089,
+      "step": 679
+    },
+    {
+      "acc_char": 0.3145,
+      "acc_token": 0.7022,
+      "epoch": 0.01,
+      "loss_char": 0.667,
+      "loss_token": 1.533,
+      "lr": "2.00e-07",
+      "norm": 5.797,
+      "step": 680
+    },
+    {
+      "acc_char": 0.3081,
+      "acc_token": 0.6984,
+      "epoch": 0.01,
+      "loss_char": 0.6948,
+      "loss_token": 1.4594,
+      "lr": "2.00e-07",
+      "norm": 5.7751,
+      "step": 681
+    },
+    {
+      "acc_char": 0.2999,
+      "acc_token": 0.6793,
+      "epoch": 0.01,
+      "loss_char": 0.7456,
+      "loss_token": 1.712,
+      "lr": "2.00e-07",
+      "norm": 6.2523,
+      "step": 682
+    },
+    {
+      "acc_char": 0.3072,
+      "acc_token": 0.6903,
+      "epoch": 0.01,
+      "loss_char": 0.7033,
+      "loss_token": 1.6286,
+      "lr": "2.00e-07",
+      "norm": 6.0363,
+      "step": 683
+    },
+    {
+      "acc_char": 0.319,
+      "acc_token": 0.71,
+      "epoch": 0.01,
+      "loss_char": 0.6502,
+      "loss_token": 1.509,
+      "lr": "2.00e-07",
+      "norm": 5.4379,
+      "step": 684
+    },
+    {
+      "acc_char": 0.2994,
+      "acc_token": 0.6809,
+      "epoch": 0.01,
+      "loss_char": 0.7212,
+      "loss_token": 1.6315,
+      "lr": "2.00e-07",
+      "norm": 5.6878,
+      "step": 685
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.692,
+      "epoch": 0.01,
+      "loss_char": 0.6758,
+      "loss_token": 1.5584,
+      "lr": "2.00e-07",
+      "norm": 5.3446,
+      "step": 686
+    },
+    {
+      "acc_char": 0.2998,
+      "acc_token": 0.6789,
+      "epoch": 0.01,
+      "loss_char": 0.7216,
+      "loss_token": 1.6356,
+      "lr": "2.00e-07",
+      "norm": 5.8065,
+      "step": 687
+    },
+    {
+      "acc_char": 0.3075,
+      "acc_token": 0.6532,
+      "epoch": 0.01,
+      "loss_char": 0.8291,
+      "loss_token": 1.6488,
+      "lr": "2.00e-07",
+      "norm": 5.4422,
+      "step": 688
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6787,
+      "epoch": 0.01,
+      "loss_char": 0.7095,
+      "loss_token": 1.5677,
+      "lr": "2.00e-07",
+      "norm": 5.7396,
+      "step": 689
+    },
+    {
+      "acc_char": 0.3176,
+      "acc_token": 0.7026,
+      "epoch": 0.01,
+      "loss_char": 0.6848,
+      "loss_token": 1.5056,
+      "lr": "2.00e-07",
+      "norm": 5.636,
+      "step": 690
+    },
+    {
+      "acc_char": 0.3101,
+      "acc_token": 0.6997,
+      "epoch": 0.01,
+      "loss_char": 0.6909,
+      "loss_token": 1.6358,
+      "lr": "2.00e-07",
+      "norm": 5.7493,
+      "step": 691
+    },
+    {
+      "acc_char": 0.3137,
+      "acc_token": 0.7156,
+      "epoch": 0.01,
+      "loss_char": 0.6279,
+      "loss_token": 1.537,
+      "lr": "2.00e-07",
+      "norm": 5.8835,
+      "step": 692
+    },
+    {
+      "acc_char": 0.2932,
+      "acc_token": 0.6654,
+      "epoch": 0.01,
+      "loss_char": 0.7754,
+      "loss_token": 1.6879,
+      "lr": "2.00e-07",
+      "norm": 5.8848,
+      "step": 693
+    },
+    {
+      "acc_char": 0.3011,
+      "acc_token": 0.6875,
+      "epoch": 0.01,
+      "loss_char": 0.7015,
+      "loss_token": 1.6103,
+      "lr": "2.00e-07",
+      "norm": 5.6533,
+      "step": 694
+    },
+    {
+      "acc_char": 0.2931,
+      "acc_token": 0.6376,
+      "epoch": 0.01,
+      "loss_char": 0.8559,
+      "loss_token": 1.7466,
+      "lr": "2.00e-07",
+      "norm": 5.4473,
+      "step": 695
+    },
+    {
+      "acc_char": 0.2941,
+      "acc_token": 0.6733,
+      "epoch": 0.01,
+      "loss_char": 0.7255,
+      "loss_token": 1.635,
+      "lr": "2.00e-07",
+      "norm": 5.7071,
+      "step": 696
+    },
+    {
+      "acc_char": 0.3102,
+      "acc_token": 0.6983,
+      "epoch": 0.01,
+      "loss_char": 0.6859,
+      "loss_token": 1.5594,
+      "lr": "2.00e-07",
+      "norm": 5.4668,
+      "step": 697
+    },
+    {
+      "acc_char": 0.3086,
+      "acc_token": 0.6872,
+      "epoch": 0.01,
+      "loss_char": 0.7307,
+      "loss_token": 1.6843,
+      "lr": "2.00e-07",
+      "norm": 5.5325,
+      "step": 698
+    },
+    {
+      "acc_char": 0.2999,
+      "acc_token": 0.6743,
+      "epoch": 0.01,
+      "loss_char": 0.7607,
+      "loss_token": 1.6332,
+      "lr": "2.00e-07",
+      "norm": 5.5354,
+      "step": 699
+    },
+    {
+      "acc_char": 0.2952,
+      "acc_token": 0.6822,
+      "epoch": 0.01,
+      "loss_char": 0.7013,
+      "loss_token": 1.6426,
+      "lr": "2.00e-07",
+      "norm": 5.5413,
+      "step": 700
+    },
+    {
+      "acc_char": 0.304,
+      "acc_token": 0.6866,
+      "epoch": 0.01,
+      "loss_char": 0.714,
+      "loss_token": 1.6075,
+      "lr": "2.00e-07",
+      "norm": 5.6022,
+      "step": 701
+    },
+    {
+      "acc_char": 0.2911,
+      "acc_token": 0.6571,
+      "epoch": 0.01,
+      "loss_char": 0.8123,
+      "loss_token": 1.6884,
+      "lr": "2.00e-07",
+      "norm": 5.6951,
+      "step": 702
+    },
+    {
+      "acc_char": 0.2935,
+      "acc_token": 0.6654,
+      "epoch": 0.01,
+      "loss_char": 0.7613,
+      "loss_token": 1.6945,
+      "lr": "2.00e-07",
+      "norm": 5.931,
+      "step": 703
+    },
+    {
+      "acc_char": 0.2941,
+      "acc_token": 0.6635,
+      "epoch": 0.01,
+      "loss_char": 0.765,
+      "loss_token": 1.7128,
+      "lr": "2.00e-07",
+      "norm": 6.0656,
+      "step": 704
+    },
+    {
+      "acc_char": 0.3081,
+      "acc_token": 0.6853,
+      "epoch": 0.01,
+      "loss_char": 0.7405,
+      "loss_token": 1.7216,
+      "lr": "2.00e-07",
+      "norm": 5.7774,
+      "step": 705
+    },
+    {
+      "acc_char": 0.3127,
+      "acc_token": 0.6943,
+      "epoch": 0.01,
+      "loss_char": 0.7052,
+      "loss_token": 1.6332,
+      "lr": "2.00e-07",
+      "norm": 5.8129,
+      "step": 706
+    },
+    {
+      "acc_char": 0.304,
+      "acc_token": 0.6762,
+      "epoch": 0.01,
+      "loss_char": 0.7689,
+      "loss_token": 1.6594,
+      "lr": "2.00e-07",
+      "norm": 5.8423,
+      "step": 707
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6949,
+      "epoch": 0.01,
+      "loss_char": 0.6868,
+      "loss_token": 1.6849,
+      "lr": "2.00e-07",
+      "norm": 5.7876,
+      "step": 708
+    },
+    {
+      "acc_char": 0.3085,
+      "acc_token": 0.6924,
+      "epoch": 0.01,
+      "loss_char": 0.6746,
+      "loss_token": 1.5821,
+      "lr": "2.00e-07",
+      "norm": 5.5369,
+      "step": 709
+    },
+    {
+      "acc_char": 0.294,
+      "acc_token": 0.6777,
+      "epoch": 0.01,
+      "loss_char": 0.7509,
+      "loss_token": 1.7132,
+      "lr": "2.00e-07",
+      "norm": 5.7105,
+      "step": 710
+    },
+    {
+      "acc_char": 0.3187,
+      "acc_token": 0.7084,
+      "epoch": 0.01,
+      "loss_char": 0.6792,
+      "loss_token": 1.5886,
+      "lr": "2.00e-07",
+      "norm": 5.4638,
+      "step": 711
+    },
+    {
+      "acc_char": 0.3075,
+      "acc_token": 0.693,
+      "epoch": 0.01,
+      "loss_char": 0.7031,
+      "loss_token": 1.6067,
+      "lr": "2.00e-07",
+      "norm": 5.5524,
+      "step": 712
+    },
+    {
+      "acc_char": 0.2849,
+      "acc_token": 0.6219,
+      "epoch": 0.01,
+      "loss_char": 0.927,
+      "loss_token": 1.8086,
+      "lr": "2.00e-07",
+      "norm": 5.673,
+      "step": 713
+    },
+    {
+      "acc_char": 0.2773,
+      "acc_token": 0.6437,
+      "epoch": 0.01,
+      "loss_char": 0.8132,
+      "loss_token": 1.7713,
+      "lr": "2.00e-07",
+      "norm": 5.9949,
+      "step": 714
+    },
+    {
+      "acc_char": 0.3059,
+      "acc_token": 0.6888,
+      "epoch": 0.01,
+      "loss_char": 0.7081,
+      "loss_token": 1.5482,
+      "lr": "2.00e-07",
+      "norm": 5.521,
+      "step": 715
+    },
+    {
+      "acc_char": 0.308,
+      "acc_token": 0.693,
+      "epoch": 0.01,
+      "loss_char": 0.7066,
+      "loss_token": 1.6414,
+      "lr": "2.00e-07",
+      "norm": 5.5016,
+      "step": 716
+    },
+    {
+      "acc_char": 0.3138,
+      "acc_token": 0.6991,
+      "epoch": 0.01,
+      "loss_char": 0.6901,
+      "loss_token": 1.5942,
+      "lr": "2.00e-07",
+      "norm": 5.4398,
+      "step": 717
+    },
+    {
+      "acc_char": 0.3008,
+      "acc_token": 0.6882,
+      "epoch": 0.01,
+      "loss_char": 0.6941,
+      "loss_token": 1.6136,
+      "lr": "2.00e-07",
+      "norm": 5.9947,
+      "step": 718
+    },
+    {
+      "acc_char": 0.2996,
+      "acc_token": 0.6727,
+      "epoch": 0.01,
+      "loss_char": 0.7845,
+      "loss_token": 1.7138,
+      "lr": "2.00e-07",
+      "norm": 5.8203,
+      "step": 719
+    },
+    {
+      "acc_char": 0.3033,
+      "acc_token": 0.6859,
+      "epoch": 0.01,
+      "loss_char": 0.7195,
+      "loss_token": 1.5991,
+      "lr": "2.00e-07",
+      "norm": 5.6985,
+      "step": 720
+    },
+    {
+      "acc_char": 0.2933,
+      "acc_token": 0.6239,
+      "epoch": 0.01,
+      "loss_char": 0.9131,
+      "loss_token": 1.6806,
+      "lr": "2.00e-07",
+      "norm": 5.6157,
+      "step": 721
+    },
+    {
+      "acc_char": 0.3045,
+      "acc_token": 0.7004,
+      "epoch": 0.01,
+      "loss_char": 0.6579,
+      "loss_token": 1.5557,
+      "lr": "2.00e-07",
+      "norm": 5.5476,
+      "step": 722
+    },
+    {
+      "acc_char": 0.325,
+      "acc_token": 0.728,
+      "epoch": 0.01,
+      "loss_char": 0.6151,
+      "loss_token": 1.445,
+      "lr": "2.00e-07",
+      "norm": 5.3312,
+      "step": 723
+    },
+    {
+      "acc_char": 0.3197,
+      "acc_token": 0.7111,
+      "epoch": 0.01,
+      "loss_char": 0.6784,
+      "loss_token": 1.46,
+      "lr": "2.00e-07",
+      "norm": 5.5799,
+      "step": 724
+    },
+    {
+      "acc_char": 0.2991,
+      "acc_token": 0.6815,
+      "epoch": 0.01,
+      "loss_char": 0.7086,
+      "loss_token": 1.6515,
+      "lr": "2.00e-07",
+      "norm": 5.9712,
+      "step": 725
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6724,
+      "epoch": 0.01,
+      "loss_char": 0.763,
+      "loss_token": 1.7044,
+      "lr": "2.00e-07",
+      "norm": 5.7853,
+      "step": 726
+    },
+    {
+      "acc_char": 0.3038,
+      "acc_token": 0.6934,
+      "epoch": 0.01,
+      "loss_char": 0.6967,
+      "loss_token": 1.572,
+      "lr": "2.00e-07",
+      "norm": 5.8128,
+      "step": 727
+    },
+    {
+      "acc_char": 0.3033,
+      "acc_token": 0.6949,
+      "epoch": 0.01,
+      "loss_char": 0.6848,
+      "loss_token": 1.522,
+      "lr": "2.00e-07",
+      "norm": 5.7481,
+      "step": 728
+    },
+    {
+      "acc_char": 0.3226,
+      "acc_token": 0.7245,
+      "epoch": 0.01,
+      "loss_char": 0.6269,
+      "loss_token": 1.5333,
+      "lr": "2.00e-07",
+      "norm": 5.6008,
+      "step": 729
+    },
+    {
+      "acc_char": 0.3065,
+      "acc_token": 0.6905,
+      "epoch": 0.01,
+      "loss_char": 0.7018,
+      "loss_token": 1.6624,
+      "lr": "2.00e-07",
+      "norm": 6.1109,
+      "step": 730
+    },
+    {
+      "acc_char": 0.3067,
+      "acc_token": 0.6836,
+      "epoch": 0.01,
+      "loss_char": 0.7501,
+      "loss_token": 1.6427,
+      "lr": "2.00e-07",
+      "norm": 6.2004,
+      "step": 731
+    },
+    {
+      "acc_char": 0.2936,
+      "acc_token": 0.6625,
+      "epoch": 0.01,
+      "loss_char": 0.7793,
+      "loss_token": 1.6704,
+      "lr": "2.00e-07",
+      "norm": 5.7747,
+      "step": 732
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.6706,
+      "epoch": 0.01,
+      "loss_char": 0.7602,
+      "loss_token": 1.6386,
+      "lr": "2.00e-07",
+      "norm": 5.8706,
+      "step": 733
+    },
+    {
+      "acc_char": 0.3177,
+      "acc_token": 0.7132,
+      "epoch": 0.01,
+      "loss_char": 0.6471,
+      "loss_token": 1.5085,
+      "lr": "2.00e-07",
+      "norm": 5.4413,
+      "step": 734
+    },
+    {
+      "acc_char": 0.3092,
+      "acc_token": 0.7031,
+      "epoch": 0.01,
+      "loss_char": 0.6671,
+      "loss_token": 1.6531,
+      "lr": "2.00e-07",
+      "norm": 6.0109,
+      "step": 735
+    },
+    {
+      "acc_char": 0.3015,
+      "acc_token": 0.6878,
+      "epoch": 0.01,
+      "loss_char": 0.6874,
+      "loss_token": 1.6372,
+      "lr": "2.00e-07",
+      "norm": 5.596,
+      "step": 736
+    },
+    {
+      "acc_char": 0.3057,
+      "acc_token": 0.6861,
+      "epoch": 0.01,
+      "loss_char": 0.7389,
+      "loss_token": 1.6141,
+      "lr": "2.00e-07",
+      "norm": 5.8621,
+      "step": 737
+    },
+    {
+      "acc_char": 0.3023,
+      "acc_token": 0.6911,
+      "epoch": 0.01,
+      "loss_char": 0.7052,
+      "loss_token": 1.5856,
+      "lr": "2.00e-07",
+      "norm": 5.5047,
+      "step": 738
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6785,
+      "epoch": 0.01,
+      "loss_char": 0.752,
+      "loss_token": 1.6367,
+      "lr": "2.00e-07",
+      "norm": 5.5477,
+      "step": 739
+    },
+    {
+      "acc_char": 0.3084,
+      "acc_token": 0.6987,
+      "epoch": 0.01,
+      "loss_char": 0.6865,
+      "loss_token": 1.5974,
+      "lr": "2.00e-07",
+      "norm": 6.0687,
+      "step": 740
+    },
+    {
+      "acc_char": 0.3089,
+      "acc_token": 0.6936,
+      "epoch": 0.01,
+      "loss_char": 0.7127,
+      "loss_token": 1.6018,
+      "lr": "2.00e-07",
+      "norm": 5.746,
+      "step": 741
+    },
+    {
+      "acc_char": 0.3024,
+      "acc_token": 0.6732,
+      "epoch": 0.01,
+      "loss_char": 0.7528,
+      "loss_token": 1.677,
+      "lr": "2.00e-07",
+      "norm": 5.7916,
+      "step": 742
+    },
+    {
+      "acc_char": 0.2957,
+      "acc_token": 0.6589,
+      "epoch": 0.01,
+      "loss_char": 0.8215,
+      "loss_token": 1.7048,
+      "lr": "2.00e-07",
+      "norm": 5.8918,
+      "step": 743
+    },
+    {
+      "acc_char": 0.3012,
+      "acc_token": 0.6779,
+      "epoch": 0.01,
+      "loss_char": 0.7707,
+      "loss_token": 1.6728,
+      "lr": "2.00e-07",
+      "norm": 5.7565,
+      "step": 744
+    },
+    {
+      "acc_char": 0.3041,
+      "acc_token": 0.6828,
+      "epoch": 0.01,
+      "loss_char": 0.7289,
+      "loss_token": 1.5819,
+      "lr": "2.00e-07",
+      "norm": 5.6077,
+      "step": 745
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6637,
+      "epoch": 0.01,
+      "loss_char": 0.7989,
+      "loss_token": 1.6926,
+      "lr": "2.00e-07",
+      "norm": 5.8018,
+      "step": 746
+    },
+    {
+      "acc_char": 0.3197,
+      "acc_token": 0.7233,
+      "epoch": 0.01,
+      "loss_char": 0.6119,
+      "loss_token": 1.6102,
+      "lr": "2.00e-07",
+      "norm": 5.7715,
+      "step": 747
+    },
+    {
+      "acc_char": 0.3048,
+      "acc_token": 0.6811,
+      "epoch": 0.01,
+      "loss_char": 0.7491,
+      "loss_token": 1.6751,
+      "lr": "2.00e-07",
+      "norm": 5.7565,
+      "step": 748
+    },
+    {
+      "acc_char": 0.3144,
+      "acc_token": 0.706,
+      "epoch": 0.01,
+      "loss_char": 0.6729,
+      "loss_token": 1.5566,
+      "lr": "2.00e-07",
+      "norm": 5.6953,
+      "step": 749
+    },
+    {
+      "acc_char": 0.3123,
+      "acc_token": 0.6989,
+      "epoch": 0.01,
+      "loss_char": 0.6951,
+      "loss_token": 1.5327,
+      "lr": "2.00e-07",
+      "norm": 5.5628,
+      "step": 750
+    },
+    {
+      "acc_char": 0.3054,
+      "acc_token": 0.7036,
+      "epoch": 0.01,
+      "loss_char": 0.6472,
+      "loss_token": 1.5634,
+      "lr": "2.00e-07",
+      "norm": 5.4744,
+      "step": 751
+    },
+    {
+      "acc_char": 0.2984,
+      "acc_token": 0.675,
+      "epoch": 0.01,
+      "loss_char": 0.7445,
+      "loss_token": 1.643,
+      "lr": "2.00e-07",
+      "norm": 5.8631,
+      "step": 752
+    },
+    {
+      "acc_char": 0.3052,
+      "acc_token": 0.6844,
+      "epoch": 0.01,
+      "loss_char": 0.7517,
+      "loss_token": 1.6013,
+      "lr": "2.00e-07",
+      "norm": 6.0664,
+      "step": 753
+    },
+    {
+      "acc_char": 0.3036,
+      "acc_token": 0.6907,
+      "epoch": 0.01,
+      "loss_char": 0.6976,
+      "loss_token": 1.6298,
+      "lr": "2.00e-07",
+      "norm": 5.4927,
+      "step": 754
+    },
+    {
+      "acc_char": 0.2969,
+      "acc_token": 0.6864,
+      "epoch": 0.01,
+      "loss_char": 0.7042,
+      "loss_token": 1.5909,
+      "lr": "2.00e-07",
+      "norm": 5.7039,
+      "step": 755
+    },
+    {
+      "acc_char": 0.2922,
+      "acc_token": 0.6709,
+      "epoch": 0.01,
+      "loss_char": 0.7614,
+      "loss_token": 1.6731,
+      "lr": "2.00e-07",
+      "norm": 5.6499,
+      "step": 756
+    },
+    {
+      "acc_char": 0.2987,
+      "acc_token": 0.6751,
+      "epoch": 0.01,
+      "loss_char": 0.7543,
+      "loss_token": 1.7016,
+      "lr": "2.00e-07",
+      "norm": 5.9883,
+      "step": 757
+    },
+    {
+      "acc_char": 0.3007,
+      "acc_token": 0.6894,
+      "epoch": 0.01,
+      "loss_char": 0.7218,
+      "loss_token": 1.6053,
+      "lr": "2.00e-07",
+      "norm": 5.7038,
+      "step": 758
+    },
+    {
+      "acc_char": 0.3117,
+      "acc_token": 0.6955,
+      "epoch": 0.01,
+      "loss_char": 0.6966,
+      "loss_token": 1.517,
+      "lr": "2.00e-07",
+      "norm": 5.3883,
+      "step": 759
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.693,
+      "epoch": 0.01,
+      "loss_char": 0.7054,
+      "loss_token": 1.5747,
+      "lr": "2.00e-07",
+      "norm": 5.5406,
+      "step": 760
+    },
+    {
+      "acc_char": 0.299,
+      "acc_token": 0.6792,
+      "epoch": 0.01,
+      "loss_char": 0.731,
+      "loss_token": 1.6095,
+      "lr": "2.00e-07",
+      "norm": 5.4993,
+      "step": 761
+    },
+    {
+      "acc_char": 0.3368,
+      "acc_token": 0.7164,
+      "epoch": 0.01,
+      "loss_char": 0.6217,
+      "loss_token": 1.5015,
+      "lr": "2.00e-07",
+      "norm": 5.1857,
+      "step": 762
+    },
+    {
+      "acc_char": 0.32,
+      "acc_token": 0.7251,
+      "epoch": 0.01,
+      "loss_char": 0.6227,
+      "loss_token": 1.5617,
+      "lr": "2.00e-07",
+      "norm": 5.6146,
+      "step": 763
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6952,
+      "epoch": 0.01,
+      "loss_char": 0.6669,
+      "loss_token": 1.5794,
+      "lr": "2.00e-07",
+      "norm": 5.564,
+      "step": 764
+    },
+    {
+      "acc_char": 0.2932,
+      "acc_token": 0.6711,
+      "epoch": 0.01,
+      "loss_char": 0.7361,
+      "loss_token": 1.6644,
+      "lr": "2.00e-07",
+      "norm": 5.783,
+      "step": 765
+    },
+    {
+      "acc_char": 0.3053,
+      "acc_token": 0.684,
+      "epoch": 0.01,
+      "loss_char": 0.7416,
+      "loss_token": 1.6462,
+      "lr": "2.00e-07",
+      "norm": 5.962,
+      "step": 766
+    },
+    {
+      "acc_char": 0.3112,
+      "acc_token": 0.6991,
+      "epoch": 0.01,
+      "loss_char": 0.6947,
+      "loss_token": 1.5991,
+      "lr": "2.00e-07",
+      "norm": 5.9584,
+      "step": 767
+    },
+    {
+      "acc_char": 0.3141,
+      "acc_token": 0.7128,
+      "epoch": 0.01,
+      "loss_char": 0.6485,
+      "loss_token": 1.5699,
+      "lr": "2.00e-07",
+      "norm": 5.4834,
+      "step": 768
+    },
+    {
+      "acc_char": 0.3241,
+      "acc_token": 0.7386,
+      "epoch": 0.01,
+      "loss_char": 0.5885,
+      "loss_token": 1.5736,
+      "lr": "2.00e-07",
+      "norm": 5.707,
+      "step": 769
+    },
+    {
+      "acc_char": 0.3106,
+      "acc_token": 0.6966,
+      "epoch": 0.01,
+      "loss_char": 0.6809,
+      "loss_token": 1.6651,
+      "lr": "2.00e-07",
+      "norm": 6.4674,
+      "step": 770
+    },
+    {
+      "acc_char": 0.3109,
+      "acc_token": 0.7015,
+      "epoch": 0.01,
+      "loss_char": 0.6787,
+      "loss_token": 1.5798,
+      "lr": "2.00e-07",
+      "norm": 5.4152,
+      "step": 771
+    },
+    {
+      "acc_char": 0.2791,
+      "acc_token": 0.6064,
+      "epoch": 0.01,
+      "loss_char": 0.9144,
+      "loss_token": 1.7602,
+      "lr": "2.00e-07",
+      "norm": 5.3641,
+      "step": 772
+    },
+    {
+      "acc_char": 0.3082,
+      "acc_token": 0.6722,
+      "epoch": 0.01,
+      "loss_char": 0.7471,
+      "loss_token": 1.6349,
+      "lr": "2.00e-07",
+      "norm": 5.1329,
+      "step": 773
+    },
+    {
+      "acc_char": 0.2971,
+      "acc_token": 0.6746,
+      "epoch": 0.01,
+      "loss_char": 0.7329,
+      "loss_token": 1.7143,
+      "lr": "2.00e-07",
+      "norm": 5.9449,
+      "step": 774
+    },
+    {
+      "acc_char": 0.2802,
+      "acc_token": 0.6183,
+      "epoch": 0.01,
+      "loss_char": 0.8729,
+      "loss_token": 1.7259,
+      "lr": "2.00e-07",
+      "norm": 5.5167,
+      "step": 775
+    },
+    {
+      "acc_char": 0.3069,
+      "acc_token": 0.6942,
+      "epoch": 0.01,
+      "loss_char": 0.6939,
+      "loss_token": 1.653,
+      "lr": "2.00e-07",
+      "norm": 5.749,
+      "step": 776
+    },
+    {
+      "acc_char": 0.2951,
+      "acc_token": 0.6668,
+      "epoch": 0.01,
+      "loss_char": 0.7659,
+      "loss_token": 1.6842,
+      "lr": "2.00e-07",
+      "norm": 5.5857,
+      "step": 777
+    },
+    {
+      "acc_char": 0.3025,
+      "acc_token": 0.6804,
+      "epoch": 0.01,
+      "loss_char": 0.7307,
+      "loss_token": 1.6822,
+      "lr": "2.00e-07",
+      "norm": 5.941,
+      "step": 778
+    },
+    {
+      "acc_char": 0.3009,
+      "acc_token": 0.6777,
+      "epoch": 0.01,
+      "loss_char": 0.7417,
+      "loss_token": 1.6323,
+      "lr": "2.00e-07",
+      "norm": 5.6713,
+      "step": 779
+    },
+    {
+      "acc_char": 0.2949,
+      "acc_token": 0.6748,
+      "epoch": 0.01,
+      "loss_char": 0.7385,
+      "loss_token": 1.7698,
+      "lr": "2.00e-07",
+      "norm": 5.7486,
+      "step": 780
+    },
+    {
+      "acc_char": 0.2982,
+      "acc_token": 0.6756,
+      "epoch": 0.01,
+      "loss_char": 0.7716,
+      "loss_token": 1.8033,
+      "lr": "2.00e-07",
+      "norm": 5.9043,
+      "step": 781
+    },
+    {
+      "acc_char": 0.3044,
+      "acc_token": 0.7008,
+      "epoch": 0.01,
+      "loss_char": 0.6561,
+      "loss_token": 1.5758,
+      "lr": "2.00e-07",
+      "norm": 5.4415,
+      "step": 782
+    },
+    {
+      "acc_char": 0.2863,
+      "acc_token": 0.6593,
+      "epoch": 0.01,
+      "loss_char": 0.7863,
+      "loss_token": 1.7233,
+      "lr": "2.00e-07",
+      "norm": 5.7104,
+      "step": 783
+    },
+    {
+      "acc_char": 0.3049,
+      "acc_token": 0.6844,
+      "epoch": 0.01,
+      "loss_char": 0.7141,
+      "loss_token": 1.6678,
+      "lr": "2.00e-07",
+      "norm": 6.1641,
+      "step": 784
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.681,
+      "epoch": 0.01,
+      "loss_char": 0.7087,
+      "loss_token": 1.6866,
+      "lr": "2.00e-07",
+      "norm": 5.9054,
+      "step": 785
+    },
+    {
+      "acc_char": 0.3028,
+      "acc_token": 0.686,
+      "epoch": 0.01,
+      "loss_char": 0.7252,
+      "loss_token": 1.6103,
+      "lr": "2.00e-07",
+      "norm": 5.7014,
+      "step": 786
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.6935,
+      "epoch": 0.01,
+      "loss_char": 0.6981,
+      "loss_token": 1.6534,
+      "lr": "2.00e-07",
+      "norm": 5.6859,
+      "step": 787
+    },
+    {
+      "acc_char": 0.2767,
+      "acc_token": 0.6098,
+      "epoch": 0.01,
+      "loss_char": 0.8904,
+      "loss_token": 1.7864,
+      "lr": "2.00e-07",
+      "norm": 5.5673,
+      "step": 788
+    },
+    {
+      "acc_char": 0.3095,
+      "acc_token": 0.6966,
+      "epoch": 0.01,
+      "loss_char": 0.6927,
+      "loss_token": 1.5377,
+      "lr": "2.00e-07",
+      "norm": 5.614,
+      "step": 789
+    },
+    {
+      "acc_char": 0.3046,
+      "acc_token": 0.6851,
+      "epoch": 0.01,
+      "loss_char": 0.7454,
+      "loss_token": 1.601,
+      "lr": "2.00e-07",
+      "norm": 5.6357,
+      "step": 790
+    },
+    {
+      "acc_char": 0.3011,
+      "acc_token": 0.6869,
+      "epoch": 0.01,
+      "loss_char": 0.7022,
+      "loss_token": 1.6659,
+      "lr": "2.00e-07",
+      "norm": 5.6749,
+      "step": 791
+    },
+    {
+      "acc_char": 0.2996,
+      "acc_token": 0.6817,
+      "epoch": 0.01,
+      "loss_char": 0.7403,
+      "loss_token": 1.6017,
+      "lr": "2.00e-07",
+      "norm": 5.5231,
+      "step": 792
+    },
+    {
+      "acc_char": 0.3026,
+      "acc_token": 0.6878,
+      "epoch": 0.01,
+      "loss_char": 0.701,
+      "loss_token": 1.5683,
+      "lr": "2.00e-07",
+      "norm": 6.018,
+      "step": 793
+    },
+    {
+      "acc_char": 0.3143,
+      "acc_token": 0.6957,
+      "epoch": 0.01,
+      "loss_char": 0.6879,
+      "loss_token": 1.6818,
+      "lr": "2.00e-07",
+      "norm": 6.0463,
+      "step": 794
+    },
+    {
+      "acc_char": 0.3123,
+      "acc_token": 0.7032,
+      "epoch": 0.01,
+      "loss_char": 0.6796,
+      "loss_token": 1.559,
+      "lr": "2.00e-07",
+      "norm": 5.3763,
+      "step": 795
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6721,
+      "epoch": 0.01,
+      "loss_char": 0.7448,
+      "loss_token": 1.6649,
+      "lr": "2.00e-07",
+      "norm": 5.6504,
+      "step": 796
+    },
+    {
+      "acc_char": 0.3108,
+      "acc_token": 0.6951,
+      "epoch": 0.01,
+      "loss_char": 0.71,
+      "loss_token": 1.5358,
+      "lr": "2.00e-07",
+      "norm": 5.5775,
+      "step": 797
+    },
+    {
+      "acc_char": 0.2889,
+      "acc_token": 0.6635,
+      "epoch": 0.01,
+      "loss_char": 0.7764,
+      "loss_token": 1.6694,
+      "lr": "2.00e-07",
+      "norm": 5.8663,
+      "step": 798
+    },
+    {
+      "acc_char": 0.3095,
+      "acc_token": 0.7013,
+      "epoch": 0.01,
+      "loss_char": 0.6619,
+      "loss_token": 1.5684,
+      "lr": "2.00e-07",
+      "norm": 5.5381,
+      "step": 799
+    },
+    {
+      "acc_char": 0.3073,
+      "acc_token": 0.6848,
+      "epoch": 0.01,
+      "loss_char": 0.7427,
+      "loss_token": 1.6565,
+      "lr": "2.00e-07",
+      "norm": 5.7262,
+      "step": 800
+    },
+    {
+      "acc_char": 0.2876,
+      "acc_token": 0.6455,
+      "epoch": 0.01,
+      "loss_char": 0.8496,
+      "loss_token": 1.76,
+      "lr": "2.00e-07",
+      "norm": 5.9846,
+      "step": 801
+    },
+    {
+      "acc_char": 0.2866,
+      "acc_token": 0.6213,
+      "epoch": 0.01,
+      "loss_char": 0.8847,
+      "loss_token": 1.7219,
+      "lr": "2.00e-07",
+      "norm": 5.528,
+      "step": 802
+    },
+    {
+      "acc_char": 0.294,
+      "acc_token": 0.6332,
+      "epoch": 0.01,
+      "loss_char": 0.8649,
+      "loss_token": 1.768,
+      "lr": "2.00e-07",
+      "norm": 5.6615,
+      "step": 803
+    },
+    {
+      "acc_char": 0.2963,
+      "acc_token": 0.6698,
+      "epoch": 0.01,
+      "loss_char": 0.7597,
+      "loss_token": 1.689,
+      "lr": "2.00e-07",
+      "norm": 5.7607,
+      "step": 804
+    },
+    {
+      "acc_char": 0.298,
+      "acc_token": 0.6756,
+      "epoch": 0.01,
+      "loss_char": 0.7265,
+      "loss_token": 1.6196,
+      "lr": "2.00e-07",
+      "norm": 5.5886,
+      "step": 805
+    },
+    {
+      "acc_char": 0.2965,
+      "acc_token": 0.6738,
+      "epoch": 0.01,
+      "loss_char": 0.7418,
+      "loss_token": 1.6831,
+      "lr": "2.00e-07",
+      "norm": 5.6464,
+      "step": 806
+    },
+    {
+      "acc_char": 0.3002,
+      "acc_token": 0.6889,
+      "epoch": 0.01,
+      "loss_char": 0.6854,
+      "loss_token": 1.6537,
+      "lr": "2.00e-07",
+      "norm": 5.7818,
+      "step": 807
+    },
+    {
+      "acc_char": 0.2867,
+      "acc_token": 0.6546,
+      "epoch": 0.01,
+      "loss_char": 0.8185,
+      "loss_token": 1.7508,
+      "lr": "2.00e-07",
+      "norm": 6.1051,
+      "step": 808
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6727,
+      "epoch": 0.01,
+      "loss_char": 0.768,
+      "loss_token": 1.6273,
+      "lr": "2.00e-07",
+      "norm": 5.7804,
+      "step": 809
+    },
+    {
+      "acc_char": 0.2798,
+      "acc_token": 0.6473,
+      "epoch": 0.01,
+      "loss_char": 0.8247,
+      "loss_token": 1.7481,
+      "lr": "2.00e-07",
+      "norm": 5.932,
+      "step": 810
+    },
+    {
+      "acc_char": 0.2984,
+      "acc_token": 0.6759,
+      "epoch": 0.01,
+      "loss_char": 0.753,
+      "loss_token": 1.6339,
+      "lr": "2.00e-07",
+      "norm": 5.5975,
+      "step": 811
+    },
+    {
+      "acc_char": 0.3008,
+      "acc_token": 0.6724,
+      "epoch": 0.01,
+      "loss_char": 0.7605,
+      "loss_token": 1.6948,
+      "lr": "2.00e-07",
+      "norm": 5.8128,
+      "step": 812
+    },
+    {
+      "acc_char": 0.3144,
+      "acc_token": 0.7123,
+      "epoch": 0.01,
+      "loss_char": 0.6638,
+      "loss_token": 1.5634,
+      "lr": "2.00e-07",
+      "norm": 5.3602,
+      "step": 813
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6854,
+      "epoch": 0.01,
+      "loss_char": 0.7074,
+      "loss_token": 1.5532,
+      "lr": "2.00e-07",
+      "norm": 5.7765,
+      "step": 814
+    },
+    {
+      "acc_char": 0.3047,
+      "acc_token": 0.6809,
+      "epoch": 0.01,
+      "loss_char": 0.7457,
+      "loss_token": 1.5699,
+      "lr": "2.00e-07",
+      "norm": 5.5826,
+      "step": 815
+    },
+    {
+      "acc_char": 0.3005,
+      "acc_token": 0.6828,
+      "epoch": 0.01,
+      "loss_char": 0.7312,
+      "loss_token": 1.559,
+      "lr": "2.00e-07",
+      "norm": 5.6165,
+      "step": 816
+    },
+    {
+      "acc_char": 0.332,
+      "acc_token": 0.7324,
+      "epoch": 0.01,
+      "loss_char": 0.6057,
+      "loss_token": 1.516,
+      "lr": "2.00e-07",
+      "norm": 5.4287,
+      "step": 817
+    },
+    {
+      "acc_char": 0.2929,
+      "acc_token": 0.6608,
+      "epoch": 0.01,
+      "loss_char": 0.779,
+      "loss_token": 1.7271,
+      "lr": "2.00e-07",
+      "norm": 5.6373,
+      "step": 818
+    },
+    {
+      "acc_char": 0.3008,
+      "acc_token": 0.6771,
+      "epoch": 0.01,
+      "loss_char": 0.7156,
+      "loss_token": 1.6334,
+      "lr": "2.00e-07",
+      "norm": 5.5915,
+      "step": 819
+    },
+    {
+      "acc_char": 0.307,
+      "acc_token": 0.6834,
+      "epoch": 0.02,
+      "loss_char": 0.7347,
+      "loss_token": 1.6243,
+      "lr": "2.00e-07",
+      "norm": 5.7642,
+      "step": 820
+    },
+    {
+      "acc_char": 0.3159,
+      "acc_token": 0.7048,
+      "epoch": 0.02,
+      "loss_char": 0.681,
+      "loss_token": 1.5652,
+      "lr": "2.00e-07",
+      "norm": 5.6864,
+      "step": 821
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6917,
+      "epoch": 0.02,
+      "loss_char": 0.6971,
+      "loss_token": 1.5949,
+      "lr": "2.00e-07",
+      "norm": 5.6179,
+      "step": 822
+    },
+    {
+      "acc_char": 0.2979,
+      "acc_token": 0.683,
+      "epoch": 0.02,
+      "loss_char": 0.7057,
+      "loss_token": 1.6006,
+      "lr": "2.00e-07",
+      "norm": 5.9728,
+      "step": 823
+    },
+    {
+      "acc_char": 0.309,
+      "acc_token": 0.6976,
+      "epoch": 0.02,
+      "loss_char": 0.6793,
+      "loss_token": 1.5622,
+      "lr": "2.00e-07",
+      "norm": 5.7353,
+      "step": 824
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6753,
+      "epoch": 0.02,
+      "loss_char": 0.7406,
+      "loss_token": 1.6784,
+      "lr": "2.00e-07",
+      "norm": 5.7657,
+      "step": 825
+    },
+    {
+      "acc_char": 0.2928,
+      "acc_token": 0.6691,
+      "epoch": 0.02,
+      "loss_char": 0.7477,
+      "loss_token": 1.6609,
+      "lr": "2.00e-07",
+      "norm": 5.9937,
+      "step": 826
+    },
+    {
+      "acc_char": 0.3049,
+      "acc_token": 0.6844,
+      "epoch": 0.02,
+      "loss_char": 0.7368,
+      "loss_token": 1.6281,
+      "lr": "2.00e-07",
+      "norm": 5.7378,
+      "step": 827
+    },
+    {
+      "acc_char": 0.2987,
+      "acc_token": 0.6795,
+      "epoch": 0.02,
+      "loss_char": 0.718,
+      "loss_token": 1.5856,
+      "lr": "2.00e-07",
+      "norm": 5.6599,
+      "step": 828
+    },
+    {
+      "acc_char": 0.3048,
+      "acc_token": 0.7048,
+      "epoch": 0.02,
+      "loss_char": 0.6446,
+      "loss_token": 1.5444,
+      "lr": "2.00e-07",
+      "norm": 5.5,
+      "step": 829
+    },
+    {
+      "acc_char": 0.3019,
+      "acc_token": 0.6809,
+      "epoch": 0.02,
+      "loss_char": 0.7433,
+      "loss_token": 1.636,
+      "lr": "2.00e-07",
+      "norm": 5.4755,
+      "step": 830
+    },
+    {
+      "acc_char": 0.3,
+      "acc_token": 0.6857,
+      "epoch": 0.02,
+      "loss_char": 0.6815,
+      "loss_token": 1.5521,
+      "lr": "2.00e-07",
+      "norm": 5.5819,
+      "step": 831
+    },
+    {
+      "acc_char": 0.2923,
+      "acc_token": 0.6756,
+      "epoch": 0.02,
+      "loss_char": 0.7457,
+      "loss_token": 1.6655,
+      "lr": "2.00e-07",
+      "norm": 6.0816,
+      "step": 832
+    },
+    {
+      "acc_char": 0.2992,
+      "acc_token": 0.6766,
+      "epoch": 0.02,
+      "loss_char": 0.7569,
+      "loss_token": 1.6527,
+      "lr": "2.00e-07",
+      "norm": 5.7566,
+      "step": 833
+    },
+    {
+      "acc_char": 0.3021,
+      "acc_token": 0.6779,
+      "epoch": 0.02,
+      "loss_char": 0.7349,
+      "loss_token": 1.6231,
+      "lr": "2.00e-07",
+      "norm": 5.5129,
+      "step": 834
+    },
+    {
+      "acc_char": 0.3048,
+      "acc_token": 0.6798,
+      "epoch": 0.02,
+      "loss_char": 0.729,
+      "loss_token": 1.5387,
+      "lr": "2.00e-07",
+      "norm": 5.4065,
+      "step": 835
+    },
+    {
+      "acc_char": 0.2927,
+      "acc_token": 0.6659,
+      "epoch": 0.02,
+      "loss_char": 0.7933,
+      "loss_token": 1.6665,
+      "lr": "2.00e-07",
+      "norm": 5.7373,
+      "step": 836
+    },
+    {
+      "acc_char": 0.3056,
+      "acc_token": 0.6878,
+      "epoch": 0.02,
+      "loss_char": 0.7152,
+      "loss_token": 1.5988,
+      "lr": "2.00e-07",
+      "norm": 6.0286,
+      "step": 837
+    },
+    {
+      "acc_char": 0.3144,
+      "acc_token": 0.6894,
+      "epoch": 0.02,
+      "loss_char": 0.7078,
+      "loss_token": 1.6339,
+      "lr": "2.00e-07",
+      "norm": 5.8277,
+      "step": 838
+    },
+    {
+      "acc_char": 0.3033,
+      "acc_token": 0.6712,
+      "epoch": 0.02,
+      "loss_char": 0.7581,
+      "loss_token": 1.6225,
+      "lr": "2.00e-07",
+      "norm": 5.7859,
+      "step": 839
+    },
+    {
+      "acc_char": 0.3126,
+      "acc_token": 0.6887,
+      "epoch": 0.02,
+      "loss_char": 0.7221,
+      "loss_token": 1.5797,
+      "lr": "2.00e-07",
+      "norm": 5.7404,
+      "step": 840
+    },
+    {
+      "acc_char": 0.3171,
+      "acc_token": 0.7202,
+      "epoch": 0.02,
+      "loss_char": 0.5949,
+      "loss_token": 1.5088,
+      "lr": "2.00e-07",
+      "norm": 5.521,
+      "step": 841
+    },
+    {
+      "acc_char": 0.2944,
+      "acc_token": 0.6671,
+      "epoch": 0.02,
+      "loss_char": 0.7784,
+      "loss_token": 1.5698,
+      "lr": "2.00e-07",
+      "norm": 5.5683,
+      "step": 842
+    },
+    {
+      "acc_char": 0.3136,
+      "acc_token": 0.6978,
+      "epoch": 0.02,
+      "loss_char": 0.6885,
+      "loss_token": 1.539,
+      "lr": "2.00e-07",
+      "norm": 5.6271,
+      "step": 843
+    },
+    {
+      "acc_char": 0.3025,
+      "acc_token": 0.6963,
+      "epoch": 0.02,
+      "loss_char": 0.6713,
+      "loss_token": 1.5745,
+      "lr": "2.00e-07",
+      "norm": 5.531,
+      "step": 844
+    },
+    {
+      "acc_char": 0.3143,
+      "acc_token": 0.7153,
+      "epoch": 0.02,
+      "loss_char": 0.6429,
+      "loss_token": 1.5588,
+      "lr": "2.00e-07",
+      "norm": 5.8076,
+      "step": 845
+    },
+    {
+      "acc_char": 0.3135,
+      "acc_token": 0.7175,
+      "epoch": 0.02,
+      "loss_char": 0.6098,
+      "loss_token": 1.5226,
+      "lr": "2.00e-07",
+      "norm": 5.5064,
+      "step": 846
+    },
+    {
+      "acc_char": 0.3018,
+      "acc_token": 0.6906,
+      "epoch": 0.02,
+      "loss_char": 0.6871,
+      "loss_token": 1.6122,
+      "lr": "2.00e-07",
+      "norm": 5.6541,
+      "step": 847
+    },
+    {
+      "acc_char": 0.312,
+      "acc_token": 0.6916,
+      "epoch": 0.02,
+      "loss_char": 0.7166,
+      "loss_token": 1.5057,
+      "lr": "2.00e-07",
+      "norm": 5.3657,
+      "step": 848
+    },
+    {
+      "acc_char": 0.3038,
+      "acc_token": 0.6949,
+      "epoch": 0.02,
+      "loss_char": 0.6808,
+      "loss_token": 1.5474,
+      "lr": "2.00e-07",
+      "norm": 5.3944,
+      "step": 849
+    },
+    {
+      "acc_char": 0.2928,
+      "acc_token": 0.6665,
+      "epoch": 0.02,
+      "loss_char": 0.7741,
+      "loss_token": 1.7224,
+      "lr": "2.00e-07",
+      "norm": 6.0661,
+      "step": 850
+    },
+    {
+      "acc_char": 0.3188,
+      "acc_token": 0.7099,
+      "epoch": 0.02,
+      "loss_char": 0.6563,
+      "loss_token": 1.5444,
+      "lr": "2.00e-07",
+      "norm": 5.7398,
+      "step": 851
+    },
+    {
+      "acc_char": 0.3117,
+      "acc_token": 0.7085,
+      "epoch": 0.02,
+      "loss_char": 0.6725,
+      "loss_token": 1.6115,
+      "lr": "2.00e-07",
+      "norm": 6.1062,
+      "step": 852
+    },
+    {
+      "acc_char": 0.2951,
+      "acc_token": 0.6786,
+      "epoch": 0.02,
+      "loss_char": 0.7135,
+      "loss_token": 1.586,
+      "lr": "2.00e-07",
+      "norm": 5.7301,
+      "step": 853
+    },
+    {
+      "acc_char": 0.3145,
+      "acc_token": 0.7019,
+      "epoch": 0.02,
+      "loss_char": 0.668,
+      "loss_token": 1.5912,
+      "lr": "2.00e-07",
+      "norm": 5.4874,
+      "step": 854
+    },
+    {
+      "acc_char": 0.3146,
+      "acc_token": 0.7107,
+      "epoch": 0.02,
+      "loss_char": 0.6492,
+      "loss_token": 1.5739,
+      "lr": "2.00e-07",
+      "norm": 5.5151,
+      "step": 855
+    },
+    {
+      "acc_char": 0.3019,
+      "acc_token": 0.6879,
+      "epoch": 0.02,
+      "loss_char": 0.7229,
+      "loss_token": 1.5687,
+      "lr": "2.00e-07",
+      "norm": 5.5906,
+      "step": 856
+    },
+    {
+      "acc_char": 0.2992,
+      "acc_token": 0.6701,
+      "epoch": 0.02,
+      "loss_char": 0.7858,
+      "loss_token": 1.6602,
+      "lr": "2.00e-07",
+      "norm": 6.0501,
+      "step": 857
+    },
+    {
+      "acc_char": 0.2985,
+      "acc_token": 0.6726,
+      "epoch": 0.02,
+      "loss_char": 0.7506,
+      "loss_token": 1.5568,
+      "lr": "2.00e-07",
+      "norm": 5.7594,
+      "step": 858
+    },
+    {
+      "acc_char": 0.283,
+      "acc_token": 0.651,
+      "epoch": 0.02,
+      "loss_char": 0.8111,
+      "loss_token": 1.7399,
+      "lr": "2.00e-07",
+      "norm": 5.8941,
+      "step": 859
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.6891,
+      "epoch": 0.02,
+      "loss_char": 0.6818,
+      "loss_token": 1.5738,
+      "lr": "2.00e-07",
+      "norm": 5.8817,
+      "step": 860
+    },
+    {
+      "acc_char": 0.3067,
+      "acc_token": 0.6701,
+      "epoch": 0.02,
+      "loss_char": 0.7807,
+      "loss_token": 1.6976,
+      "lr": "2.00e-07",
+      "norm": 5.9143,
+      "step": 861
+    },
+    {
+      "acc_char": 0.3062,
+      "acc_token": 0.7005,
+      "epoch": 0.02,
+      "loss_char": 0.6652,
+      "loss_token": 1.5703,
+      "lr": "2.00e-07",
+      "norm": 5.8149,
+      "step": 862
+    },
+    {
+      "acc_char": 0.3088,
+      "acc_token": 0.7021,
+      "epoch": 0.02,
+      "loss_char": 0.6614,
+      "loss_token": 1.5577,
+      "lr": "2.00e-07",
+      "norm": 5.619,
+      "step": 863
+    },
+    {
+      "acc_char": 0.31,
+      "acc_token": 0.6976,
+      "epoch": 0.02,
+      "loss_char": 0.6798,
+      "loss_token": 1.5745,
+      "lr": "2.00e-07",
+      "norm": 5.7408,
+      "step": 864
+    },
+    {
+      "acc_char": 0.3075,
+      "acc_token": 0.6924,
+      "epoch": 0.02,
+      "loss_char": 0.7154,
+      "loss_token": 1.5676,
+      "lr": "2.00e-07",
+      "norm": 5.6581,
+      "step": 865
+    },
+    {
+      "acc_char": 0.3092,
+      "acc_token": 0.709,
+      "epoch": 0.02,
+      "loss_char": 0.6473,
+      "loss_token": 1.4729,
+      "lr": "2.00e-07",
+      "norm": 5.2921,
+      "step": 866
+    },
+    {
+      "acc_char": 0.2904,
+      "acc_token": 0.6714,
+      "epoch": 0.02,
+      "loss_char": 0.7233,
+      "loss_token": 1.6225,
+      "lr": "2.00e-07",
+      "norm": 5.5076,
+      "step": 867
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.6996,
+      "epoch": 0.02,
+      "loss_char": 0.6696,
+      "loss_token": 1.6228,
+      "lr": "2.00e-07",
+      "norm": 5.6327,
+      "step": 868
+    },
+    {
+      "acc_char": 0.2991,
+      "acc_token": 0.6821,
+      "epoch": 0.02,
+      "loss_char": 0.7096,
+      "loss_token": 1.6779,
+      "lr": "2.00e-07",
+      "norm": 5.7236,
+      "step": 869
+    },
+    {
+      "acc_char": 0.2987,
+      "acc_token": 0.6916,
+      "epoch": 0.02,
+      "loss_char": 0.6837,
+      "loss_token": 1.6087,
+      "lr": "2.00e-07",
+      "norm": 5.6206,
+      "step": 870
+    },
+    {
+      "acc_char": 0.3074,
+      "acc_token": 0.6979,
+      "epoch": 0.02,
+      "loss_char": 0.663,
+      "loss_token": 1.5133,
+      "lr": "2.00e-07",
+      "norm": 5.5679,
+      "step": 871
+    },
+    {
+      "acc_char": 0.2908,
+      "acc_token": 0.6608,
+      "epoch": 0.02,
+      "loss_char": 0.7855,
+      "loss_token": 1.6526,
+      "lr": "2.00e-07",
+      "norm": 5.7404,
+      "step": 872
+    },
+    {
+      "acc_char": 0.2957,
+      "acc_token": 0.6746,
+      "epoch": 0.02,
+      "loss_char": 0.7258,
+      "loss_token": 1.6516,
+      "lr": "2.00e-07",
+      "norm": 5.6479,
+      "step": 873
+    },
+    {
+      "acc_char": 0.305,
+      "acc_token": 0.6764,
+      "epoch": 0.02,
+      "loss_char": 0.7713,
+      "loss_token": 1.6556,
+      "lr": "2.00e-07",
+      "norm": 5.7678,
+      "step": 874
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6732,
+      "epoch": 0.02,
+      "loss_char": 0.7498,
+      "loss_token": 1.737,
+      "lr": "2.00e-07",
+      "norm": 5.9202,
+      "step": 875
+    },
+    {
+      "acc_char": 0.3012,
+      "acc_token": 0.6684,
+      "epoch": 0.02,
+      "loss_char": 0.785,
+      "loss_token": 1.6429,
+      "lr": "2.00e-07",
+      "norm": 5.6248,
+      "step": 876
+    },
+    {
+      "acc_char": 0.3156,
+      "acc_token": 0.7224,
+      "epoch": 0.02,
+      "loss_char": 0.6204,
+      "loss_token": 1.5703,
+      "lr": "2.00e-07",
+      "norm": 5.5374,
+      "step": 877
+    },
+    {
+      "acc_char": 0.2873,
+      "acc_token": 0.6655,
+      "epoch": 0.02,
+      "loss_char": 0.7559,
+      "loss_token": 1.707,
+      "lr": "2.00e-07",
+      "norm": 5.889,
+      "step": 878
+    },
+    {
+      "acc_char": 0.2936,
+      "acc_token": 0.6815,
+      "epoch": 0.02,
+      "loss_char": 0.7047,
+      "loss_token": 1.621,
+      "lr": "2.00e-07",
+      "norm": 5.7028,
+      "step": 879
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6811,
+      "epoch": 0.02,
+      "loss_char": 0.7095,
+      "loss_token": 1.5844,
+      "lr": "2.00e-07",
+      "norm": 5.574,
+      "step": 880
+    },
+    {
+      "acc_char": 0.3103,
+      "acc_token": 0.6643,
+      "epoch": 0.02,
+      "loss_char": 0.8205,
+      "loss_token": 1.7376,
+      "lr": "2.00e-07",
+      "norm": 5.7651,
+      "step": 881
+    },
+    {
+      "acc_char": 0.2889,
+      "acc_token": 0.6652,
+      "epoch": 0.02,
+      "loss_char": 0.7952,
+      "loss_token": 1.7347,
+      "lr": "2.00e-07",
+      "norm": 11.8228,
+      "step": 882
+    },
+    {
+      "acc_char": 0.2937,
+      "acc_token": 0.664,
+      "epoch": 0.02,
+      "loss_char": 0.7846,
+      "loss_token": 1.6037,
+      "lr": "2.00e-07",
+      "norm": 5.556,
+      "step": 883
+    },
+    {
+      "acc_char": 0.3034,
+      "acc_token": 0.6775,
+      "epoch": 0.02,
+      "loss_char": 0.7591,
+      "loss_token": 1.6535,
+      "lr": "2.00e-07",
+      "norm": 5.5967,
+      "step": 884
+    },
+    {
+      "acc_char": 0.3091,
+      "acc_token": 0.6898,
+      "epoch": 0.02,
+      "loss_char": 0.7304,
+      "loss_token": 1.594,
+      "lr": "2.00e-07",
+      "norm": 5.4046,
+      "step": 885
+    },
+    {
+      "acc_char": 0.2999,
+      "acc_token": 0.6882,
+      "epoch": 0.02,
+      "loss_char": 0.6872,
+      "loss_token": 1.6064,
+      "lr": "2.00e-07",
+      "norm": 5.8358,
+      "step": 886
+    },
+    {
+      "acc_char": 0.3042,
+      "acc_token": 0.6817,
+      "epoch": 0.02,
+      "loss_char": 0.7236,
+      "loss_token": 1.7433,
+      "lr": "2.00e-07",
+      "norm": 5.7178,
+      "step": 887
+    },
+    {
+      "acc_char": 0.2988,
+      "acc_token": 0.6754,
+      "epoch": 0.02,
+      "loss_char": 0.729,
+      "loss_token": 1.6565,
+      "lr": "2.00e-07",
+      "norm": 5.9591,
+      "step": 888
+    },
+    {
+      "acc_char": 0.2986,
+      "acc_token": 0.6801,
+      "epoch": 0.02,
+      "loss_char": 0.7253,
+      "loss_token": 1.6613,
+      "lr": "2.00e-07",
+      "norm": 5.7309,
+      "step": 889
+    },
+    {
+      "acc_char": 0.295,
+      "acc_token": 0.6737,
+      "epoch": 0.02,
+      "loss_char": 0.7399,
+      "loss_token": 1.6015,
+      "lr": "2.00e-07",
+      "norm": 5.4777,
+      "step": 890
+    },
+    {
+      "acc_char": 0.2896,
+      "acc_token": 0.666,
+      "epoch": 0.02,
+      "loss_char": 0.7702,
+      "loss_token": 1.6573,
+      "lr": "2.00e-07",
+      "norm": 6.1023,
+      "step": 891
+    },
+    {
+      "acc_char": 0.2845,
+      "acc_token": 0.6454,
+      "epoch": 0.02,
+      "loss_char": 0.8636,
+      "loss_token": 1.7583,
+      "lr": "2.00e-07",
+      "norm": 5.9048,
+      "step": 892
+    },
+    {
+      "acc_char": 0.2889,
+      "acc_token": 0.6415,
+      "epoch": 0.02,
+      "loss_char": 0.8748,
+      "loss_token": 1.7602,
+      "lr": "2.00e-07",
+      "norm": 5.9197,
+      "step": 893
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6774,
+      "epoch": 0.02,
+      "loss_char": 0.7214,
+      "loss_token": 1.6364,
+      "lr": "2.00e-07",
+      "norm": 5.4568,
+      "step": 894
+    },
+    {
+      "acc_char": 0.3245,
+      "acc_token": 0.7212,
+      "epoch": 0.02,
+      "loss_char": 0.6279,
+      "loss_token": 1.466,
+      "lr": "2.00e-07",
+      "norm": 5.5303,
+      "step": 895
+    },
+    {
+      "acc_char": 0.3241,
+      "acc_token": 0.7322,
+      "epoch": 0.02,
+      "loss_char": 0.5878,
+      "loss_token": 1.4649,
+      "lr": "2.00e-07",
+      "norm": 5.4387,
+      "step": 896
+    },
+    {
+      "acc_char": 0.3078,
+      "acc_token": 0.6898,
+      "epoch": 0.02,
+      "loss_char": 0.7139,
+      "loss_token": 1.5937,
+      "lr": "2.00e-07",
+      "norm": 5.5562,
+      "step": 897
+    },
+    {
+      "acc_char": 0.2982,
+      "acc_token": 0.6713,
+      "epoch": 0.02,
+      "loss_char": 0.7894,
+      "loss_token": 1.685,
+      "lr": "2.00e-07",
+      "norm": 5.6968,
+      "step": 898
+    },
+    {
+      "acc_char": 0.2958,
+      "acc_token": 0.6649,
+      "epoch": 0.02,
+      "loss_char": 0.7826,
+      "loss_token": 1.7361,
+      "lr": "2.00e-07",
+      "norm": 6.044,
+      "step": 899
+    },
+    {
+      "acc_char": 0.3078,
+      "acc_token": 0.6906,
+      "epoch": 0.02,
+      "loss_char": 0.7139,
+      "loss_token": 1.5897,
+      "lr": "2.00e-07",
+      "norm": 5.6098,
+      "step": 900
+    },
+    {
+      "acc_char": 0.2989,
+      "acc_token": 0.6848,
+      "epoch": 0.02,
+      "loss_char": 0.7267,
+      "loss_token": 1.6437,
+      "lr": "2.00e-07",
+      "norm": 6.034,
+      "step": 901
+    },
+    {
+      "acc_char": 0.3091,
+      "acc_token": 0.6975,
+      "epoch": 0.02,
+      "loss_char": 0.7161,
+      "loss_token": 1.5962,
+      "lr": "2.00e-07",
+      "norm": 5.8044,
+      "step": 902
+    },
+    {
+      "acc_char": 0.308,
+      "acc_token": 0.6939,
+      "epoch": 0.02,
+      "loss_char": 0.6884,
+      "loss_token": 1.585,
+      "lr": "2.00e-07",
+      "norm": 5.7306,
+      "step": 903
+    },
+    {
+      "acc_char": 0.3086,
+      "acc_token": 0.6925,
+      "epoch": 0.02,
+      "loss_char": 0.7179,
+      "loss_token": 1.6073,
+      "lr": "2.00e-07",
+      "norm": 5.7369,
+      "step": 904
+    },
+    {
+      "acc_char": 0.2908,
+      "acc_token": 0.6717,
+      "epoch": 0.02,
+      "loss_char": 0.7543,
+      "loss_token": 1.6996,
+      "lr": "2.00e-07",
+      "norm": 5.7256,
+      "step": 905
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6958,
+      "epoch": 0.02,
+      "loss_char": 0.6856,
+      "loss_token": 1.585,
+      "lr": "2.00e-07",
+      "norm": 5.6225,
+      "step": 906
+    },
+    {
+      "acc_char": 0.3088,
+      "acc_token": 0.6996,
+      "epoch": 0.02,
+      "loss_char": 0.6747,
+      "loss_token": 1.5768,
+      "lr": "2.00e-07",
+      "norm": 5.7417,
+      "step": 907
+    },
+    {
+      "acc_char": 0.3177,
+      "acc_token": 0.7195,
+      "epoch": 0.02,
+      "loss_char": 0.6313,
+      "loss_token": 1.5231,
+      "lr": "2.00e-07",
+      "norm": 5.7196,
+      "step": 908
+    },
+    {
+      "acc_char": 0.3083,
+      "acc_token": 0.6958,
+      "epoch": 0.02,
+      "loss_char": 0.6694,
+      "loss_token": 1.635,
+      "lr": "2.00e-07",
+      "norm": 6.0173,
+      "step": 909
+    },
+    {
+      "acc_char": 0.2915,
+      "acc_token": 0.6698,
+      "epoch": 0.02,
+      "loss_char": 0.7286,
+      "loss_token": 1.6788,
+      "lr": "2.00e-07",
+      "norm": 5.6674,
+      "step": 910
+    },
+    {
+      "acc_char": 0.3072,
+      "acc_token": 0.689,
+      "epoch": 0.02,
+      "loss_char": 0.7164,
+      "loss_token": 1.6722,
+      "lr": "2.00e-07",
+      "norm": 5.83,
+      "step": 911
+    },
+    {
+      "acc_char": 0.2934,
+      "acc_token": 0.6323,
+      "epoch": 0.02,
+      "loss_char": 0.8456,
+      "loss_token": 1.7197,
+      "lr": "2.00e-07",
+      "norm": 5.5935,
+      "step": 912
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.679,
+      "epoch": 0.02,
+      "loss_char": 0.728,
+      "loss_token": 1.6382,
+      "lr": "2.00e-07",
+      "norm": 5.7187,
+      "step": 913
+    },
+    {
+      "acc_char": 0.3033,
+      "acc_token": 0.6892,
+      "epoch": 0.02,
+      "loss_char": 0.7052,
+      "loss_token": 1.5839,
+      "lr": "2.00e-07",
+      "norm": 5.7948,
+      "step": 914
+    },
+    {
+      "acc_char": 0.3211,
+      "acc_token": 0.7173,
+      "epoch": 0.02,
+      "loss_char": 0.6406,
+      "loss_token": 1.5502,
+      "lr": "2.00e-07",
+      "norm": 5.5802,
+      "step": 915
+    },
+    {
+      "acc_char": 0.2919,
+      "acc_token": 0.6749,
+      "epoch": 0.02,
+      "loss_char": 0.7306,
+      "loss_token": 1.6292,
+      "lr": "2.00e-07",
+      "norm": 5.6494,
+      "step": 916
+    },
+    {
+      "acc_char": 0.2947,
+      "acc_token": 0.6714,
+      "epoch": 0.02,
+      "loss_char": 0.7284,
+      "loss_token": 1.692,
+      "lr": "2.00e-07",
+      "norm": 5.5497,
+      "step": 917
+    },
+    {
+      "acc_char": 0.3015,
+      "acc_token": 0.6835,
+      "epoch": 0.02,
+      "loss_char": 0.7163,
+      "loss_token": 1.5713,
+      "lr": "2.00e-07",
+      "norm": 5.8363,
+      "step": 918
+    },
+    {
+      "acc_char": 0.3016,
+      "acc_token": 0.6719,
+      "epoch": 0.02,
+      "loss_char": 0.7656,
+      "loss_token": 1.7156,
+      "lr": "2.00e-07",
+      "norm": 5.9585,
+      "step": 919
+    },
+    {
+      "acc_char": 0.2974,
+      "acc_token": 0.686,
+      "epoch": 0.02,
+      "loss_char": 0.706,
+      "loss_token": 1.6797,
+      "lr": "2.00e-07",
+      "norm": 5.858,
+      "step": 920
+    },
+    {
+      "acc_char": 0.3023,
+      "acc_token": 0.6833,
+      "epoch": 0.02,
+      "loss_char": 0.737,
+      "loss_token": 1.6113,
+      "lr": "2.00e-07",
+      "norm": 5.7094,
+      "step": 921
+    },
+    {
+      "acc_char": 0.2893,
+      "acc_token": 0.6675,
+      "epoch": 0.02,
+      "loss_char": 0.7461,
+      "loss_token": 1.6808,
+      "lr": "2.00e-07",
+      "norm": 5.8515,
+      "step": 922
+    },
+    {
+      "acc_char": 0.2925,
+      "acc_token": 0.634,
+      "epoch": 0.02,
+      "loss_char": 0.8618,
+      "loss_token": 1.7723,
+      "lr": "2.00e-07",
+      "norm": 5.5071,
+      "step": 923
+    },
+    {
+      "acc_char": 0.3084,
+      "acc_token": 0.7052,
+      "epoch": 0.02,
+      "loss_char": 0.6395,
+      "loss_token": 1.6118,
+      "lr": "2.00e-07",
+      "norm": 6.0085,
+      "step": 924
+    },
+    {
+      "acc_char": 0.3117,
+      "acc_token": 0.6981,
+      "epoch": 0.02,
+      "loss_char": 0.686,
+      "loss_token": 1.5,
+      "lr": "2.00e-07",
+      "norm": 6.0124,
+      "step": 925
+    },
+    {
+      "acc_char": 0.3039,
+      "acc_token": 0.6853,
+      "epoch": 0.02,
+      "loss_char": 0.7184,
+      "loss_token": 1.6244,
+      "lr": "2.00e-07",
+      "norm": 5.9832,
+      "step": 926
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6876,
+      "epoch": 0.02,
+      "loss_char": 0.6786,
+      "loss_token": 1.6284,
+      "lr": "2.00e-07",
+      "norm": 5.725,
+      "step": 927
+    },
+    {
+      "acc_char": 0.3066,
+      "acc_token": 0.7078,
+      "epoch": 0.02,
+      "loss_char": 0.6441,
+      "loss_token": 1.4689,
+      "lr": "2.00e-07",
+      "norm": 5.4178,
+      "step": 928
+    },
+    {
+      "acc_char": 0.3015,
+      "acc_token": 0.6397,
+      "epoch": 0.02,
+      "loss_char": 0.8701,
+      "loss_token": 1.7746,
+      "lr": "2.00e-07",
+      "norm": 5.6169,
+      "step": 929
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6762,
+      "epoch": 0.02,
+      "loss_char": 0.7557,
+      "loss_token": 1.6987,
+      "lr": "2.00e-07",
+      "norm": 5.8867,
+      "step": 930
+    },
+    {
+      "acc_char": 0.2994,
+      "acc_token": 0.672,
+      "epoch": 0.02,
+      "loss_char": 0.7681,
+      "loss_token": 1.7252,
+      "lr": "2.00e-07",
+      "norm": 5.9723,
+      "step": 931
+    },
+    {
+      "acc_char": 0.3008,
+      "acc_token": 0.677,
+      "epoch": 0.02,
+      "loss_char": 0.7565,
+      "loss_token": 1.706,
+      "lr": "2.00e-07",
+      "norm": 5.9506,
+      "step": 932
+    },
+    {
+      "acc_char": 0.3066,
+      "acc_token": 0.6942,
+      "epoch": 0.02,
+      "loss_char": 0.6873,
+      "loss_token": 1.5338,
+      "lr": "2.00e-07",
+      "norm": 5.9416,
+      "step": 933
+    },
+    {
+      "acc_char": 0.3081,
+      "acc_token": 0.7002,
+      "epoch": 0.02,
+      "loss_char": 0.6803,
+      "loss_token": 1.5462,
+      "lr": "2.00e-07",
+      "norm": 5.6716,
+      "step": 934
+    },
+    {
+      "acc_char": 0.3191,
+      "acc_token": 0.7165,
+      "epoch": 0.02,
+      "loss_char": 0.6339,
+      "loss_token": 1.493,
+      "lr": "2.00e-07",
+      "norm": 5.4216,
+      "step": 935
+    },
+    {
+      "acc_char": 0.2913,
+      "acc_token": 0.6643,
+      "epoch": 0.02,
+      "loss_char": 0.7805,
+      "loss_token": 1.7313,
+      "lr": "2.00e-07",
+      "norm": 5.9457,
+      "step": 936
+    },
+    {
+      "acc_char": 0.3171,
+      "acc_token": 0.7187,
+      "epoch": 0.02,
+      "loss_char": 0.6301,
+      "loss_token": 1.5637,
+      "lr": "2.00e-07",
+      "norm": 5.7545,
+      "step": 937
+    },
+    {
+      "acc_char": 0.3084,
+      "acc_token": 0.6958,
+      "epoch": 0.02,
+      "loss_char": 0.6851,
+      "loss_token": 1.5865,
+      "lr": "2.00e-07",
+      "norm": 5.792,
+      "step": 938
+    },
+    {
+      "acc_char": 0.2989,
+      "acc_token": 0.6709,
+      "epoch": 0.02,
+      "loss_char": 0.7519,
+      "loss_token": 1.6043,
+      "lr": "2.00e-07",
+      "norm": 5.531,
+      "step": 939
+    },
+    {
+      "acc_char": 0.2991,
+      "acc_token": 0.6756,
+      "epoch": 0.02,
+      "loss_char": 0.7521,
+      "loss_token": 1.6099,
+      "lr": "2.00e-07",
+      "norm": 5.4071,
+      "step": 940
+    },
+    {
+      "acc_char": 0.2772,
+      "acc_token": 0.6351,
+      "epoch": 0.02,
+      "loss_char": 0.8474,
+      "loss_token": 1.7509,
+      "lr": "2.00e-07",
+      "norm": 5.956,
+      "step": 941
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6867,
+      "epoch": 0.02,
+      "loss_char": 0.7244,
+      "loss_token": 1.66,
+      "lr": "2.00e-07",
+      "norm": 5.5928,
+      "step": 942
+    },
+    {
+      "acc_char": 0.2985,
+      "acc_token": 0.6779,
+      "epoch": 0.02,
+      "loss_char": 0.7319,
+      "loss_token": 1.7502,
+      "lr": "2.00e-07",
+      "norm": 5.6872,
+      "step": 943
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6762,
+      "epoch": 0.02,
+      "loss_char": 0.7591,
+      "loss_token": 1.6157,
+      "lr": "2.00e-07",
+      "norm": 5.5877,
+      "step": 944
+    },
+    {
+      "acc_char": 0.2953,
+      "acc_token": 0.6704,
+      "epoch": 0.02,
+      "loss_char": 0.7453,
+      "loss_token": 1.549,
+      "lr": "2.00e-07",
+      "norm": 5.4127,
+      "step": 945
+    },
+    {
+      "acc_char": 0.3025,
+      "acc_token": 0.6776,
+      "epoch": 0.02,
+      "loss_char": 0.7399,
+      "loss_token": 1.5651,
+      "lr": "2.00e-07",
+      "norm": 5.3875,
+      "step": 946
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6851,
+      "epoch": 0.02,
+      "loss_char": 0.7664,
+      "loss_token": 1.6692,
+      "lr": "2.00e-07",
+      "norm": 6.3977,
+      "step": 947
+    },
+    {
+      "acc_char": 0.2952,
+      "acc_token": 0.6804,
+      "epoch": 0.02,
+      "loss_char": 0.7196,
+      "loss_token": 1.5708,
+      "lr": "2.00e-07",
+      "norm": 5.3926,
+      "step": 948
+    },
+    {
+      "acc_char": 0.3033,
+      "acc_token": 0.6881,
+      "epoch": 0.02,
+      "loss_char": 0.7196,
+      "loss_token": 1.6602,
+      "lr": "2.00e-07",
+      "norm": 5.7608,
+      "step": 949
+    },
+    {
+      "acc_char": 0.2967,
+      "acc_token": 0.6757,
+      "epoch": 0.02,
+      "loss_char": 0.7133,
+      "loss_token": 1.4978,
+      "lr": "2.00e-07",
+      "norm": 5.4931,
+      "step": 950
+    },
+    {
+      "acc_char": 0.2858,
+      "acc_token": 0.6277,
+      "epoch": 0.02,
+      "loss_char": 0.8623,
+      "loss_token": 1.6958,
+      "lr": "2.00e-07",
+      "norm": 5.3998,
+      "step": 951
+    },
+    {
+      "acc_char": 0.2997,
+      "acc_token": 0.6793,
+      "epoch": 0.02,
+      "loss_char": 0.7242,
+      "loss_token": 1.5484,
+      "lr": "2.00e-07",
+      "norm": 5.8509,
+      "step": 952
+    },
+    {
+      "acc_char": 0.2958,
+      "acc_token": 0.6784,
+      "epoch": 0.02,
+      "loss_char": 0.7124,
+      "loss_token": 1.6386,
+      "lr": "2.00e-07",
+      "norm": 5.6745,
+      "step": 953
+    },
+    {
+      "acc_char": 0.3102,
+      "acc_token": 0.6879,
+      "epoch": 0.02,
+      "loss_char": 0.7138,
+      "loss_token": 1.5752,
+      "lr": "2.00e-07",
+      "norm": 5.6627,
+      "step": 954
+    },
+    {
+      "acc_char": 0.314,
+      "acc_token": 0.7037,
+      "epoch": 0.02,
+      "loss_char": 0.6819,
+      "loss_token": 1.6394,
+      "lr": "2.00e-07",
+      "norm": 5.8967,
+      "step": 955
+    },
+    {
+      "acc_char": 0.295,
+      "acc_token": 0.6637,
+      "epoch": 0.02,
+      "loss_char": 0.787,
+      "loss_token": 1.7269,
+      "lr": "2.00e-07",
+      "norm": 6.0271,
+      "step": 956
+    },
+    {
+      "acc_char": 0.2884,
+      "acc_token": 0.6678,
+      "epoch": 0.02,
+      "loss_char": 0.7588,
+      "loss_token": 1.6674,
+      "lr": "2.00e-07",
+      "norm": 5.669,
+      "step": 957
+    },
+    {
+      "acc_char": 0.2978,
+      "acc_token": 0.6848,
+      "epoch": 0.02,
+      "loss_char": 0.7159,
+      "loss_token": 1.567,
+      "lr": "2.00e-07",
+      "norm": 5.3287,
+      "step": 958
+    },
+    {
+      "acc_char": 0.294,
+      "acc_token": 0.6637,
+      "epoch": 0.02,
+      "loss_char": 0.7728,
+      "loss_token": 1.6319,
+      "lr": "2.00e-07",
+      "norm": 5.9871,
+      "step": 959
+    },
+    {
+      "acc_char": 0.2899,
+      "acc_token": 0.672,
+      "epoch": 0.02,
+      "loss_char": 0.7255,
+      "loss_token": 1.6181,
+      "lr": "2.00e-07",
+      "norm": 5.4521,
+      "step": 960
+    },
+    {
+      "acc_char": 0.3153,
+      "acc_token": 0.7082,
+      "epoch": 0.02,
+      "loss_char": 0.6676,
+      "loss_token": 1.6058,
+      "lr": "2.00e-07",
+      "norm": 6.0036,
+      "step": 961
+    },
+    {
+      "acc_char": 0.2989,
+      "acc_token": 0.6886,
+      "epoch": 0.02,
+      "loss_char": 0.6855,
+      "loss_token": 1.6063,
+      "lr": "2.00e-07",
+      "norm": 5.8428,
+      "step": 962
+    },
+    {
+      "acc_char": 0.2953,
+      "acc_token": 0.6828,
+      "epoch": 0.02,
+      "loss_char": 0.6914,
+      "loss_token": 1.6176,
+      "lr": "2.00e-07",
+      "norm": 5.704,
+      "step": 963
+    },
+    {
+      "acc_char": 0.2963,
+      "acc_token": 0.6692,
+      "epoch": 0.02,
+      "loss_char": 0.7661,
+      "loss_token": 1.6847,
+      "lr": "2.00e-07",
+      "norm": 5.8579,
+      "step": 964
+    },
+    {
+      "acc_char": 0.303,
+      "acc_token": 0.6859,
+      "epoch": 0.02,
+      "loss_char": 0.7015,
+      "loss_token": 1.5478,
+      "lr": "2.00e-07",
+      "norm": 6.1629,
+      "step": 965
+    },
+    {
+      "acc_char": 0.3015,
+      "acc_token": 0.6883,
+      "epoch": 0.02,
+      "loss_char": 0.7104,
+      "loss_token": 1.6628,
+      "lr": "2.00e-07",
+      "norm": 5.7657,
+      "step": 966
+    },
+    {
+      "acc_char": 0.2957,
+      "acc_token": 0.6608,
+      "epoch": 0.02,
+      "loss_char": 0.7851,
+      "loss_token": 1.6136,
+      "lr": "2.00e-07",
+      "norm": 5.6574,
+      "step": 967
+    },
+    {
+      "acc_char": 0.2931,
+      "acc_token": 0.6716,
+      "epoch": 0.02,
+      "loss_char": 0.7285,
+      "loss_token": 1.5881,
+      "lr": "2.00e-07",
+      "norm": 5.9718,
+      "step": 968
+    },
+    {
+      "acc_char": 0.31,
+      "acc_token": 0.7031,
+      "epoch": 0.02,
+      "loss_char": 0.6727,
+      "loss_token": 1.5546,
+      "lr": "2.00e-07",
+      "norm": 5.628,
+      "step": 969
+    },
+    {
+      "acc_char": 0.308,
+      "acc_token": 0.6854,
+      "epoch": 0.02,
+      "loss_char": 0.7318,
+      "loss_token": 1.6251,
+      "lr": "2.00e-07",
+      "norm": 5.7469,
+      "step": 970
+    },
+    {
+      "acc_char": 0.3036,
+      "acc_token": 0.6873,
+      "epoch": 0.02,
+      "loss_char": 0.7109,
+      "loss_token": 1.6241,
+      "lr": "2.00e-07",
+      "norm": 5.4993,
+      "step": 971
+    },
+    {
+      "acc_char": 0.2966,
+      "acc_token": 0.6401,
+      "epoch": 0.02,
+      "loss_char": 0.8435,
+      "loss_token": 1.7002,
+      "lr": "2.00e-07",
+      "norm": 5.4046,
+      "step": 972
+    },
+    {
+      "acc_char": 0.2896,
+      "acc_token": 0.6557,
+      "epoch": 0.02,
+      "loss_char": 0.7916,
+      "loss_token": 1.6553,
+      "lr": "2.00e-07",
+      "norm": 5.9539,
+      "step": 973
+    },
+    {
+      "acc_char": 0.3168,
+      "acc_token": 0.7225,
+      "epoch": 0.02,
+      "loss_char": 0.6251,
+      "loss_token": 1.5137,
+      "lr": "2.00e-07",
+      "norm": 5.4914,
+      "step": 974
+    },
+    {
+      "acc_char": 0.3009,
+      "acc_token": 0.6696,
+      "epoch": 0.02,
+      "loss_char": 0.7631,
+      "loss_token": 1.597,
+      "lr": "2.00e-07",
+      "norm": 5.8429,
+      "step": 975
+    },
+    {
+      "acc_char": 0.3112,
+      "acc_token": 0.7,
+      "epoch": 0.02,
+      "loss_char": 0.6922,
+      "loss_token": 1.5982,
+      "lr": "2.00e-07",
+      "norm": 5.5825,
+      "step": 976
+    },
+    {
+      "acc_char": 0.2958,
+      "acc_token": 0.6691,
+      "epoch": 0.02,
+      "loss_char": 0.7646,
+      "loss_token": 1.6608,
+      "lr": "2.00e-07",
+      "norm": 5.9167,
+      "step": 977
+    },
+    {
+      "acc_char": 0.2957,
+      "acc_token": 0.6676,
+      "epoch": 0.02,
+      "loss_char": 0.7549,
+      "loss_token": 1.6464,
+      "lr": "2.00e-07",
+      "norm": 5.6152,
+      "step": 978
+    },
+    {
+      "acc_char": 0.2952,
+      "acc_token": 0.6787,
+      "epoch": 0.02,
+      "loss_char": 0.7135,
+      "loss_token": 1.5947,
+      "lr": "2.00e-07",
+      "norm": 5.8565,
+      "step": 979
+    },
+    {
+      "acc_char": 0.2982,
+      "acc_token": 0.6622,
+      "epoch": 0.02,
+      "loss_char": 0.7917,
+      "loss_token": 1.61,
+      "lr": "2.00e-07",
+      "norm": 5.9463,
+      "step": 980
+    },
+    {
+      "acc_char": 0.3189,
+      "acc_token": 0.708,
+      "epoch": 0.02,
+      "loss_char": 0.6644,
+      "loss_token": 1.5372,
+      "lr": "2.00e-07",
+      "norm": 5.4684,
+      "step": 981
+    },
+    {
+      "acc_char": 0.3102,
+      "acc_token": 0.7074,
+      "epoch": 0.02,
+      "loss_char": 0.658,
+      "loss_token": 1.5625,
+      "lr": "2.00e-07",
+      "norm": 5.8835,
+      "step": 982
+    },
+    {
+      "acc_char": 0.306,
+      "acc_token": 0.6826,
+      "epoch": 0.02,
+      "loss_char": 0.7437,
+      "loss_token": 1.5536,
+      "lr": "2.00e-07",
+      "norm": 5.7835,
+      "step": 983
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6837,
+      "epoch": 0.02,
+      "loss_char": 0.7336,
+      "loss_token": 1.6808,
+      "lr": "2.00e-07",
+      "norm": 6.0116,
+      "step": 984
+    },
+    {
+      "acc_char": 0.3081,
+      "acc_token": 0.7027,
+      "epoch": 0.02,
+      "loss_char": 0.6454,
+      "loss_token": 1.4823,
+      "lr": "2.00e-07",
+      "norm": 5.621,
+      "step": 985
+    },
+    {
+      "acc_char": 0.3,
+      "acc_token": 0.674,
+      "epoch": 0.02,
+      "loss_char": 0.7582,
+      "loss_token": 1.7042,
+      "lr": "2.00e-07",
+      "norm": 5.4723,
+      "step": 986
+    },
+    {
+      "acc_char": 0.3021,
+      "acc_token": 0.6848,
+      "epoch": 0.02,
+      "loss_char": 0.721,
+      "loss_token": 1.5896,
+      "lr": "2.00e-07",
+      "norm": 5.5576,
+      "step": 987
+    },
+    {
+      "acc_char": 0.2973,
+      "acc_token": 0.675,
+      "epoch": 0.02,
+      "loss_char": 0.7316,
+      "loss_token": 1.6535,
+      "lr": "2.00e-07",
+      "norm": 6.3758,
+      "step": 988
+    },
+    {
+      "acc_char": 0.2932,
+      "acc_token": 0.6709,
+      "epoch": 0.02,
+      "loss_char": 0.7234,
+      "loss_token": 1.6849,
+      "lr": "2.00e-07",
+      "norm": 6.0452,
+      "step": 989
+    },
+    {
+      "acc_char": 0.3049,
+      "acc_token": 0.6874,
+      "epoch": 0.02,
+      "loss_char": 0.7124,
+      "loss_token": 1.6068,
+      "lr": "2.00e-07",
+      "norm": 5.6446,
+      "step": 990
+    },
+    {
+      "acc_char": 0.3056,
+      "acc_token": 0.6794,
+      "epoch": 0.02,
+      "loss_char": 0.7833,
+      "loss_token": 1.6207,
+      "lr": "2.00e-07",
+      "norm": 5.8083,
+      "step": 991
+    },
+    {
+      "acc_char": 0.3044,
+      "acc_token": 0.7026,
+      "epoch": 0.02,
+      "loss_char": 0.6677,
+      "loss_token": 1.6228,
+      "lr": "2.00e-07",
+      "norm": 5.8483,
+      "step": 992
+    },
+    {
+      "acc_char": 0.2968,
+      "acc_token": 0.6706,
+      "epoch": 0.02,
+      "loss_char": 0.7411,
+      "loss_token": 1.6051,
+      "lr": "2.00e-07",
+      "norm": 5.6637,
+      "step": 993
+    },
+    {
+      "acc_char": 0.2939,
+      "acc_token": 0.6371,
+      "epoch": 0.02,
+      "loss_char": 0.8373,
+      "loss_token": 1.6046,
+      "lr": "2.00e-07",
+      "norm": 5.2617,
+      "step": 994
+    },
+    {
+      "acc_char": 0.2967,
+      "acc_token": 0.6698,
+      "epoch": 0.02,
+      "loss_char": 0.7605,
+      "loss_token": 1.7639,
+      "lr": "2.00e-07",
+      "norm": 6.172,
+      "step": 995
+    },
+    {
+      "acc_char": 0.3038,
+      "acc_token": 0.6939,
+      "epoch": 0.02,
+      "loss_char": 0.665,
+      "loss_token": 1.5734,
+      "lr": "2.00e-07",
+      "norm": 5.5388,
+      "step": 996
+    },
+    {
+      "acc_char": 0.2934,
+      "acc_token": 0.6633,
+      "epoch": 0.02,
+      "loss_char": 0.7859,
+      "loss_token": 1.7921,
+      "lr": "2.00e-07",
+      "norm": 5.8554,
+      "step": 997
+    },
+    {
+      "acc_char": 0.309,
+      "acc_token": 0.6991,
+      "epoch": 0.02,
+      "loss_char": 0.6785,
+      "loss_token": 1.5595,
+      "lr": "2.00e-07",
+      "norm": 5.6781,
+      "step": 998
+    },
+    {
+      "acc_char": 0.2965,
+      "acc_token": 0.6636,
+      "epoch": 0.02,
+      "loss_char": 0.796,
+      "loss_token": 1.8083,
+      "lr": "2.00e-07",
+      "norm": 6.3303,
+      "step": 999
+    },
+    {
+      "acc_char": 0.2988,
+      "acc_token": 0.6845,
+      "epoch": 0.02,
+      "loss_char": 0.6922,
+      "loss_token": 1.5965,
+      "lr": "2.00e-07",
+      "norm": 5.5313,
+      "step": 1000
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6778,
+      "epoch": 0.02,
+      "loss_char": 0.7077,
+      "loss_token": 1.6731,
+      "lr": "2.00e-07",
+      "norm": 5.7537,
+      "step": 1001
+    },
+    {
+      "acc_char": 0.3086,
+      "acc_token": 0.6934,
+      "epoch": 0.02,
+      "loss_char": 0.7039,
+      "loss_token": 1.6362,
+      "lr": "2.00e-07",
+      "norm": 5.7014,
+      "step": 1002
+    },
+    {
+      "acc_char": 0.3148,
+      "acc_token": 0.6991,
+      "epoch": 0.02,
+      "loss_char": 0.6969,
+      "loss_token": 1.5178,
+      "lr": "2.00e-07",
+      "norm": 5.6035,
+      "step": 1003
+    },
+    {
+      "acc_char": 0.3199,
+      "acc_token": 0.7144,
+      "epoch": 0.02,
+      "loss_char": 0.648,
+      "loss_token": 1.5337,
+      "lr": "2.00e-07",
+      "norm": 5.6349,
+      "step": 1004
+    },
+    {
+      "acc_char": 0.2994,
+      "acc_token": 0.6819,
+      "epoch": 0.02,
+      "loss_char": 0.7368,
+      "loss_token": 1.6663,
+      "lr": "2.00e-07",
+      "norm": 6.0038,
+      "step": 1005
+    },
+    {
+      "acc_char": 0.3102,
+      "acc_token": 0.7065,
+      "epoch": 0.02,
+      "loss_char": 0.6671,
+      "loss_token": 1.4836,
+      "lr": "2.00e-07",
+      "norm": 5.5694,
+      "step": 1006
+    },
+    {
+      "acc_char": 0.3098,
+      "acc_token": 0.7015,
+      "epoch": 0.02,
+      "loss_char": 0.6681,
+      "loss_token": 1.5047,
+      "lr": "2.00e-07",
+      "norm": 5.6658,
+      "step": 1007
+    },
+    {
+      "acc_char": 0.3133,
+      "acc_token": 0.696,
+      "epoch": 0.02,
+      "loss_char": 0.7178,
+      "loss_token": 1.5725,
+      "lr": "2.00e-07",
+      "norm": 5.6305,
+      "step": 1008
+    },
+    {
+      "acc_char": 0.3118,
+      "acc_token": 0.7075,
+      "epoch": 0.02,
+      "loss_char": 0.6562,
+      "loss_token": 1.6071,
+      "lr": "2.00e-07",
+      "norm": 5.8318,
+      "step": 1009
+    },
+    {
+      "acc_char": 0.299,
+      "acc_token": 0.6798,
+      "epoch": 0.02,
+      "loss_char": 0.7177,
+      "loss_token": 1.6163,
+      "lr": "2.00e-07",
+      "norm": 6.018,
+      "step": 1010
+    },
+    {
+      "acc_char": 0.3028,
+      "acc_token": 0.6844,
+      "epoch": 0.02,
+      "loss_char": 0.7275,
+      "loss_token": 1.6145,
+      "lr": "2.00e-07",
+      "norm": 5.5871,
+      "step": 1011
+    },
+    {
+      "acc_char": 0.307,
+      "acc_token": 0.6897,
+      "epoch": 0.02,
+      "loss_char": 0.692,
+      "loss_token": 1.6421,
+      "lr": "2.00e-07",
+      "norm": 5.5714,
+      "step": 1012
+    },
+    {
+      "acc_char": 0.3302,
+      "acc_token": 0.7278,
+      "epoch": 0.02,
+      "loss_char": 0.6332,
+      "loss_token": 1.4216,
+      "lr": "2.00e-07",
+      "norm": 5.5367,
+      "step": 1013
+    },
+    {
+      "acc_char": 0.301,
+      "acc_token": 0.6804,
+      "epoch": 0.02,
+      "loss_char": 0.715,
+      "loss_token": 1.5587,
+      "lr": "2.00e-07",
+      "norm": 5.5248,
+      "step": 1014
+    },
+    {
+      "acc_char": 0.312,
+      "acc_token": 0.7007,
+      "epoch": 0.02,
+      "loss_char": 0.6859,
+      "loss_token": 1.5663,
+      "lr": "2.00e-07",
+      "norm": 5.7662,
+      "step": 1015
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6801,
+      "epoch": 0.02,
+      "loss_char": 0.7289,
+      "loss_token": 1.7149,
+      "lr": "2.00e-07",
+      "norm": 6.2105,
+      "step": 1016
+    },
+    {
+      "acc_char": 0.2963,
+      "acc_token": 0.6727,
+      "epoch": 0.02,
+      "loss_char": 0.7366,
+      "loss_token": 1.6419,
+      "lr": "2.00e-07",
+      "norm": 5.5471,
+      "step": 1017
+    },
+    {
+      "acc_char": 0.2947,
+      "acc_token": 0.6715,
+      "epoch": 0.02,
+      "loss_char": 0.7429,
+      "loss_token": 1.6365,
+      "lr": "2.00e-07",
+      "norm": 5.8369,
+      "step": 1018
+    },
+    {
+      "acc_char": 0.2834,
+      "acc_token": 0.6051,
+      "epoch": 0.02,
+      "loss_char": 0.9581,
+      "loss_token": 1.7791,
+      "lr": "2.00e-07",
+      "norm": 5.8307,
+      "step": 1019
+    },
+    {
+      "acc_char": 0.3075,
+      "acc_token": 0.7017,
+      "epoch": 0.02,
+      "loss_char": 0.6708,
+      "loss_token": 1.6284,
+      "lr": "2.00e-07",
+      "norm": 5.6093,
+      "step": 1020
+    },
+    {
+      "acc_char": 0.2883,
+      "acc_token": 0.6585,
+      "epoch": 0.02,
+      "loss_char": 0.7865,
+      "loss_token": 1.6614,
+      "lr": "2.00e-07",
+      "norm": 5.5822,
+      "step": 1021
+    },
+    {
+      "acc_char": 0.3119,
+      "acc_token": 0.7002,
+      "epoch": 0.02,
+      "loss_char": 0.6822,
+      "loss_token": 1.522,
+      "lr": "2.00e-07",
+      "norm": 5.5264,
+      "step": 1022
+    },
+    {
+      "acc_char": 0.3014,
+      "acc_token": 0.6818,
+      "epoch": 0.02,
+      "loss_char": 0.7099,
+      "loss_token": 1.6397,
+      "lr": "2.00e-07",
+      "norm": 6.0578,
+      "step": 1023
+    },
+    {
+      "acc_char": 0.2979,
+      "acc_token": 0.6682,
+      "epoch": 0.02,
+      "loss_char": 0.7661,
+      "loss_token": 1.5846,
+      "lr": "2.00e-07",
+      "norm": 5.6649,
+      "step": 1024
+    },
+    {
+      "acc_char": 0.3115,
+      "acc_token": 0.7107,
+      "epoch": 0.02,
+      "loss_char": 0.6384,
+      "loss_token": 1.4569,
+      "lr": "2.00e-07",
+      "norm": 5.8935,
+      "step": 1025
+    },
+    {
+      "acc_char": 0.3036,
+      "acc_token": 0.6913,
+      "epoch": 0.02,
+      "loss_char": 0.7082,
+      "loss_token": 1.6987,
+      "lr": "2.00e-07",
+      "norm": 5.8197,
+      "step": 1026
+    },
+    {
+      "acc_char": 0.3048,
+      "acc_token": 0.701,
+      "epoch": 0.02,
+      "loss_char": 0.6587,
+      "loss_token": 1.5762,
+      "lr": "2.00e-07",
+      "norm": 5.9085,
+      "step": 1027
+    },
+    {
+      "acc_char": 0.2939,
+      "acc_token": 0.664,
+      "epoch": 0.02,
+      "loss_char": 0.7915,
+      "loss_token": 1.6195,
+      "lr": "2.00e-07",
+      "norm": 6.1612,
+      "step": 1028
+    },
+    {
+      "acc_char": 0.2942,
+      "acc_token": 0.6608,
+      "epoch": 0.02,
+      "loss_char": 0.795,
+      "loss_token": 1.7062,
+      "lr": "2.00e-07",
+      "norm": 5.9826,
+      "step": 1029
+    },
+    {
+      "acc_char": 0.302,
+      "acc_token": 0.6942,
+      "epoch": 0.02,
+      "loss_char": 0.679,
+      "loss_token": 1.5238,
+      "lr": "2.00e-07",
+      "norm": 5.5649,
+      "step": 1030
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6598,
+      "epoch": 0.02,
+      "loss_char": 0.7892,
+      "loss_token": 1.6886,
+      "lr": "2.00e-07",
+      "norm": 5.8533,
+      "step": 1031
+    },
+    {
+      "acc_char": 0.3043,
+      "acc_token": 0.6845,
+      "epoch": 0.02,
+      "loss_char": 0.7398,
+      "loss_token": 1.6375,
+      "lr": "2.00e-07",
+      "norm": 5.9787,
+      "step": 1032
+    },
+    {
+      "acc_char": 0.3008,
+      "acc_token": 0.6733,
+      "epoch": 0.02,
+      "loss_char": 0.7634,
+      "loss_token": 1.7155,
+      "lr": "2.00e-07",
+      "norm": 6.2085,
+      "step": 1033
+    },
+    {
+      "acc_char": 0.3073,
+      "acc_token": 0.7069,
+      "epoch": 0.02,
+      "loss_char": 0.646,
+      "loss_token": 1.5893,
+      "lr": "2.00e-07",
+      "norm": 5.4341,
+      "step": 1034
+    },
+    {
+      "acc_char": 0.2876,
+      "acc_token": 0.6591,
+      "epoch": 0.02,
+      "loss_char": 0.7724,
+      "loss_token": 1.714,
+      "lr": "2.00e-07",
+      "norm": 6.0691,
+      "step": 1035
+    },
+    {
+      "acc_char": 0.3212,
+      "acc_token": 0.694,
+      "epoch": 0.02,
+      "loss_char": 0.7451,
+      "loss_token": 1.6876,
+      "lr": "2.00e-07",
+      "norm": 5.8539,
+      "step": 1036
+    },
+    {
+      "acc_char": 0.3072,
+      "acc_token": 0.6944,
+      "epoch": 0.02,
+      "loss_char": 0.694,
+      "loss_token": 1.644,
+      "lr": "2.00e-07",
+      "norm": 5.9415,
+      "step": 1037
+    },
+    {
+      "acc_char": 0.3083,
+      "acc_token": 0.704,
+      "epoch": 0.02,
+      "loss_char": 0.663,
+      "loss_token": 1.5913,
+      "lr": "2.00e-07",
+      "norm": 5.752,
+      "step": 1038
+    },
+    {
+      "acc_char": 0.3034,
+      "acc_token": 0.6902,
+      "epoch": 0.02,
+      "loss_char": 0.6879,
+      "loss_token": 1.5627,
+      "lr": "2.00e-07",
+      "norm": 5.4879,
+      "step": 1039
+    },
+    {
+      "acc_char": 0.3045,
+      "acc_token": 0.693,
+      "epoch": 0.02,
+      "loss_char": 0.7045,
+      "loss_token": 1.6542,
+      "lr": "2.00e-07",
+      "norm": 5.9541,
+      "step": 1040
+    },
+    {
+      "acc_char": 0.2985,
+      "acc_token": 0.6826,
+      "epoch": 0.02,
+      "loss_char": 0.7156,
+      "loss_token": 1.6487,
+      "lr": "2.00e-07",
+      "norm": 5.9876,
+      "step": 1041
+    },
+    {
+      "acc_char": 0.2876,
+      "acc_token": 0.6663,
+      "epoch": 0.02,
+      "loss_char": 0.7454,
+      "loss_token": 1.7047,
+      "lr": "2.00e-07",
+      "norm": 5.9158,
+      "step": 1042
+    },
+    {
+      "acc_char": 0.312,
+      "acc_token": 0.6979,
+      "epoch": 0.02,
+      "loss_char": 0.7015,
+      "loss_token": 1.6362,
+      "lr": "2.00e-07",
+      "norm": 5.7544,
+      "step": 1043
+    },
+    {
+      "acc_char": 0.317,
+      "acc_token": 0.7015,
+      "epoch": 0.02,
+      "loss_char": 0.6757,
+      "loss_token": 1.5832,
+      "lr": "2.00e-07",
+      "norm": 5.8245,
+      "step": 1044
+    },
+    {
+      "acc_char": 0.2898,
+      "acc_token": 0.6603,
+      "epoch": 0.02,
+      "loss_char": 0.7839,
+      "loss_token": 1.6956,
+      "lr": "2.00e-07",
+      "norm": 5.586,
+      "step": 1045
+    },
+    {
+      "acc_char": 0.3107,
+      "acc_token": 0.7142,
+      "epoch": 0.02,
+      "loss_char": 0.6085,
+      "loss_token": 1.5081,
+      "lr": "2.00e-07",
+      "norm": 5.6408,
+      "step": 1046
+    },
+    {
+      "acc_char": 0.307,
+      "acc_token": 0.6913,
+      "epoch": 0.02,
+      "loss_char": 0.719,
+      "loss_token": 1.6445,
+      "lr": "2.00e-07",
+      "norm": 6.0086,
+      "step": 1047
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6932,
+      "epoch": 0.02,
+      "loss_char": 0.6583,
+      "loss_token": 1.59,
+      "lr": "2.00e-07",
+      "norm": 5.6165,
+      "step": 1048
+    },
+    {
+      "acc_char": 0.2898,
+      "acc_token": 0.6699,
+      "epoch": 0.02,
+      "loss_char": 0.7697,
+      "loss_token": 1.6934,
+      "lr": "2.00e-07",
+      "norm": 5.8678,
+      "step": 1049
+    },
+    {
+      "acc_char": 0.2896,
+      "acc_token": 0.6657,
+      "epoch": 0.02,
+      "loss_char": 0.726,
+      "loss_token": 1.6075,
+      "lr": "2.00e-07",
+      "norm": 5.7553,
+      "step": 1050
+    },
+    {
+      "acc_char": 0.3099,
+      "acc_token": 0.6864,
+      "epoch": 0.02,
+      "loss_char": 0.7305,
+      "loss_token": 1.5594,
+      "lr": "2.00e-07",
+      "norm": 5.5492,
+      "step": 1051
+    },
+    {
+      "acc_char": 0.308,
+      "acc_token": 0.698,
+      "epoch": 0.02,
+      "loss_char": 0.6745,
+      "loss_token": 1.6243,
+      "lr": "2.00e-07",
+      "norm": 5.6144,
+      "step": 1052
+    },
+    {
+      "acc_char": 0.3049,
+      "acc_token": 0.6837,
+      "epoch": 0.02,
+      "loss_char": 0.7032,
+      "loss_token": 1.5993,
+      "lr": "2.00e-07",
+      "norm": 5.7339,
+      "step": 1053
+    },
+    {
+      "acc_char": 0.2919,
+      "acc_token": 0.6712,
+      "epoch": 0.02,
+      "loss_char": 0.7139,
+      "loss_token": 1.6253,
+      "lr": "2.00e-07",
+      "norm": 5.5032,
+      "step": 1054
+    },
+    {
+      "acc_char": 0.3036,
+      "acc_token": 0.6947,
+      "epoch": 0.02,
+      "loss_char": 0.6816,
+      "loss_token": 1.5841,
+      "lr": "2.00e-07",
+      "norm": 5.6046,
+      "step": 1055
+    },
+    {
+      "acc_char": 0.3039,
+      "acc_token": 0.6752,
+      "epoch": 0.02,
+      "loss_char": 0.7526,
+      "loss_token": 1.5791,
+      "lr": "2.00e-07",
+      "norm": 5.6113,
+      "step": 1056
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.6852,
+      "epoch": 0.02,
+      "loss_char": 0.6962,
+      "loss_token": 1.5514,
+      "lr": "2.00e-07",
+      "norm": 5.402,
+      "step": 1057
+    },
+    {
+      "acc_char": 0.2951,
+      "acc_token": 0.6855,
+      "epoch": 0.02,
+      "loss_char": 0.6932,
+      "loss_token": 1.6079,
+      "lr": "2.00e-07",
+      "norm": 5.5452,
+      "step": 1058
+    },
+    {
+      "acc_char": 0.3009,
+      "acc_token": 0.6767,
+      "epoch": 0.02,
+      "loss_char": 0.7406,
+      "loss_token": 1.654,
+      "lr": "2.00e-07",
+      "norm": 5.6495,
+      "step": 1059
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.6819,
+      "epoch": 0.02,
+      "loss_char": 0.7202,
+      "loss_token": 1.6393,
+      "lr": "2.00e-07",
+      "norm": 5.7668,
+      "step": 1060
+    },
+    {
+      "acc_char": 0.3053,
+      "acc_token": 0.6853,
+      "epoch": 0.02,
+      "loss_char": 0.7316,
+      "loss_token": 1.6124,
+      "lr": "2.00e-07",
+      "norm": 5.7434,
+      "step": 1061
+    },
+    {
+      "acc_char": 0.3086,
+      "acc_token": 0.6933,
+      "epoch": 0.02,
+      "loss_char": 0.7046,
+      "loss_token": 1.5938,
+      "lr": "2.00e-07",
+      "norm": 5.7935,
+      "step": 1062
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.6864,
+      "epoch": 0.02,
+      "loss_char": 0.6721,
+      "loss_token": 1.5034,
+      "lr": "2.00e-07",
+      "norm": 5.5376,
+      "step": 1063
+    },
+    {
+      "acc_char": 0.3153,
+      "acc_token": 0.6985,
+      "epoch": 0.02,
+      "loss_char": 0.7039,
+      "loss_token": 1.629,
+      "lr": "2.00e-07",
+      "norm": 5.7942,
+      "step": 1064
+    },
+    {
+      "acc_char": 0.3056,
+      "acc_token": 0.6922,
+      "epoch": 0.02,
+      "loss_char": 0.6849,
+      "loss_token": 1.5634,
+      "lr": "2.00e-07",
+      "norm": 5.574,
+      "step": 1065
+    },
+    {
+      "acc_char": 0.3042,
+      "acc_token": 0.6885,
+      "epoch": 0.02,
+      "loss_char": 0.7123,
+      "loss_token": 1.6346,
+      "lr": "2.00e-07",
+      "norm": 5.748,
+      "step": 1066
+    },
+    {
+      "acc_char": 0.3022,
+      "acc_token": 0.6913,
+      "epoch": 0.02,
+      "loss_char": 0.6892,
+      "loss_token": 1.6414,
+      "lr": "2.00e-07",
+      "norm": 5.6367,
+      "step": 1067
+    },
+    {
+      "acc_char": 0.3201,
+      "acc_token": 0.7164,
+      "epoch": 0.02,
+      "loss_char": 0.6417,
+      "loss_token": 1.4198,
+      "lr": "2.00e-07",
+      "norm": 5.763,
+      "step": 1068
+    },
+    {
+      "acc_char": 0.3084,
+      "acc_token": 0.7033,
+      "epoch": 0.02,
+      "loss_char": 0.6387,
+      "loss_token": 1.4254,
+      "lr": "2.00e-07",
+      "norm": 5.5338,
+      "step": 1069
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6755,
+      "epoch": 0.02,
+      "loss_char": 0.7654,
+      "loss_token": 1.66,
+      "lr": "2.00e-07",
+      "norm": 5.725,
+      "step": 1070
+    },
+    {
+      "acc_char": 0.3093,
+      "acc_token": 0.6938,
+      "epoch": 0.02,
+      "loss_char": 0.6978,
+      "loss_token": 1.4869,
+      "lr": "2.00e-07",
+      "norm": 5.6716,
+      "step": 1071
+    },
+    {
+      "acc_char": 0.3168,
+      "acc_token": 0.704,
+      "epoch": 0.02,
+      "loss_char": 0.6935,
+      "loss_token": 1.5235,
+      "lr": "2.00e-07",
+      "norm": 5.7914,
+      "step": 1072
+    },
+    {
+      "acc_char": 0.2949,
+      "acc_token": 0.6807,
+      "epoch": 0.02,
+      "loss_char": 0.7211,
+      "loss_token": 1.4897,
+      "lr": "2.00e-07",
+      "norm": 5.769,
+      "step": 1073
+    },
+    {
+      "acc_char": 0.2976,
+      "acc_token": 0.67,
+      "epoch": 0.02,
+      "loss_char": 0.7538,
+      "loss_token": 1.5103,
+      "lr": "2.00e-07",
+      "norm": 5.6134,
+      "step": 1074
+    },
+    {
+      "acc_char": 0.3044,
+      "acc_token": 0.7082,
+      "epoch": 0.02,
+      "loss_char": 0.6335,
+      "loss_token": 1.5032,
+      "lr": "2.00e-07",
+      "norm": 5.7971,
+      "step": 1075
+    },
+    {
+      "acc_char": 0.3111,
+      "acc_token": 0.6939,
+      "epoch": 0.02,
+      "loss_char": 0.6864,
+      "loss_token": 1.5623,
+      "lr": "2.00e-07",
+      "norm": 6.0235,
+      "step": 1076
+    },
+    {
+      "acc_char": 0.298,
+      "acc_token": 0.677,
+      "epoch": 0.02,
+      "loss_char": 0.7431,
+      "loss_token": 1.6234,
+      "lr": "2.00e-07",
+      "norm": 5.7891,
+      "step": 1077
+    },
+    {
+      "acc_char": 0.2977,
+      "acc_token": 0.6496,
+      "epoch": 0.02,
+      "loss_char": 0.8182,
+      "loss_token": 1.7089,
+      "lr": "2.00e-07",
+      "norm": 5.4175,
+      "step": 1078
+    },
+    {
+      "acc_char": 0.3125,
+      "acc_token": 0.7023,
+      "epoch": 0.02,
+      "loss_char": 0.6682,
+      "loss_token": 1.5609,
+      "lr": "2.00e-07",
+      "norm": 5.61,
+      "step": 1079
+    },
+    {
+      "acc_char": 0.289,
+      "acc_token": 0.629,
+      "epoch": 0.02,
+      "loss_char": 0.8491,
+      "loss_token": 1.7577,
+      "lr": "2.00e-07",
+      "norm": 5.6545,
+      "step": 1080
+    },
+    {
+      "acc_char": 0.2979,
+      "acc_token": 0.6751,
+      "epoch": 0.02,
+      "loss_char": 0.742,
+      "loss_token": 1.6155,
+      "lr": "2.00e-07",
+      "norm": 6.1317,
+      "step": 1081
+    },
+    {
+      "acc_char": 0.2807,
+      "acc_token": 0.6569,
+      "epoch": 0.02,
+      "loss_char": 0.7763,
+      "loss_token": 1.6378,
+      "lr": "2.00e-07",
+      "norm": 5.7644,
+      "step": 1082
+    },
+    {
+      "acc_char": 0.3042,
+      "acc_token": 0.6817,
+      "epoch": 0.02,
+      "loss_char": 0.725,
+      "loss_token": 1.6469,
+      "lr": "2.00e-07",
+      "norm": 6.1361,
+      "step": 1083
+    },
+    {
+      "acc_char": 0.3027,
+      "acc_token": 0.6684,
+      "epoch": 0.02,
+      "loss_char": 0.802,
+      "loss_token": 1.6769,
+      "lr": "2.00e-07",
+      "norm": 5.7604,
+      "step": 1084
+    },
+    {
+      "acc_char": 0.3024,
+      "acc_token": 0.7005,
+      "epoch": 0.02,
+      "loss_char": 0.6753,
+      "loss_token": 1.6225,
+      "lr": "2.00e-07",
+      "norm": 6.0857,
+      "step": 1085
+    },
+    {
+      "acc_char": 0.3024,
+      "acc_token": 0.6782,
+      "epoch": 0.02,
+      "loss_char": 0.7807,
+      "loss_token": 1.6282,
+      "lr": "2.00e-07",
+      "norm": 5.5427,
+      "step": 1086
+    },
+    {
+      "acc_char": 0.3037,
+      "acc_token": 0.6775,
+      "epoch": 0.02,
+      "loss_char": 0.7505,
+      "loss_token": 1.6065,
+      "lr": "2.00e-07",
+      "norm": 5.8749,
+      "step": 1087
+    },
+    {
+      "acc_char": 0.305,
+      "acc_token": 0.6957,
+      "epoch": 0.02,
+      "loss_char": 0.6722,
+      "loss_token": 1.5519,
+      "lr": "2.00e-07",
+      "norm": 6.1559,
+      "step": 1088
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6814,
+      "epoch": 0.02,
+      "loss_char": 0.7442,
+      "loss_token": 1.6725,
+      "lr": "2.00e-07",
+      "norm": 5.7716,
+      "step": 1089
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6795,
+      "epoch": 0.02,
+      "loss_char": 0.735,
+      "loss_token": 1.6029,
+      "lr": "2.00e-07",
+      "norm": 5.6591,
+      "step": 1090
+    },
+    {
+      "acc_char": 0.3149,
+      "acc_token": 0.7071,
+      "epoch": 0.02,
+      "loss_char": 0.6583,
+      "loss_token": 1.5001,
+      "lr": "2.00e-07",
+      "norm": 6.1907,
+      "step": 1091
+    },
+    {
+      "acc_char": 0.3112,
+      "acc_token": 0.6987,
+      "epoch": 0.02,
+      "loss_char": 0.6887,
+      "loss_token": 1.649,
+      "lr": "2.00e-07",
+      "norm": 5.9268,
+      "step": 1092
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.6854,
+      "epoch": 0.02,
+      "loss_char": 0.7085,
+      "loss_token": 1.6117,
+      "lr": "2.00e-07",
+      "norm": 5.5384,
+      "step": 1093
+    },
+    {
+      "acc_char": 0.3178,
+      "acc_token": 0.709,
+      "epoch": 0.02,
+      "loss_char": 0.6628,
+      "loss_token": 1.5315,
+      "lr": "2.00e-07",
+      "norm": 5.8061,
+      "step": 1094
+    },
+    {
+      "acc_char": 0.2962,
+      "acc_token": 0.6681,
+      "epoch": 0.02,
+      "loss_char": 0.793,
+      "loss_token": 1.7314,
+      "lr": "2.00e-07",
+      "norm": 6.2168,
+      "step": 1095
+    },
+    {
+      "acc_char": 0.2913,
+      "acc_token": 0.6629,
+      "epoch": 0.02,
+      "loss_char": 0.799,
+      "loss_token": 1.6578,
+      "lr": "2.00e-07",
+      "norm": 5.7005,
+      "step": 1096
+    },
+    {
+      "acc_char": 0.3066,
+      "acc_token": 0.6833,
+      "epoch": 0.02,
+      "loss_char": 0.7418,
+      "loss_token": 1.6199,
+      "lr": "2.00e-07",
+      "norm": 5.6565,
+      "step": 1097
+    },
+    {
+      "acc_char": 0.3096,
+      "acc_token": 0.6978,
+      "epoch": 0.02,
+      "loss_char": 0.6901,
+      "loss_token": 1.5497,
+      "lr": "2.00e-07",
+      "norm": 5.6842,
+      "step": 1098
+    },
+    {
+      "acc_char": 0.2991,
+      "acc_token": 0.6748,
+      "epoch": 0.02,
+      "loss_char": 0.7516,
+      "loss_token": 1.6768,
+      "lr": "2.00e-07",
+      "norm": 5.9045,
+      "step": 1099
+    },
+    {
+      "acc_char": 0.308,
+      "acc_token": 0.6957,
+      "epoch": 0.02,
+      "loss_char": 0.7002,
+      "loss_token": 1.5442,
+      "lr": "2.00e-07",
+      "norm": 5.5604,
+      "step": 1100
+    },
+    {
+      "acc_char": 0.3073,
+      "acc_token": 0.6942,
+      "epoch": 0.02,
+      "loss_char": 0.6821,
+      "loss_token": 1.6096,
+      "lr": "2.00e-07",
+      "norm": 5.6615,
+      "step": 1101
+    },
+    {
+      "acc_char": 0.2967,
+      "acc_token": 0.676,
+      "epoch": 0.02,
+      "loss_char": 0.7305,
+      "loss_token": 1.6356,
+      "lr": "2.00e-07",
+      "norm": 5.9914,
+      "step": 1102
+    },
+    {
+      "acc_char": 0.3041,
+      "acc_token": 0.6879,
+      "epoch": 0.02,
+      "loss_char": 0.7077,
+      "loss_token": 1.6373,
+      "lr": "2.00e-07",
+      "norm": 5.6571,
+      "step": 1103
+    },
+    {
+      "acc_char": 0.3059,
+      "acc_token": 0.692,
+      "epoch": 0.02,
+      "loss_char": 0.7083,
+      "loss_token": 1.5886,
+      "lr": "2.00e-07",
+      "norm": 5.7888,
+      "step": 1104
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6927,
+      "epoch": 0.02,
+      "loss_char": 0.6797,
+      "loss_token": 1.5879,
+      "lr": "2.00e-07",
+      "norm": 5.5456,
+      "step": 1105
+    },
+    {
+      "acc_char": 0.3108,
+      "acc_token": 0.7049,
+      "epoch": 0.02,
+      "loss_char": 0.6782,
+      "loss_token": 1.6428,
+      "lr": "2.00e-07",
+      "norm": 5.7501,
+      "step": 1106
+    },
+    {
+      "acc_char": 0.3124,
+      "acc_token": 0.7192,
+      "epoch": 0.02,
+      "loss_char": 0.6138,
+      "loss_token": 1.4979,
+      "lr": "2.00e-07",
+      "norm": 5.5096,
+      "step": 1107
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6847,
+      "epoch": 0.02,
+      "loss_char": 0.7125,
+      "loss_token": 1.6176,
+      "lr": "2.00e-07",
+      "norm": 5.9919,
+      "step": 1108
+    },
+    {
+      "acc_char": 0.2978,
+      "acc_token": 0.6398,
+      "epoch": 0.02,
+      "loss_char": 0.8575,
+      "loss_token": 1.8423,
+      "lr": "2.00e-07",
+      "norm": 5.9944,
+      "step": 1109
+    },
+    {
+      "acc_char": 0.336,
+      "acc_token": 0.7513,
+      "epoch": 0.02,
+      "loss_char": 0.5394,
+      "loss_token": 1.4633,
+      "lr": "2.00e-07",
+      "norm": 5.5493,
+      "step": 1110
+    },
+    {
+      "acc_char": 0.2986,
+      "acc_token": 0.6811,
+      "epoch": 0.02,
+      "loss_char": 0.7307,
+      "loss_token": 1.6021,
+      "lr": "2.00e-07",
+      "norm": 5.977,
+      "step": 1111
+    },
+    {
+      "acc_char": 0.3055,
+      "acc_token": 0.6839,
+      "epoch": 0.02,
+      "loss_char": 0.7099,
+      "loss_token": 1.6421,
+      "lr": "2.00e-07",
+      "norm": 6.1366,
+      "step": 1112
+    },
+    {
+      "acc_char": 0.2972,
+      "acc_token": 0.6873,
+      "epoch": 0.02,
+      "loss_char": 0.6962,
+      "loss_token": 1.526,
+      "lr": "2.00e-07",
+      "norm": 5.6157,
+      "step": 1113
+    },
+    {
+      "acc_char": 0.3069,
+      "acc_token": 0.658,
+      "epoch": 0.02,
+      "loss_char": 0.8129,
+      "loss_token": 1.6929,
+      "lr": "2.00e-07",
+      "norm": 5.2494,
+      "step": 1114
+    },
+    {
+      "acc_char": 0.3093,
+      "acc_token": 0.7066,
+      "epoch": 0.02,
+      "loss_char": 0.6525,
+      "loss_token": 1.5155,
+      "lr": "2.00e-07",
+      "norm": 5.5897,
+      "step": 1115
+    },
+    {
+      "acc_char": 0.3035,
+      "acc_token": 0.6805,
+      "epoch": 0.02,
+      "loss_char": 0.778,
+      "loss_token": 1.6953,
+      "lr": "2.00e-07",
+      "norm": 5.9045,
+      "step": 1116
+    },
+    {
+      "acc_char": 0.3072,
+      "acc_token": 0.6812,
+      "epoch": 0.02,
+      "loss_char": 0.7494,
+      "loss_token": 1.6054,
+      "lr": "2.00e-07",
+      "norm": 5.7761,
+      "step": 1117
+    },
+    {
+      "acc_char": 0.3175,
+      "acc_token": 0.7154,
+      "epoch": 0.02,
+      "loss_char": 0.6509,
+      "loss_token": 1.6047,
+      "lr": "2.00e-07",
+      "norm": 5.8353,
+      "step": 1118
+    },
+    {
+      "acc_char": 0.2921,
+      "acc_token": 0.6718,
+      "epoch": 0.02,
+      "loss_char": 0.7321,
+      "loss_token": 1.6241,
+      "lr": "2.00e-07",
+      "norm": 5.8048,
+      "step": 1119
+    },
+    {
+      "acc_char": 0.3005,
+      "acc_token": 0.6899,
+      "epoch": 0.02,
+      "loss_char": 0.6899,
+      "loss_token": 1.6578,
+      "lr": "2.00e-07",
+      "norm": 5.7674,
+      "step": 1120
+    },
+    {
+      "acc_char": 0.2967,
+      "acc_token": 0.6599,
+      "epoch": 0.02,
+      "loss_char": 0.7958,
+      "loss_token": 1.7595,
+      "lr": "2.00e-07",
+      "norm": 5.7997,
+      "step": 1121
+    },
+    {
+      "acc_char": 0.2968,
+      "acc_token": 0.6699,
+      "epoch": 0.02,
+      "loss_char": 0.7646,
+      "loss_token": 1.6413,
+      "lr": "2.00e-07",
+      "norm": 6.2408,
+      "step": 1122
+    },
+    {
+      "acc_char": 0.3,
+      "acc_token": 0.674,
+      "epoch": 0.02,
+      "loss_char": 0.7489,
+      "loss_token": 1.6281,
+      "lr": "2.00e-07",
+      "norm": 6.1318,
+      "step": 1123
+    },
+    {
+      "acc_char": 0.305,
+      "acc_token": 0.6771,
+      "epoch": 0.02,
+      "loss_char": 0.7268,
+      "loss_token": 1.6143,
+      "lr": "2.00e-07",
+      "norm": 5.8492,
+      "step": 1124
+    },
+    {
+      "acc_char": 0.3118,
+      "acc_token": 0.7104,
+      "epoch": 0.02,
+      "loss_char": 0.6392,
+      "loss_token": 1.4847,
+      "lr": "2.00e-07",
+      "norm": 5.6128,
+      "step": 1125
+    },
+    {
+      "acc_char": 0.3019,
+      "acc_token": 0.6886,
+      "epoch": 0.02,
+      "loss_char": 0.695,
+      "loss_token": 1.6002,
+      "lr": "2.00e-07",
+      "norm": 5.6202,
+      "step": 1126
+    },
+    {
+      "acc_char": 0.2998,
+      "acc_token": 0.6811,
+      "epoch": 0.02,
+      "loss_char": 0.7349,
+      "loss_token": 1.6541,
+      "lr": "2.00e-07",
+      "norm": 5.9175,
+      "step": 1127
+    },
+    {
+      "acc_char": 0.3002,
+      "acc_token": 0.6721,
+      "epoch": 0.02,
+      "loss_char": 0.8037,
+      "loss_token": 1.6463,
+      "lr": "2.00e-07",
+      "norm": 5.9228,
+      "step": 1128
+    },
+    {
+      "acc_char": 0.2939,
+      "acc_token": 0.6658,
+      "epoch": 0.02,
+      "loss_char": 0.7702,
+      "loss_token": 1.6422,
+      "lr": "2.00e-07",
+      "norm": 5.8841,
+      "step": 1129
+    },
+    {
+      "acc_char": 0.3095,
+      "acc_token": 0.7018,
+      "epoch": 0.02,
+      "loss_char": 0.6496,
+      "loss_token": 1.5094,
+      "lr": "2.00e-07",
+      "norm": 5.4645,
+      "step": 1130
+    },
+    {
+      "acc_char": 0.2889,
+      "acc_token": 0.6654,
+      "epoch": 0.02,
+      "loss_char": 0.7496,
+      "loss_token": 1.6055,
+      "lr": "2.00e-07",
+      "norm": 5.7124,
+      "step": 1131
+    },
+    {
+      "acc_char": 0.303,
+      "acc_token": 0.6727,
+      "epoch": 0.02,
+      "loss_char": 0.748,
+      "loss_token": 1.6193,
+      "lr": "2.00e-07",
+      "norm": 5.7687,
+      "step": 1132
+    },
+    {
+      "acc_char": 0.2895,
+      "acc_token": 0.6682,
+      "epoch": 0.02,
+      "loss_char": 0.7351,
+      "loss_token": 1.7073,
+      "lr": "2.00e-07",
+      "norm": 5.9363,
+      "step": 1133
+    },
+    {
+      "acc_char": 0.3023,
+      "acc_token": 0.6841,
+      "epoch": 0.02,
+      "loss_char": 0.7325,
+      "loss_token": 1.68,
+      "lr": "2.00e-07",
+      "norm": 5.6545,
+      "step": 1134
+    },
+    {
+      "acc_char": 0.3073,
+      "acc_token": 0.6845,
+      "epoch": 0.02,
+      "loss_char": 0.7218,
+      "loss_token": 1.6811,
+      "lr": "2.00e-07",
+      "norm": 5.7013,
+      "step": 1135
+    },
+    {
+      "acc_char": 0.3153,
+      "acc_token": 0.7098,
+      "epoch": 0.02,
+      "loss_char": 0.6496,
+      "loss_token": 1.6311,
+      "lr": "2.00e-07",
+      "norm": 5.502,
+      "step": 1136
+    },
+    {
+      "acc_char": 0.3176,
+      "acc_token": 0.7143,
+      "epoch": 0.02,
+      "loss_char": 0.6434,
+      "loss_token": 1.4789,
+      "lr": "2.00e-07",
+      "norm": 5.409,
+      "step": 1137
+    },
+    {
+      "acc_char": 0.2948,
+      "acc_token": 0.6224,
+      "epoch": 0.02,
+      "loss_char": 0.9341,
+      "loss_token": 1.8052,
+      "lr": "2.00e-07",
+      "norm": 5.7219,
+      "step": 1138
+    },
+    {
+      "acc_char": 0.314,
+      "acc_token": 0.7197,
+      "epoch": 0.02,
+      "loss_char": 0.6184,
+      "loss_token": 1.4714,
+      "lr": "2.00e-07",
+      "norm": 5.2736,
+      "step": 1139
+    },
+    {
+      "acc_char": 0.3037,
+      "acc_token": 0.6837,
+      "epoch": 0.02,
+      "loss_char": 0.7257,
+      "loss_token": 1.6964,
+      "lr": "2.00e-07",
+      "norm": 5.7293,
+      "step": 1140
+    },
+    {
+      "acc_char": 0.2996,
+      "acc_token": 0.6801,
+      "epoch": 0.02,
+      "loss_char": 0.7177,
+      "loss_token": 1.6569,
+      "lr": "2.00e-07",
+      "norm": 5.8446,
+      "step": 1141
+    },
+    {
+      "acc_char": 0.3144,
+      "acc_token": 0.7021,
+      "epoch": 0.02,
+      "loss_char": 0.6756,
+      "loss_token": 1.5498,
+      "lr": "2.00e-07",
+      "norm": 5.4456,
+      "step": 1142
+    },
+    {
+      "acc_char": 0.3091,
+      "acc_token": 0.7007,
+      "epoch": 0.02,
+      "loss_char": 0.6462,
+      "loss_token": 1.5228,
+      "lr": "2.00e-07",
+      "norm": 5.4793,
+      "step": 1143
+    },
+    {
+      "acc_char": 0.3141,
+      "acc_token": 0.7055,
+      "epoch": 0.02,
+      "loss_char": 0.6789,
+      "loss_token": 1.5552,
+      "lr": "2.00e-07",
+      "norm": 5.6754,
+      "step": 1144
+    },
+    {
+      "acc_char": 0.3149,
+      "acc_token": 0.7041,
+      "epoch": 0.02,
+      "loss_char": 0.677,
+      "loss_token": 1.6774,
+      "lr": "2.00e-07",
+      "norm": 5.6087,
+      "step": 1145
+    },
+    {
+      "acc_char": 0.2854,
+      "acc_token": 0.6525,
+      "epoch": 0.02,
+      "loss_char": 0.8141,
+      "loss_token": 1.7267,
+      "lr": "2.00e-07",
+      "norm": 6.0019,
+      "step": 1146
+    },
+    {
+      "acc_char": 0.3099,
+      "acc_token": 0.7048,
+      "epoch": 0.02,
+      "loss_char": 0.6781,
+      "loss_token": 1.5686,
+      "lr": "2.00e-07",
+      "norm": 5.6791,
+      "step": 1147
+    },
+    {
+      "acc_char": 0.296,
+      "acc_token": 0.675,
+      "epoch": 0.02,
+      "loss_char": 0.7463,
+      "loss_token": 1.679,
+      "lr": "2.00e-07",
+      "norm": 5.5865,
+      "step": 1148
+    },
+    {
+      "acc_char": 0.3138,
+      "acc_token": 0.7211,
+      "epoch": 0.02,
+      "loss_char": 0.6145,
+      "loss_token": 1.5026,
+      "lr": "2.00e-07",
+      "norm": 5.7165,
+      "step": 1149
+    },
+    {
+      "acc_char": 0.3025,
+      "acc_token": 0.6962,
+      "epoch": 0.02,
+      "loss_char": 0.657,
+      "loss_token": 1.5813,
+      "lr": "2.00e-07",
+      "norm": 5.806,
+      "step": 1150
+    },
+    {
+      "acc_char": 0.3103,
+      "acc_token": 0.7005,
+      "epoch": 0.02,
+      "loss_char": 0.6963,
+      "loss_token": 1.6262,
+      "lr": "2.00e-07",
+      "norm": 5.6984,
+      "step": 1151
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6378,
+      "epoch": 0.02,
+      "loss_char": 0.8329,
+      "loss_token": 1.7329,
+      "lr": "2.00e-07",
+      "norm": 5.4393,
+      "step": 1152
+    },
+    {
+      "acc_char": 0.2946,
+      "acc_token": 0.6716,
+      "epoch": 0.02,
+      "loss_char": 0.7424,
+      "loss_token": 1.5892,
+      "lr": "2.00e-07",
+      "norm": 6.0617,
+      "step": 1153
+    },
+    {
+      "acc_char": 0.3046,
+      "acc_token": 0.6855,
+      "epoch": 0.02,
+      "loss_char": 0.7098,
+      "loss_token": 1.6979,
+      "lr": "2.00e-07",
+      "norm": 5.6377,
+      "step": 1154
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6955,
+      "epoch": 0.02,
+      "loss_char": 0.6606,
+      "loss_token": 1.5007,
+      "lr": "2.00e-07",
+      "norm": 5.4087,
+      "step": 1155
+    },
+    {
+      "acc_char": 0.2943,
+      "acc_token": 0.6602,
+      "epoch": 0.02,
+      "loss_char": 0.7823,
+      "loss_token": 1.6413,
+      "lr": "2.00e-07",
+      "norm": 5.8073,
+      "step": 1156
+    },
+    {
+      "acc_char": 0.2912,
+      "acc_token": 0.698,
+      "epoch": 0.02,
+      "loss_char": 0.6182,
+      "loss_token": 1.5402,
+      "lr": "2.00e-07",
+      "norm": 5.8262,
+      "step": 1157
+    },
+    {
+      "acc_char": 0.3198,
+      "acc_token": 0.7236,
+      "epoch": 0.02,
+      "loss_char": 0.6016,
+      "loss_token": 1.4951,
+      "lr": "2.00e-07",
+      "norm": 5.5455,
+      "step": 1158
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6685,
+      "epoch": 0.02,
+      "loss_char": 0.7937,
+      "loss_token": 1.6673,
+      "lr": "2.00e-07",
+      "norm": 5.9008,
+      "step": 1159
+    },
+    {
+      "acc_char": 0.3045,
+      "acc_token": 0.6974,
+      "epoch": 0.02,
+      "loss_char": 0.6667,
+      "loss_token": 1.5969,
+      "lr": "2.00e-07",
+      "norm": 5.8918,
+      "step": 1160
+    },
+    {
+      "acc_char": 0.3024,
+      "acc_token": 0.685,
+      "epoch": 0.02,
+      "loss_char": 0.7058,
+      "loss_token": 1.5962,
+      "lr": "2.00e-07",
+      "norm": 5.6316,
+      "step": 1161
+    },
+    {
+      "acc_char": 0.3056,
+      "acc_token": 0.6877,
+      "epoch": 0.02,
+      "loss_char": 0.7297,
+      "loss_token": 1.6399,
+      "lr": "2.00e-07",
+      "norm": 5.7485,
+      "step": 1162
+    },
+    {
+      "acc_char": 0.2939,
+      "acc_token": 0.6676,
+      "epoch": 0.02,
+      "loss_char": 0.7448,
+      "loss_token": 1.6946,
+      "lr": "2.00e-07",
+      "norm": 5.8275,
+      "step": 1163
+    },
+    {
+      "acc_char": 0.3009,
+      "acc_token": 0.6583,
+      "epoch": 0.02,
+      "loss_char": 0.7806,
+      "loss_token": 1.698,
+      "lr": "2.00e-07",
+      "norm": 5.6786,
+      "step": 1164
+    },
+    {
+      "acc_char": 0.2888,
+      "acc_token": 0.6536,
+      "epoch": 0.02,
+      "loss_char": 0.8236,
+      "loss_token": 1.7791,
+      "lr": "2.00e-07",
+      "norm": 6.2519,
+      "step": 1165
+    },
+    {
+      "acc_char": 0.3082,
+      "acc_token": 0.6946,
+      "epoch": 0.02,
+      "loss_char": 0.7213,
+      "loss_token": 1.6737,
+      "lr": "2.00e-07",
+      "norm": 5.6594,
+      "step": 1166
+    },
+    {
+      "acc_char": 0.2911,
+      "acc_token": 0.6662,
+      "epoch": 0.02,
+      "loss_char": 0.7801,
+      "loss_token": 1.7293,
+      "lr": "2.00e-07",
+      "norm": 5.8279,
+      "step": 1167
+    },
+    {
+      "acc_char": 0.2991,
+      "acc_token": 0.6862,
+      "epoch": 0.02,
+      "loss_char": 0.6893,
+      "loss_token": 1.6088,
+      "lr": "2.00e-07",
+      "norm": 5.7553,
+      "step": 1168
+    },
+    {
+      "acc_char": 0.3086,
+      "acc_token": 0.703,
+      "epoch": 0.02,
+      "loss_char": 0.6888,
+      "loss_token": 1.5782,
+      "lr": "2.00e-07",
+      "norm": 5.7771,
+      "step": 1169
+    },
+    {
+      "acc_char": 0.3103,
+      "acc_token": 0.713,
+      "epoch": 0.02,
+      "loss_char": 0.6274,
+      "loss_token": 1.5235,
+      "lr": "2.00e-07",
+      "norm": 5.775,
+      "step": 1170
+    },
+    {
+      "acc_char": 0.2932,
+      "acc_token": 0.6684,
+      "epoch": 0.02,
+      "loss_char": 0.7567,
+      "loss_token": 1.5922,
+      "lr": "2.00e-07",
+      "norm": 5.9991,
+      "step": 1171
+    },
+    {
+      "acc_char": 0.3162,
+      "acc_token": 0.7083,
+      "epoch": 0.02,
+      "loss_char": 0.6703,
+      "loss_token": 1.6602,
+      "lr": "2.00e-07",
+      "norm": 5.4752,
+      "step": 1172
+    },
+    {
+      "acc_char": 0.2892,
+      "acc_token": 0.6303,
+      "epoch": 0.02,
+      "loss_char": 0.8684,
+      "loss_token": 1.7126,
+      "lr": "2.00e-07",
+      "norm": 5.755,
+      "step": 1173
+    },
+    {
+      "acc_char": 0.2921,
+      "acc_token": 0.666,
+      "epoch": 0.02,
+      "loss_char": 0.7583,
+      "loss_token": 1.6371,
+      "lr": "2.00e-07",
+      "norm": 5.6528,
+      "step": 1174
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.686,
+      "epoch": 0.02,
+      "loss_char": 0.6875,
+      "loss_token": 1.5647,
+      "lr": "2.00e-07",
+      "norm": 5.4077,
+      "step": 1175
+    },
+    {
+      "acc_char": 0.3114,
+      "acc_token": 0.6969,
+      "epoch": 0.02,
+      "loss_char": 0.6888,
+      "loss_token": 1.5857,
+      "lr": "2.00e-07",
+      "norm": 5.8141,
+      "step": 1176
+    },
+    {
+      "acc_char": 0.3214,
+      "acc_token": 0.7065,
+      "epoch": 0.02,
+      "loss_char": 0.6635,
+      "loss_token": 1.4552,
+      "lr": "2.00e-07",
+      "norm": 5.7959,
+      "step": 1177
+    },
+    {
+      "acc_char": 0.3034,
+      "acc_token": 0.6808,
+      "epoch": 0.02,
+      "loss_char": 0.7381,
+      "loss_token": 1.7253,
+      "lr": "2.00e-07",
+      "norm": 5.9402,
+      "step": 1178
+    },
+    {
+      "acc_char": 0.3135,
+      "acc_token": 0.7093,
+      "epoch": 0.02,
+      "loss_char": 0.6598,
+      "loss_token": 1.5537,
+      "lr": "2.00e-07",
+      "norm": 5.4749,
+      "step": 1179
+    },
+    {
+      "acc_char": 0.2853,
+      "acc_token": 0.6284,
+      "epoch": 0.02,
+      "loss_char": 0.8759,
+      "loss_token": 1.733,
+      "lr": "2.00e-07",
+      "norm": 5.8521,
+      "step": 1180
+    },
+    {
+      "acc_char": 0.274,
+      "acc_token": 0.6284,
+      "epoch": 0.02,
+      "loss_char": 0.8762,
+      "loss_token": 1.7769,
+      "lr": "2.00e-07",
+      "norm": 5.8663,
+      "step": 1181
+    },
+    {
+      "acc_char": 0.3098,
+      "acc_token": 0.6998,
+      "epoch": 0.02,
+      "loss_char": 0.6929,
+      "loss_token": 1.5764,
+      "lr": "2.00e-07",
+      "norm": 5.6745,
+      "step": 1182
+    },
+    {
+      "acc_char": 0.3059,
+      "acc_token": 0.6963,
+      "epoch": 0.02,
+      "loss_char": 0.6876,
+      "loss_token": 1.6232,
+      "lr": "2.00e-07",
+      "norm": 5.8334,
+      "step": 1183
+    },
+    {
+      "acc_char": 0.3013,
+      "acc_token": 0.6847,
+      "epoch": 0.02,
+      "loss_char": 0.6982,
+      "loss_token": 1.5633,
+      "lr": "2.00e-07",
+      "norm": 5.6752,
+      "step": 1184
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6781,
+      "epoch": 0.02,
+      "loss_char": 0.7393,
+      "loss_token": 1.6662,
+      "lr": "2.00e-07",
+      "norm": 5.9113,
+      "step": 1185
+    },
+    {
+      "acc_char": 0.2929,
+      "acc_token": 0.6713,
+      "epoch": 0.02,
+      "loss_char": 0.7428,
+      "loss_token": 1.6182,
+      "lr": "2.00e-07",
+      "norm": 6.1931,
+      "step": 1186
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6851,
+      "epoch": 0.02,
+      "loss_char": 0.6948,
+      "loss_token": 1.6335,
+      "lr": "2.00e-07",
+      "norm": 5.8183,
+      "step": 1187
+    },
+    {
+      "acc_char": 0.3169,
+      "acc_token": 0.708,
+      "epoch": 0.02,
+      "loss_char": 0.6941,
+      "loss_token": 1.5985,
+      "lr": "2.00e-07",
+      "norm": 6.1227,
+      "step": 1188
+    },
+    {
+      "acc_char": 0.2947,
+      "acc_token": 0.6685,
+      "epoch": 0.02,
+      "loss_char": 0.7599,
+      "loss_token": 1.6387,
+      "lr": "2.00e-07",
+      "norm": 5.9399,
+      "step": 1189
+    },
+    {
+      "acc_char": 0.2882,
+      "acc_token": 0.6734,
+      "epoch": 0.02,
+      "loss_char": 0.7478,
+      "loss_token": 1.6861,
+      "lr": "2.00e-07",
+      "norm": 5.8545,
+      "step": 1190
+    },
+    {
+      "acc_char": 0.2996,
+      "acc_token": 0.6772,
+      "epoch": 0.02,
+      "loss_char": 0.7301,
+      "loss_token": 1.6765,
+      "lr": "2.00e-07",
+      "norm": 6.0008,
+      "step": 1191
+    },
+    {
+      "acc_char": 0.3056,
+      "acc_token": 0.6815,
+      "epoch": 0.02,
+      "loss_char": 0.7253,
+      "loss_token": 1.5029,
+      "lr": "2.00e-07",
+      "norm": 5.5761,
+      "step": 1192
+    },
+    {
+      "acc_char": 0.2976,
+      "acc_token": 0.6713,
+      "epoch": 0.02,
+      "loss_char": 0.7522,
+      "loss_token": 1.6588,
+      "lr": "2.00e-07",
+      "norm": 5.679,
+      "step": 1193
+    },
+    {
+      "acc_char": 0.2938,
+      "acc_token": 0.668,
+      "epoch": 0.02,
+      "loss_char": 0.7532,
+      "loss_token": 1.6303,
+      "lr": "2.00e-07",
+      "norm": 5.6541,
+      "step": 1194
+    },
+    {
+      "acc_char": 0.3037,
+      "acc_token": 0.6695,
+      "epoch": 0.02,
+      "loss_char": 0.7715,
+      "loss_token": 1.6237,
+      "lr": "2.00e-07",
+      "norm": 6.1823,
+      "step": 1195
+    },
+    {
+      "acc_char": 0.298,
+      "acc_token": 0.6762,
+      "epoch": 0.02,
+      "loss_char": 0.7422,
+      "loss_token": 1.5991,
+      "lr": "2.00e-07",
+      "norm": 5.8473,
+      "step": 1196
+    },
+    {
+      "acc_char": 0.3136,
+      "acc_token": 0.7075,
+      "epoch": 0.02,
+      "loss_char": 0.6773,
+      "loss_token": 1.5642,
+      "lr": "2.00e-07",
+      "norm": 5.6722,
+      "step": 1197
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6751,
+      "epoch": 0.02,
+      "loss_char": 0.7141,
+      "loss_token": 1.6704,
+      "lr": "2.00e-07",
+      "norm": 6.141,
+      "step": 1198
+    },
+    {
+      "acc_char": 0.296,
+      "acc_token": 0.6694,
+      "epoch": 0.02,
+      "loss_char": 0.7446,
+      "loss_token": 1.6256,
+      "lr": "2.00e-07",
+      "norm": 5.942,
+      "step": 1199
+    },
+    {
+      "acc_char": 0.2992,
+      "acc_token": 0.6891,
+      "epoch": 0.02,
+      "loss_char": 0.7081,
+      "loss_token": 1.5872,
+      "lr": "2.00e-07",
+      "norm": 5.6528,
+      "step": 1200
+    },
+    {
+      "acc_char": 0.3113,
+      "acc_token": 0.6944,
+      "epoch": 0.02,
+      "loss_char": 0.6927,
+      "loss_token": 1.5479,
+      "lr": "2.00e-07",
+      "norm": 5.574,
+      "step": 1201
+    },
+    {
+      "acc_char": 0.2949,
+      "acc_token": 0.6881,
+      "epoch": 0.02,
+      "loss_char": 0.6741,
+      "loss_token": 1.5658,
+      "lr": "2.00e-07",
+      "norm": 5.6061,
+      "step": 1202
+    },
+    {
+      "acc_char": 0.3162,
+      "acc_token": 0.7157,
+      "epoch": 0.02,
+      "loss_char": 0.6284,
+      "loss_token": 1.5231,
+      "lr": "2.00e-07",
+      "norm": 5.5183,
+      "step": 1203
+    },
+    {
+      "acc_char": 0.2949,
+      "acc_token": 0.6586,
+      "epoch": 0.02,
+      "loss_char": 0.7916,
+      "loss_token": 1.725,
+      "lr": "2.00e-07",
+      "norm": 5.9887,
+      "step": 1204
+    },
+    {
+      "acc_char": 0.3105,
+      "acc_token": 0.7052,
+      "epoch": 0.02,
+      "loss_char": 0.6655,
+      "loss_token": 1.6009,
+      "lr": "2.00e-07",
+      "norm": 5.7779,
+      "step": 1205
+    },
+    {
+      "acc_char": 0.3019,
+      "acc_token": 0.6792,
+      "epoch": 0.02,
+      "loss_char": 0.7376,
+      "loss_token": 1.5953,
+      "lr": "2.00e-07",
+      "norm": 5.8897,
+      "step": 1206
+    },
+    {
+      "acc_char": 0.3059,
+      "acc_token": 0.6879,
+      "epoch": 0.02,
+      "loss_char": 0.7213,
+      "loss_token": 1.5719,
+      "lr": "2.00e-07",
+      "norm": 5.6923,
+      "step": 1207
+    },
+    {
+      "acc_char": 0.2865,
+      "acc_token": 0.6294,
+      "epoch": 0.02,
+      "loss_char": 0.8252,
+      "loss_token": 1.7565,
+      "lr": "2.00e-07",
+      "norm": 5.7013,
+      "step": 1208
+    },
+    {
+      "acc_char": 0.3107,
+      "acc_token": 0.6958,
+      "epoch": 0.02,
+      "loss_char": 0.7085,
+      "loss_token": 1.6167,
+      "lr": "2.00e-07",
+      "norm": 5.6654,
+      "step": 1209
+    },
+    {
+      "acc_char": 0.3115,
+      "acc_token": 0.6773,
+      "epoch": 0.02,
+      "loss_char": 0.7292,
+      "loss_token": 1.5463,
+      "lr": "2.00e-07",
+      "norm": 5.4154,
+      "step": 1210
+    },
+    {
+      "acc_char": 0.3196,
+      "acc_token": 0.7027,
+      "epoch": 0.02,
+      "loss_char": 0.6707,
+      "loss_token": 1.5965,
+      "lr": "2.00e-07",
+      "norm": 6.0887,
+      "step": 1211
+    },
+    {
+      "acc_char": 0.3162,
+      "acc_token": 0.7024,
+      "epoch": 0.02,
+      "loss_char": 0.6739,
+      "loss_token": 1.5327,
+      "lr": "2.00e-07",
+      "norm": 5.9695,
+      "step": 1212
+    },
+    {
+      "acc_char": 0.3201,
+      "acc_token": 0.7112,
+      "epoch": 0.02,
+      "loss_char": 0.6603,
+      "loss_token": 1.4594,
+      "lr": "2.00e-07",
+      "norm": 5.4631,
+      "step": 1213
+    },
+    {
+      "acc_char": 0.3123,
+      "acc_token": 0.7183,
+      "epoch": 0.02,
+      "loss_char": 0.606,
+      "loss_token": 1.4774,
+      "lr": "2.00e-07",
+      "norm": 5.5688,
+      "step": 1214
+    },
+    {
+      "acc_char": 0.308,
+      "acc_token": 0.6938,
+      "epoch": 0.02,
+      "loss_char": 0.6814,
+      "loss_token": 1.525,
+      "lr": "2.00e-07",
+      "norm": 5.4292,
+      "step": 1215
+    },
+    {
+      "acc_char": 0.3054,
+      "acc_token": 0.6819,
+      "epoch": 0.02,
+      "loss_char": 0.7364,
+      "loss_token": 1.6021,
+      "lr": "2.00e-07",
+      "norm": 5.5374,
+      "step": 1216
+    },
+    {
+      "acc_char": 0.3228,
+      "acc_token": 0.7098,
+      "epoch": 0.02,
+      "loss_char": 0.6828,
+      "loss_token": 1.5927,
+      "lr": "2.00e-07",
+      "norm": 6.0024,
+      "step": 1217
+    },
+    {
+      "acc_char": 0.3128,
+      "acc_token": 0.69,
+      "epoch": 0.02,
+      "loss_char": 0.717,
+      "loss_token": 1.61,
+      "lr": "2.00e-07",
+      "norm": 5.756,
+      "step": 1218
+    },
+    {
+      "acc_char": 0.3006,
+      "acc_token": 0.684,
+      "epoch": 0.02,
+      "loss_char": 0.7142,
+      "loss_token": 1.6364,
+      "lr": "2.00e-07",
+      "norm": 5.6081,
+      "step": 1219
+    },
+    {
+      "acc_char": 0.2954,
+      "acc_token": 0.6756,
+      "epoch": 0.02,
+      "loss_char": 0.7117,
+      "loss_token": 1.5118,
+      "lr": "2.00e-07",
+      "norm": 5.7647,
+      "step": 1220
+    },
+    {
+      "acc_char": 0.3062,
+      "acc_token": 0.7,
+      "epoch": 0.02,
+      "loss_char": 0.6784,
+      "loss_token": 1.6177,
+      "lr": "2.00e-07",
+      "norm": 5.5468,
+      "step": 1221
+    },
+    {
+      "acc_char": 0.2938,
+      "acc_token": 0.6753,
+      "epoch": 0.02,
+      "loss_char": 0.7119,
+      "loss_token": 1.5672,
+      "lr": "2.00e-07",
+      "norm": 6.0324,
+      "step": 1222
+    },
+    {
+      "acc_char": 0.3025,
+      "acc_token": 0.6842,
+      "epoch": 0.02,
+      "loss_char": 0.7241,
+      "loss_token": 1.6795,
+      "lr": "2.00e-07",
+      "norm": 6.0055,
+      "step": 1223
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6746,
+      "epoch": 0.02,
+      "loss_char": 0.7454,
+      "loss_token": 1.6286,
+      "lr": "2.00e-07",
+      "norm": 5.8116,
+      "step": 1224
+    },
+    {
+      "acc_char": 0.2895,
+      "acc_token": 0.6561,
+      "epoch": 0.02,
+      "loss_char": 0.7919,
+      "loss_token": 1.7589,
+      "lr": "2.00e-07",
+      "norm": 6.1568,
+      "step": 1225
+    },
+    {
+      "acc_char": 0.3112,
+      "acc_token": 0.6893,
+      "epoch": 0.02,
+      "loss_char": 0.7214,
+      "loss_token": 1.5514,
+      "lr": "2.00e-07",
+      "norm": 6.0475,
+      "step": 1226
+    },
+    {
+      "acc_char": 0.3006,
+      "acc_token": 0.6802,
+      "epoch": 0.02,
+      "loss_char": 0.7417,
+      "loss_token": 1.6405,
+      "lr": "2.00e-07",
+      "norm": 5.888,
+      "step": 1227
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6747,
+      "epoch": 0.02,
+      "loss_char": 0.7373,
+      "loss_token": 1.6025,
+      "lr": "2.00e-07",
+      "norm": 5.9288,
+      "step": 1228
+    },
+    {
+      "acc_char": 0.3079,
+      "acc_token": 0.6781,
+      "epoch": 0.02,
+      "loss_char": 0.7676,
+      "loss_token": 1.6012,
+      "lr": "2.00e-07",
+      "norm": 5.8617,
+      "step": 1229
+    },
+    {
+      "acc_char": 0.3014,
+      "acc_token": 0.6851,
+      "epoch": 0.02,
+      "loss_char": 0.7017,
+      "loss_token": 1.6302,
+      "lr": "2.00e-07",
+      "norm": 5.8773,
+      "step": 1230
+    },
+    {
+      "acc_char": 0.2894,
+      "acc_token": 0.6596,
+      "epoch": 0.02,
+      "loss_char": 0.7961,
+      "loss_token": 1.6606,
+      "lr": "2.00e-07",
+      "norm": 5.7945,
+      "step": 1231
+    },
+    {
+      "acc_char": 0.3134,
+      "acc_token": 0.696,
+      "epoch": 0.02,
+      "loss_char": 0.7082,
+      "loss_token": 1.5876,
+      "lr": "2.00e-07",
+      "norm": 5.7545,
+      "step": 1232
+    },
+    {
+      "acc_char": 0.2867,
+      "acc_token": 0.6524,
+      "epoch": 0.02,
+      "loss_char": 0.7828,
+      "loss_token": 1.6832,
+      "lr": "2.00e-07",
+      "norm": 5.6264,
+      "step": 1233
+    },
+    {
+      "acc_char": 0.2902,
+      "acc_token": 0.6328,
+      "epoch": 0.02,
+      "loss_char": 0.913,
+      "loss_token": 1.7744,
+      "lr": "2.00e-07",
+      "norm": 6.6499,
+      "step": 1234
+    },
+    {
+      "acc_char": 0.315,
+      "acc_token": 0.7204,
+      "epoch": 0.02,
+      "loss_char": 0.623,
+      "loss_token": 1.422,
+      "lr": "2.00e-07",
+      "norm": 5.5104,
+      "step": 1235
+    },
+    {
+      "acc_char": 0.3039,
+      "acc_token": 0.6784,
+      "epoch": 0.02,
+      "loss_char": 0.7408,
+      "loss_token": 1.6017,
+      "lr": "2.00e-07",
+      "norm": 2.732,
+      "step": 1236
+    },
+    {
+      "acc_char": 0.3078,
+      "acc_token": 0.6977,
+      "epoch": 0.02,
+      "loss_char": 0.6765,
+      "loss_token": 1.6248,
+      "lr": "2.00e-07",
+      "norm": 5.7629,
+      "step": 1237
+    },
+    {
+      "acc_char": 0.3089,
+      "acc_token": 0.7052,
+      "epoch": 0.02,
+      "loss_char": 0.6535,
+      "loss_token": 1.5843,
+      "lr": "2.00e-07",
+      "norm": 5.6524,
+      "step": 1238
+    },
+    {
+      "acc_char": 0.319,
+      "acc_token": 0.7262,
+      "epoch": 0.02,
+      "loss_char": 0.615,
+      "loss_token": 1.5164,
+      "lr": "2.00e-07",
+      "norm": 5.5803,
+      "step": 1239
+    },
+    {
+      "acc_char": 0.2951,
+      "acc_token": 0.6735,
+      "epoch": 0.02,
+      "loss_char": 0.7513,
+      "loss_token": 1.6776,
+      "lr": "2.00e-07",
+      "norm": 5.7461,
+      "step": 1240
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6742,
+      "epoch": 0.02,
+      "loss_char": 0.7417,
+      "loss_token": 1.6581,
+      "lr": "2.00e-07",
+      "norm": 5.9435,
+      "step": 1241
+    },
+    {
+      "acc_char": 0.2895,
+      "acc_token": 0.6463,
+      "epoch": 0.02,
+      "loss_char": 0.8473,
+      "loss_token": 1.8006,
+      "lr": "2.00e-07",
+      "norm": 6.0168,
+      "step": 1242
+    },
+    {
+      "acc_char": 0.3064,
+      "acc_token": 0.7035,
+      "epoch": 0.02,
+      "loss_char": 0.6319,
+      "loss_token": 1.5342,
+      "lr": "2.00e-07",
+      "norm": 5.755,
+      "step": 1243
+    },
+    {
+      "acc_char": 0.3018,
+      "acc_token": 0.6873,
+      "epoch": 0.02,
+      "loss_char": 0.6987,
+      "loss_token": 1.5935,
+      "lr": "2.00e-07",
+      "norm": 5.5453,
+      "step": 1244
+    },
+    {
+      "acc_char": 0.2961,
+      "acc_token": 0.6757,
+      "epoch": 0.02,
+      "loss_char": 0.7513,
+      "loss_token": 1.6462,
+      "lr": "2.00e-07",
+      "norm": 7.3667,
+      "step": 1245
+    },
+    {
+      "acc_char": 0.31,
+      "acc_token": 0.6903,
+      "epoch": 0.02,
+      "loss_char": 0.7214,
+      "loss_token": 1.5986,
+      "lr": "2.00e-07",
+      "norm": 5.6851,
+      "step": 1246
+    },
+    {
+      "acc_char": 0.3029,
+      "acc_token": 0.6882,
+      "epoch": 0.02,
+      "loss_char": 0.7119,
+      "loss_token": 1.6474,
+      "lr": "2.00e-07",
+      "norm": 5.829,
+      "step": 1247
+    },
+    {
+      "acc_char": 0.3095,
+      "acc_token": 0.7146,
+      "epoch": 0.02,
+      "loss_char": 0.6218,
+      "loss_token": 1.5106,
+      "lr": "2.00e-07",
+      "norm": 5.4342,
+      "step": 1248
+    },
+    {
+      "acc_char": 0.2975,
+      "acc_token": 0.6854,
+      "epoch": 0.02,
+      "loss_char": 0.6857,
+      "loss_token": 1.6246,
+      "lr": "2.00e-07",
+      "norm": 5.7823,
+      "step": 1249
+    },
+    {
+      "acc_char": 0.3036,
+      "acc_token": 0.6944,
+      "epoch": 0.02,
+      "loss_char": 0.6883,
+      "loss_token": 1.5675,
+      "lr": "2.00e-07",
+      "norm": 5.4636,
+      "step": 1250
+    },
+    {
+      "acc_char": 0.3069,
+      "acc_token": 0.6922,
+      "epoch": 0.02,
+      "loss_char": 0.7031,
+      "loss_token": 1.5125,
+      "lr": "2.00e-07",
+      "norm": 5.4007,
+      "step": 1251
+    },
+    {
+      "acc_char": 0.3049,
+      "acc_token": 0.6812,
+      "epoch": 0.02,
+      "loss_char": 0.7204,
+      "loss_token": 1.5756,
+      "lr": "2.00e-07",
+      "norm": 5.9017,
+      "step": 1252
+    },
+    {
+      "acc_char": 0.3148,
+      "acc_token": 0.7214,
+      "epoch": 0.02,
+      "loss_char": 0.6013,
+      "loss_token": 1.4596,
+      "lr": "2.00e-07",
+      "norm": 5.4911,
+      "step": 1253
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.6694,
+      "epoch": 0.02,
+      "loss_char": 0.7707,
+      "loss_token": 1.5119,
+      "lr": "2.00e-07",
+      "norm": 5.656,
+      "step": 1254
+    },
+    {
+      "acc_char": 0.3015,
+      "acc_token": 0.6944,
+      "epoch": 0.02,
+      "loss_char": 0.6839,
+      "loss_token": 1.6382,
+      "lr": "2.00e-07",
+      "norm": 5.6153,
+      "step": 1255
+    },
+    {
+      "acc_char": 0.3001,
+      "acc_token": 0.6474,
+      "epoch": 0.02,
+      "loss_char": 0.8154,
+      "loss_token": 1.6608,
+      "lr": "2.00e-07",
+      "norm": 5.5675,
+      "step": 1256
+    },
+    {
+      "acc_char": 0.3161,
+      "acc_token": 0.704,
+      "epoch": 0.02,
+      "loss_char": 0.6828,
+      "loss_token": 1.5635,
+      "lr": "2.00e-07",
+      "norm": 5.5283,
+      "step": 1257
+    },
+    {
+      "acc_char": 0.3108,
+      "acc_token": 0.6962,
+      "epoch": 0.02,
+      "loss_char": 0.6988,
+      "loss_token": 1.5846,
+      "lr": "2.00e-07",
+      "norm": 5.6018,
+      "step": 1258
+    },
+    {
+      "acc_char": 0.2944,
+      "acc_token": 0.6763,
+      "epoch": 0.02,
+      "loss_char": 0.7116,
+      "loss_token": 1.607,
+      "lr": "2.00e-07",
+      "norm": 5.6171,
+      "step": 1259
+    },
+    {
+      "acc_char": 0.2987,
+      "acc_token": 0.6785,
+      "epoch": 0.02,
+      "loss_char": 0.7239,
+      "loss_token": 1.6913,
+      "lr": "2.00e-07",
+      "norm": 5.7759,
+      "step": 1260
+    },
+    {
+      "acc_char": 0.3042,
+      "acc_token": 0.6966,
+      "epoch": 0.02,
+      "loss_char": 0.6699,
+      "loss_token": 1.5636,
+      "lr": "2.00e-07",
+      "norm": 6.2067,
+      "step": 1261
+    },
+    {
+      "acc_char": 0.2987,
+      "acc_token": 0.6694,
+      "epoch": 0.02,
+      "loss_char": 0.7628,
+      "loss_token": 1.6383,
+      "lr": "2.00e-07",
+      "norm": 5.6774,
+      "step": 1262
+    },
+    {
+      "acc_char": 0.2891,
+      "acc_token": 0.6604,
+      "epoch": 0.02,
+      "loss_char": 0.7724,
+      "loss_token": 1.6981,
+      "lr": "2.00e-07",
+      "norm": 5.9932,
+      "step": 1263
+    },
+    {
+      "acc_char": 0.2997,
+      "acc_token": 0.6786,
+      "epoch": 0.02,
+      "loss_char": 0.7339,
+      "loss_token": 1.5778,
+      "lr": "2.00e-07",
+      "norm": 5.6281,
+      "step": 1264
+    },
+    {
+      "acc_char": 0.2989,
+      "acc_token": 0.6753,
+      "epoch": 0.02,
+      "loss_char": 0.7396,
+      "loss_token": 1.5632,
+      "lr": "2.00e-07",
+      "norm": 5.6482,
+      "step": 1265
+    },
+    {
+      "acc_char": 0.3088,
+      "acc_token": 0.7037,
+      "epoch": 0.02,
+      "loss_char": 0.6675,
+      "loss_token": 1.4307,
+      "lr": "2.00e-07",
+      "norm": 5.6762,
+      "step": 1266
+    },
+    {
+      "acc_char": 0.3097,
+      "acc_token": 0.7041,
+      "epoch": 0.02,
+      "loss_char": 0.6608,
+      "loss_token": 1.5348,
+      "lr": "2.00e-07",
+      "norm": 5.767,
+      "step": 1267
+    },
+    {
+      "acc_char": 0.2969,
+      "acc_token": 0.6725,
+      "epoch": 0.02,
+      "loss_char": 0.7345,
+      "loss_token": 1.68,
+      "lr": "2.00e-07",
+      "norm": 5.801,
+      "step": 1268
+    },
+    {
+      "acc_char": 0.3099,
+      "acc_token": 0.6968,
+      "epoch": 0.02,
+      "loss_char": 0.6917,
+      "loss_token": 1.5728,
+      "lr": "2.00e-07",
+      "norm": 5.5568,
+      "step": 1269
+    },
+    {
+      "acc_char": 0.3226,
+      "acc_token": 0.6871,
+      "epoch": 0.02,
+      "loss_char": 0.7358,
+      "loss_token": 1.5833,
+      "lr": "2.00e-07",
+      "norm": 5.5733,
+      "step": 1270
+    },
+    {
+      "acc_char": 0.3057,
+      "acc_token": 0.6873,
+      "epoch": 0.02,
+      "loss_char": 0.7212,
+      "loss_token": 1.669,
+      "lr": "2.00e-07",
+      "norm": 5.8983,
+      "step": 1271
+    },
+    {
+      "acc_char": 0.3064,
+      "acc_token": 0.6861,
+      "epoch": 0.02,
+      "loss_char": 0.7191,
+      "loss_token": 1.6987,
+      "lr": "2.00e-07",
+      "norm": 6.279,
+      "step": 1272
+    },
+    {
+      "acc_char": 0.2931,
+      "acc_token": 0.6685,
+      "epoch": 0.02,
+      "loss_char": 0.75,
+      "loss_token": 1.5613,
+      "lr": "2.00e-07",
+      "norm": 5.8721,
+      "step": 1273
+    },
+    {
+      "acc_char": 0.2968,
+      "acc_token": 0.6677,
+      "epoch": 0.02,
+      "loss_char": 0.7697,
+      "loss_token": 1.639,
+      "lr": "2.00e-07",
+      "norm": 5.499,
+      "step": 1274
+    },
+    {
+      "acc_char": 0.3112,
+      "acc_token": 0.6988,
+      "epoch": 0.02,
+      "loss_char": 0.6827,
+      "loss_token": 1.5999,
+      "lr": "2.00e-07",
+      "norm": 5.6463,
+      "step": 1275
+    },
+    {
+      "acc_char": 0.3173,
+      "acc_token": 0.7016,
+      "epoch": 0.02,
+      "loss_char": 0.6897,
+      "loss_token": 1.515,
+      "lr": "2.00e-07",
+      "norm": 5.4869,
+      "step": 1276
+    },
+    {
+      "acc_char": 0.2954,
+      "acc_token": 0.6773,
+      "epoch": 0.02,
+      "loss_char": 0.7179,
+      "loss_token": 1.6558,
+      "lr": "2.00e-07",
+      "norm": 5.7433,
+      "step": 1277
+    },
+    {
+      "acc_char": 0.3029,
+      "acc_token": 0.6768,
+      "epoch": 0.02,
+      "loss_char": 0.7316,
+      "loss_token": 1.5784,
+      "lr": "2.00e-07",
+      "norm": 5.8841,
+      "step": 1278
+    },
+    {
+      "acc_char": 0.2838,
+      "acc_token": 0.6648,
+      "epoch": 0.02,
+      "loss_char": 0.7332,
+      "loss_token": 1.621,
+      "lr": "2.00e-07",
+      "norm": 5.819,
+      "step": 1279
+    },
+    {
+      "acc_char": 0.309,
+      "acc_token": 0.6922,
+      "epoch": 0.02,
+      "loss_char": 0.6954,
+      "loss_token": 1.5557,
+      "lr": "2.00e-07",
+      "norm": 5.6622,
+      "step": 1280
+    },
+    {
+      "acc_char": 0.2977,
+      "acc_token": 0.6724,
+      "epoch": 0.02,
+      "loss_char": 0.7497,
+      "loss_token": 1.6469,
+      "lr": "2.00e-07",
+      "norm": 6.0357,
+      "step": 1281
+    },
+    {
+      "acc_char": 0.2991,
+      "acc_token": 0.693,
+      "epoch": 0.02,
+      "loss_char": 0.6675,
+      "loss_token": 1.5335,
+      "lr": "2.00e-07",
+      "norm": 5.398,
+      "step": 1282
+    },
+    {
+      "acc_char": 0.2906,
+      "acc_token": 0.6389,
+      "epoch": 0.02,
+      "loss_char": 0.8624,
+      "loss_token": 1.8026,
+      "lr": "2.00e-07",
+      "norm": 5.4403,
+      "step": 1283
+    },
+    {
+      "acc_char": 0.3027,
+      "acc_token": 0.6786,
+      "epoch": 0.02,
+      "loss_char": 0.7183,
+      "loss_token": 1.567,
+      "lr": "2.00e-07",
+      "norm": 5.4143,
+      "step": 1284
+    },
+    {
+      "acc_char": 0.2908,
+      "acc_token": 0.6395,
+      "epoch": 0.02,
+      "loss_char": 0.8086,
+      "loss_token": 1.671,
+      "lr": "2.00e-07",
+      "norm": 5.4947,
+      "step": 1285
+    },
+    {
+      "acc_char": 0.3081,
+      "acc_token": 0.6976,
+      "epoch": 0.02,
+      "loss_char": 0.674,
+      "loss_token": 1.5783,
+      "lr": "2.00e-07",
+      "norm": 5.4502,
+      "step": 1286
+    },
+    {
+      "acc_char": 0.2987,
+      "acc_token": 0.6731,
+      "epoch": 0.02,
+      "loss_char": 0.7479,
+      "loss_token": 1.6846,
+      "lr": "2.00e-07",
+      "norm": 5.959,
+      "step": 1287
+    },
+    {
+      "acc_char": 0.3125,
+      "acc_token": 0.6928,
+      "epoch": 0.02,
+      "loss_char": 0.7215,
+      "loss_token": 1.5817,
+      "lr": "2.00e-07",
+      "norm": 5.5225,
+      "step": 1288
+    },
+    {
+      "acc_char": 0.2924,
+      "acc_token": 0.6698,
+      "epoch": 0.02,
+      "loss_char": 0.7396,
+      "loss_token": 1.5918,
+      "lr": "2.00e-07",
+      "norm": 5.7961,
+      "step": 1289
+    },
+    {
+      "acc_char": 0.3046,
+      "acc_token": 0.6958,
+      "epoch": 0.02,
+      "loss_char": 0.6932,
+      "loss_token": 1.5588,
+      "lr": "2.00e-07",
+      "norm": 5.6947,
+      "step": 1290
+    },
+    {
+      "acc_char": 0.2926,
+      "acc_token": 0.6789,
+      "epoch": 0.02,
+      "loss_char": 0.6949,
+      "loss_token": 1.6479,
+      "lr": "2.00e-07",
+      "norm": 5.9985,
+      "step": 1291
+    },
+    {
+      "acc_char": 0.2909,
+      "acc_token": 0.6287,
+      "epoch": 0.02,
+      "loss_char": 0.8573,
+      "loss_token": 1.7134,
+      "lr": "2.00e-07",
+      "norm": 5.5973,
+      "step": 1292
+    },
+    {
+      "acc_char": 0.3081,
+      "acc_token": 0.702,
+      "epoch": 0.02,
+      "loss_char": 0.655,
+      "loss_token": 1.5874,
+      "lr": "2.00e-07",
+      "norm": 5.5553,
+      "step": 1293
+    },
+    {
+      "acc_char": 0.2935,
+      "acc_token": 0.6674,
+      "epoch": 0.02,
+      "loss_char": 0.7717,
+      "loss_token": 1.6123,
+      "lr": "2.00e-07",
+      "norm": 5.7581,
+      "step": 1294
+    },
+    {
+      "acc_char": 0.3015,
+      "acc_token": 0.6914,
+      "epoch": 0.02,
+      "loss_char": 0.6618,
+      "loss_token": 1.5497,
+      "lr": "2.00e-07",
+      "norm": 5.5086,
+      "step": 1295
+    },
+    {
+      "acc_char": 0.303,
+      "acc_token": 0.6814,
+      "epoch": 0.02,
+      "loss_char": 0.7289,
+      "loss_token": 1.6631,
+      "lr": "2.00e-07",
+      "norm": 5.679,
+      "step": 1296
+    },
+    {
+      "acc_char": 0.2966,
+      "acc_token": 0.682,
+      "epoch": 0.02,
+      "loss_char": 0.7135,
+      "loss_token": 1.5486,
+      "lr": "2.00e-07",
+      "norm": 5.9415,
+      "step": 1297
+    },
+    {
+      "acc_char": 0.3011,
+      "acc_token": 0.6693,
+      "epoch": 0.02,
+      "loss_char": 0.777,
+      "loss_token": 1.6275,
+      "lr": "2.00e-07",
+      "norm": 6.2361,
+      "step": 1298
+    },
+    {
+      "acc_char": 0.3191,
+      "acc_token": 0.706,
+      "epoch": 0.02,
+      "loss_char": 0.6841,
+      "loss_token": 1.5494,
+      "lr": "2.00e-07",
+      "norm": 5.7965,
+      "step": 1299
+    },
+    {
+      "acc_char": 0.3061,
+      "acc_token": 0.6798,
+      "epoch": 0.02,
+      "loss_char": 0.7659,
+      "loss_token": 1.6338,
+      "lr": "2.00e-07",
+      "norm": 5.727,
+      "step": 1300
+    },
+    {
+      "acc_char": 0.2989,
+      "acc_token": 0.6424,
+      "epoch": 0.02,
+      "loss_char": 0.8275,
+      "loss_token": 1.6732,
+      "lr": "2.00e-07",
+      "norm": 5.9103,
+      "step": 1301
+    },
+    {
+      "acc_char": 0.3073,
+      "acc_token": 0.6837,
+      "epoch": 0.02,
+      "loss_char": 0.7156,
+      "loss_token": 1.5905,
+      "lr": "2.00e-07",
+      "norm": 5.7508,
+      "step": 1302
+    },
+    {
+      "acc_char": 0.3092,
+      "acc_token": 0.7093,
+      "epoch": 0.02,
+      "loss_char": 0.6465,
+      "loss_token": 1.5352,
+      "lr": "2.00e-07",
+      "norm": 5.3712,
+      "step": 1303
+    },
+    {
+      "acc_char": 0.3046,
+      "acc_token": 0.6859,
+      "epoch": 0.02,
+      "loss_char": 0.7196,
+      "loss_token": 1.6089,
+      "lr": "2.00e-07",
+      "norm": 5.7123,
+      "step": 1304
+    },
+    {
+      "acc_char": 0.295,
+      "acc_token": 0.6642,
+      "epoch": 0.02,
+      "loss_char": 0.7776,
+      "loss_token": 1.6871,
+      "lr": "2.00e-07",
+      "norm": 5.9281,
+      "step": 1305
+    },
+    {
+      "acc_char": 0.3124,
+      "acc_token": 0.7126,
+      "epoch": 0.02,
+      "loss_char": 0.6216,
+      "loss_token": 1.4293,
+      "lr": "2.00e-07",
+      "norm": 5.403,
+      "step": 1306
+    },
+    {
+      "acc_char": 0.2982,
+      "acc_token": 0.6654,
+      "epoch": 0.02,
+      "loss_char": 0.7713,
+      "loss_token": 1.7133,
+      "lr": "2.00e-07",
+      "norm": 6.1354,
+      "step": 1307
+    },
+    {
+      "acc_char": 0.2972,
+      "acc_token": 0.6834,
+      "epoch": 0.02,
+      "loss_char": 0.7293,
+      "loss_token": 1.5969,
+      "lr": "2.00e-07",
+      "norm": 5.713,
+      "step": 1308
+    },
+    {
+      "acc_char": 0.3121,
+      "acc_token": 0.6991,
+      "epoch": 0.02,
+      "loss_char": 0.6729,
+      "loss_token": 1.5578,
+      "lr": "2.00e-07",
+      "norm": 5.6073,
+      "step": 1309
+    },
+    {
+      "acc_char": 0.3035,
+      "acc_token": 0.6826,
+      "epoch": 0.02,
+      "loss_char": 0.7313,
+      "loss_token": 1.647,
+      "lr": "2.00e-07",
+      "norm": 5.9066,
+      "step": 1310
+    },
+    {
+      "acc_char": 0.3092,
+      "acc_token": 0.6894,
+      "epoch": 0.02,
+      "loss_char": 0.712,
+      "loss_token": 1.5379,
+      "lr": "2.00e-07",
+      "norm": 5.5333,
+      "step": 1311
+    },
+    {
+      "acc_char": 0.3037,
+      "acc_token": 0.6957,
+      "epoch": 0.02,
+      "loss_char": 0.6818,
+      "loss_token": 1.5175,
+      "lr": "2.00e-07",
+      "norm": 6.0989,
+      "step": 1312
+    },
+    {
+      "acc_char": 0.3087,
+      "acc_token": 0.6977,
+      "epoch": 0.02,
+      "loss_char": 0.6731,
+      "loss_token": 1.5837,
+      "lr": "2.00e-07",
+      "norm": 5.8096,
+      "step": 1313
+    },
+    {
+      "acc_char": 0.3152,
+      "acc_token": 0.7096,
+      "epoch": 0.02,
+      "loss_char": 0.6565,
+      "loss_token": 1.5564,
+      "lr": "2.00e-07",
+      "norm": 5.4834,
+      "step": 1314
+    },
+    {
+      "acc_char": 0.311,
+      "acc_token": 0.7085,
+      "epoch": 0.02,
+      "loss_char": 0.6535,
+      "loss_token": 1.5775,
+      "lr": "2.00e-07",
+      "norm": 5.5097,
+      "step": 1315
+    },
+    {
+      "acc_char": 0.2956,
+      "acc_token": 0.6631,
+      "epoch": 0.02,
+      "loss_char": 0.7923,
+      "loss_token": 1.713,
+      "lr": "2.00e-07",
+      "norm": 5.9354,
+      "step": 1316
+    },
+    {
+      "acc_char": 0.3073,
+      "acc_token": 0.6969,
+      "epoch": 0.02,
+      "loss_char": 0.6925,
+      "loss_token": 1.6065,
+      "lr": "2.00e-07",
+      "norm": 6.1533,
+      "step": 1317
+    },
+    {
+      "acc_char": 0.3259,
+      "acc_token": 0.7287,
+      "epoch": 0.02,
+      "loss_char": 0.5966,
+      "loss_token": 1.4401,
+      "lr": "2.00e-07",
+      "norm": 5.7195,
+      "step": 1318
+    },
+    {
+      "acc_char": 0.3141,
+      "acc_token": 0.7055,
+      "epoch": 0.02,
+      "loss_char": 0.6725,
+      "loss_token": 1.5512,
+      "lr": "2.00e-07",
+      "norm": 5.7712,
+      "step": 1319
+    },
+    {
+      "acc_char": 0.3002,
+      "acc_token": 0.6793,
+      "epoch": 0.02,
+      "loss_char": 0.7412,
+      "loss_token": 1.587,
+      "lr": "2.00e-07",
+      "norm": 5.6212,
+      "step": 1320
+    },
+    {
+      "acc_char": 0.3085,
+      "acc_token": 0.699,
+      "epoch": 0.02,
+      "loss_char": 0.6889,
+      "loss_token": 1.6217,
+      "lr": "2.00e-07",
+      "norm": 5.915,
+      "step": 1321
+    },
+    {
+      "acc_char": 0.3151,
+      "acc_token": 0.6983,
+      "epoch": 0.02,
+      "loss_char": 0.7024,
+      "loss_token": 1.581,
+      "lr": "2.00e-07",
+      "norm": 5.8306,
+      "step": 1322
+    },
+    {
+      "acc_char": 0.3029,
+      "acc_token": 0.6853,
+      "epoch": 0.02,
+      "loss_char": 0.7054,
+      "loss_token": 1.6941,
+      "lr": "2.00e-07",
+      "norm": 5.7394,
+      "step": 1323
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.706,
+      "epoch": 0.02,
+      "loss_char": 0.6812,
+      "loss_token": 1.5992,
+      "lr": "2.00e-07",
+      "norm": 5.8773,
+      "step": 1324
+    },
+    {
+      "acc_char": 0.3093,
+      "acc_token": 0.6971,
+      "epoch": 0.02,
+      "loss_char": 0.6921,
+      "loss_token": 1.5585,
+      "lr": "2.00e-07",
+      "norm": 5.6204,
+      "step": 1325
+    },
+    {
+      "acc_char": 0.2947,
+      "acc_token": 0.6644,
+      "epoch": 0.02,
+      "loss_char": 0.7743,
+      "loss_token": 1.6958,
+      "lr": "2.00e-07",
+      "norm": 5.856,
+      "step": 1326
+    },
+    {
+      "acc_char": 0.3062,
+      "acc_token": 0.6977,
+      "epoch": 0.02,
+      "loss_char": 0.6612,
+      "loss_token": 1.5455,
+      "lr": "2.00e-07",
+      "norm": 5.5659,
+      "step": 1327
+    },
+    {
+      "acc_char": 0.3122,
+      "acc_token": 0.6991,
+      "epoch": 0.02,
+      "loss_char": 0.6977,
+      "loss_token": 1.5435,
+      "lr": "2.00e-07",
+      "norm": 5.9666,
+      "step": 1328
+    },
+    {
+      "acc_char": 0.2953,
+      "acc_token": 0.6732,
+      "epoch": 0.02,
+      "loss_char": 0.7336,
+      "loss_token": 1.7345,
+      "lr": "2.00e-07",
+      "norm": 5.6002,
+      "step": 1329
+    },
+    {
+      "acc_char": 0.3132,
+      "acc_token": 0.7081,
+      "epoch": 0.02,
+      "loss_char": 0.6456,
+      "loss_token": 1.5861,
+      "lr": "2.00e-07",
+      "norm": 5.6949,
+      "step": 1330
+    },
+    {
+      "acc_char": 0.302,
+      "acc_token": 0.6938,
+      "epoch": 0.02,
+      "loss_char": 0.6581,
+      "loss_token": 1.5993,
+      "lr": "2.00e-07",
+      "norm": 6.3812,
+      "step": 1331
+    },
+    {
+      "acc_char": 0.3111,
+      "acc_token": 0.7023,
+      "epoch": 0.02,
+      "loss_char": 0.6768,
+      "loss_token": 1.5597,
+      "lr": "2.00e-07",
+      "norm": 5.5679,
+      "step": 1332
+    },
+    {
+      "acc_char": 0.3062,
+      "acc_token": 0.6841,
+      "epoch": 0.02,
+      "loss_char": 0.7264,
+      "loss_token": 1.6529,
+      "lr": "2.00e-07",
+      "norm": 5.9286,
+      "step": 1333
+    },
+    {
+      "acc_char": 0.3059,
+      "acc_token": 0.7084,
+      "epoch": 0.02,
+      "loss_char": 0.6333,
+      "loss_token": 1.492,
+      "lr": "2.00e-07",
+      "norm": 5.4609,
+      "step": 1334
+    },
+    {
+      "acc_char": 0.3002,
+      "acc_token": 0.6916,
+      "epoch": 0.02,
+      "loss_char": 0.666,
+      "loss_token": 1.5257,
+      "lr": "2.00e-07",
+      "norm": 5.5032,
+      "step": 1335
+    },
+    {
+      "acc_char": 0.3002,
+      "acc_token": 0.6851,
+      "epoch": 0.02,
+      "loss_char": 0.715,
+      "loss_token": 1.669,
+      "lr": "2.00e-07",
+      "norm": 5.89,
+      "step": 1336
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6808,
+      "epoch": 0.02,
+      "loss_char": 0.726,
+      "loss_token": 1.6113,
+      "lr": "2.00e-07",
+      "norm": 5.6423,
+      "step": 1337
+    },
+    {
+      "acc_char": 0.301,
+      "acc_token": 0.6771,
+      "epoch": 0.02,
+      "loss_char": 0.736,
+      "loss_token": 1.6483,
+      "lr": "2.00e-07",
+      "norm": 6.1082,
+      "step": 1338
+    },
+    {
+      "acc_char": 0.2952,
+      "acc_token": 0.6747,
+      "epoch": 0.02,
+      "loss_char": 0.769,
+      "loss_token": 1.6764,
+      "lr": "2.00e-07",
+      "norm": 5.9451,
+      "step": 1339
+    },
+    {
+      "acc_char": 0.2922,
+      "acc_token": 0.667,
+      "epoch": 0.02,
+      "loss_char": 0.7571,
+      "loss_token": 1.6656,
+      "lr": "2.00e-07",
+      "norm": 5.824,
+      "step": 1340
+    },
+    {
+      "acc_char": 0.2967,
+      "acc_token": 0.6644,
+      "epoch": 0.02,
+      "loss_char": 0.8147,
+      "loss_token": 1.7101,
+      "lr": "2.00e-07",
+      "norm": 5.9317,
+      "step": 1341
+    },
+    {
+      "acc_char": 0.2979,
+      "acc_token": 0.6884,
+      "epoch": 0.02,
+      "loss_char": 0.7192,
+      "loss_token": 1.6335,
+      "lr": "2.00e-07",
+      "norm": 5.9221,
+      "step": 1342
+    },
+    {
+      "acc_char": 0.3071,
+      "acc_token": 0.6797,
+      "epoch": 0.02,
+      "loss_char": 0.7358,
+      "loss_token": 1.6014,
+      "lr": "2.00e-07",
+      "norm": 5.824,
+      "step": 1343
+    },
+    {
+      "acc_char": 0.3119,
+      "acc_token": 0.6764,
+      "epoch": 0.02,
+      "loss_char": 0.7539,
+      "loss_token": 1.7648,
+      "lr": "2.00e-07",
+      "norm": 6.1267,
+      "step": 1344
+    },
+    {
+      "acc_char": 0.2983,
+      "acc_token": 0.6722,
+      "epoch": 0.02,
+      "loss_char": 0.7617,
+      "loss_token": 1.63,
+      "lr": "2.00e-07",
+      "norm": 5.8447,
+      "step": 1345
+    },
+    {
+      "acc_char": 0.3093,
+      "acc_token": 0.7008,
+      "epoch": 0.02,
+      "loss_char": 0.6767,
+      "loss_token": 1.5799,
+      "lr": "2.00e-07",
+      "norm": 5.5396,
+      "step": 1346
+    },
+    {
+      "acc_char": 0.2933,
+      "acc_token": 0.6702,
+      "epoch": 0.02,
+      "loss_char": 0.7454,
+      "loss_token": 1.6384,
+      "lr": "2.00e-07",
+      "norm": 5.709,
+      "step": 1347
+    },
+    {
+      "acc_char": 0.2914,
+      "acc_token": 0.6709,
+      "epoch": 0.02,
+      "loss_char": 0.7337,
+      "loss_token": 1.592,
+      "lr": "2.00e-07",
+      "norm": 5.3753,
+      "step": 1348
+    },
+    {
+      "acc_char": 0.3056,
+      "acc_token": 0.6938,
+      "epoch": 0.02,
+      "loss_char": 0.6961,
+      "loss_token": 1.6395,
+      "lr": "2.00e-07",
+      "norm": 5.6499,
+      "step": 1349
+    },
+    {
+      "acc_char": 0.3064,
+      "acc_token": 0.677,
+      "epoch": 0.02,
+      "loss_char": 0.7697,
+      "loss_token": 1.7278,
+      "lr": "2.00e-07",
+      "norm": 6.3553,
+      "step": 1350
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.669,
+      "epoch": 0.02,
+      "loss_char": 0.8016,
+      "loss_token": 1.6509,
+      "lr": "2.00e-07",
+      "norm": 6.1227,
+      "step": 1351
+    },
+    {
+      "acc_char": 0.2973,
+      "acc_token": 0.6812,
+      "epoch": 0.02,
+      "loss_char": 0.7068,
+      "loss_token": 1.6572,
+      "lr": "2.00e-07",
+      "norm": 5.6062,
+      "step": 1352
+    },
+    {
+      "acc_char": 0.3016,
+      "acc_token": 0.7348,
+      "epoch": 0.02,
+      "loss_char": 0.545,
+      "loss_token": 1.4678,
+      "lr": "2.00e-07",
+      "norm": 5.5256,
+      "step": 1353
+    },
+    {
+      "acc_char": 0.2859,
+      "acc_token": 0.6595,
+      "epoch": 0.02,
+      "loss_char": 0.7825,
+      "loss_token": 1.7206,
+      "lr": "2.00e-07",
+      "norm": 5.6409,
+      "step": 1354
+    },
+    {
+      "acc_char": 0.3046,
+      "acc_token": 0.6902,
+      "epoch": 0.02,
+      "loss_char": 0.696,
+      "loss_token": 1.6006,
+      "lr": "2.00e-07",
+      "norm": 6.1147,
+      "step": 1355
+    },
+    {
+      "acc_char": 0.3,
+      "acc_token": 0.666,
+      "epoch": 0.02,
+      "loss_char": 0.805,
+      "loss_token": 1.7478,
+      "lr": "2.00e-07",
+      "norm": 5.9633,
+      "step": 1356
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.6793,
+      "epoch": 0.02,
+      "loss_char": 0.7234,
+      "loss_token": 1.6125,
+      "lr": "2.00e-07",
+      "norm": 5.8055,
+      "step": 1357
+    },
+    {
+      "acc_char": 0.3124,
+      "acc_token": 0.704,
+      "epoch": 0.02,
+      "loss_char": 0.6558,
+      "loss_token": 1.5236,
+      "lr": "2.00e-07",
+      "norm": 5.401,
+      "step": 1358
+    },
+    {
+      "acc_char": 0.3125,
+      "acc_token": 0.7084,
+      "epoch": 0.02,
+      "loss_char": 0.652,
+      "loss_token": 1.5135,
+      "lr": "2.00e-07",
+      "norm": 5.7235,
+      "step": 1359
+    },
+    {
+      "acc_char": 0.3183,
+      "acc_token": 0.7207,
+      "epoch": 0.02,
+      "loss_char": 0.6271,
+      "loss_token": 1.503,
+      "lr": "2.00e-07",
+      "norm": 5.5718,
+      "step": 1360
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.6753,
+      "epoch": 0.02,
+      "loss_char": 0.7615,
+      "loss_token": 1.6756,
+      "lr": "2.00e-07",
+      "norm": 5.9743,
+      "step": 1361
+    },
+    {
+      "acc_char": 0.3221,
+      "acc_token": 0.7169,
+      "epoch": 0.02,
+      "loss_char": 0.6669,
+      "loss_token": 1.5328,
+      "lr": "2.00e-07",
+      "norm": 5.9282,
+      "step": 1362
+    },
+    {
+      "acc_char": 0.2839,
+      "acc_token": 0.6536,
+      "epoch": 0.02,
+      "loss_char": 0.7904,
+      "loss_token": 1.7676,
+      "lr": "2.00e-07",
+      "norm": 6.4369,
+      "step": 1363
+    },
+    {
+      "acc_char": 0.3128,
+      "acc_token": 0.6961,
+      "epoch": 0.02,
+      "loss_char": 0.6898,
+      "loss_token": 1.542,
+      "lr": "2.00e-07",
+      "norm": 5.5994,
+      "step": 1364
+    },
+    {
+      "acc_char": 0.2969,
+      "acc_token": 0.6742,
+      "epoch": 0.02,
+      "loss_char": 0.7332,
+      "loss_token": 1.6879,
+      "lr": "2.00e-07",
+      "norm": 6.1954,
+      "step": 1365
+    },
+    {
+      "acc_char": 0.2925,
+      "acc_token": 0.6585,
+      "epoch": 0.03,
+      "loss_char": 0.7898,
+      "loss_token": 1.7,
+      "lr": "2.00e-07",
+      "norm": 6.0185,
+      "step": 1366
+    },
+    {
+      "acc_char": 0.2989,
+      "acc_token": 0.6691,
+      "epoch": 0.03,
+      "loss_char": 0.756,
+      "loss_token": 1.6255,
+      "lr": "2.00e-07",
+      "norm": 6.3259,
+      "step": 1367
+    },
+    {
+      "acc_char": 0.3114,
+      "acc_token": 0.7018,
+      "epoch": 0.03,
+      "loss_char": 0.6807,
+      "loss_token": 1.669,
+      "lr": "2.00e-07",
+      "norm": 5.9932,
+      "step": 1368
+    },
+    {
+      "acc_char": 0.2956,
+      "acc_token": 0.6205,
+      "epoch": 0.03,
+      "loss_char": 0.7947,
+      "loss_token": 1.6081,
+      "lr": "2.00e-07",
+      "norm": 5.4512,
+      "step": 1369
+    },
+    {
+      "acc_char": 0.3231,
+      "acc_token": 0.7104,
+      "epoch": 0.03,
+      "loss_char": 0.672,
+      "loss_token": 1.5403,
+      "lr": "2.00e-07",
+      "norm": 6.2026,
+      "step": 1370
+    },
+    {
+      "acc_char": 0.3035,
+      "acc_token": 0.6891,
+      "epoch": 0.03,
+      "loss_char": 0.7086,
+      "loss_token": 1.5575,
+      "lr": "2.00e-07",
+      "norm": 5.5939,
+      "step": 1371
+    },
+    {
+      "acc_char": 0.3126,
+      "acc_token": 0.7002,
+      "epoch": 0.03,
+      "loss_char": 0.7236,
+      "loss_token": 1.6196,
+      "lr": "2.00e-07",
+      "norm": 5.4226,
+      "step": 1372
+    },
+    {
+      "acc_char": 0.3135,
+      "acc_token": 0.7063,
+      "epoch": 0.03,
+      "loss_char": 0.6697,
+      "loss_token": 1.4176,
+      "lr": "2.00e-07",
+      "norm": 5.5692,
+      "step": 1373
+    },
+    {
+      "acc_char": 0.3084,
+      "acc_token": 0.6858,
+      "epoch": 0.03,
+      "loss_char": 0.7254,
+      "loss_token": 1.6299,
+      "lr": "2.00e-07",
+      "norm": 5.6955,
+      "step": 1374
+    },
+    {
+      "acc_char": 0.2914,
+      "acc_token": 0.6656,
+      "epoch": 0.03,
+      "loss_char": 0.7616,
+      "loss_token": 1.6862,
+      "lr": "2.00e-07",
+      "norm": 6.1186,
+      "step": 1375
+    },
+    {
+      "acc_char": 0.303,
+      "acc_token": 0.6843,
+      "epoch": 0.03,
+      "loss_char": 0.7165,
+      "loss_token": 1.5999,
+      "lr": "2.00e-07",
+      "norm": 6.0434,
+      "step": 1376
+    },
+    {
+      "acc_char": 0.305,
+      "acc_token": 0.6946,
+      "epoch": 0.03,
+      "loss_char": 0.6948,
+      "loss_token": 1.5296,
+      "lr": "2.00e-07",
+      "norm": 5.7242,
+      "step": 1377
+    },
+    {
+      "acc_char": 0.312,
+      "acc_token": 0.6938,
+      "epoch": 0.03,
+      "loss_char": 0.717,
+      "loss_token": 1.5859,
+      "lr": "2.00e-07",
+      "norm": 5.6534,
+      "step": 1378
+    },
+    {
+      "acc_char": 0.2975,
+      "acc_token": 0.6662,
+      "epoch": 0.03,
+      "loss_char": 0.7601,
+      "loss_token": 1.638,
+      "lr": "2.00e-07",
+      "norm": 6.0556,
+      "step": 1379
+    },
+    {
+      "acc_char": 0.302,
+      "acc_token": 0.677,
+      "epoch": 0.03,
+      "loss_char": 0.7621,
+      "loss_token": 1.6764,
+      "lr": "2.00e-07",
+      "norm": 5.7392,
+      "step": 1380
+    },
+    {
+      "acc_char": 0.3123,
+      "acc_token": 0.7012,
+      "epoch": 0.03,
+      "loss_char": 0.6862,
+      "loss_token": 1.6282,
+      "lr": "2.00e-07",
+      "norm": 5.8201,
+      "step": 1381
+    },
+    {
+      "acc_char": 0.3136,
+      "acc_token": 0.7041,
+      "epoch": 0.03,
+      "loss_char": 0.6634,
+      "loss_token": 1.4637,
+      "lr": "2.00e-07",
+      "norm": 5.4737,
+      "step": 1382
+    },
+    {
+      "acc_char": 0.3047,
+      "acc_token": 0.6831,
+      "epoch": 0.03,
+      "loss_char": 0.7225,
+      "loss_token": 1.6879,
+      "lr": "2.00e-07",
+      "norm": 5.8436,
+      "step": 1383
+    },
+    {
+      "acc_char": 0.3059,
+      "acc_token": 0.6894,
+      "epoch": 0.03,
+      "loss_char": 0.7192,
+      "loss_token": 1.633,
+      "lr": "2.00e-07",
+      "norm": 5.9087,
+      "step": 1384
+    },
+    {
+      "acc_char": 0.3098,
+      "acc_token": 0.7056,
+      "epoch": 0.03,
+      "loss_char": 0.6613,
+      "loss_token": 1.4887,
+      "lr": "2.00e-07",
+      "norm": 5.5427,
+      "step": 1385
+    },
+    {
+      "acc_char": 0.2913,
+      "acc_token": 0.6533,
+      "epoch": 0.03,
+      "loss_char": 0.8128,
+      "loss_token": 1.6987,
+      "lr": "2.00e-07",
+      "norm": 5.9285,
+      "step": 1386
+    },
+    {
+      "acc_char": 0.306,
+      "acc_token": 0.6957,
+      "epoch": 0.03,
+      "loss_char": 0.6774,
+      "loss_token": 1.5901,
+      "lr": "2.00e-07",
+      "norm": 5.6079,
+      "step": 1387
+    },
+    {
+      "acc_char": 0.3,
+      "acc_token": 0.6777,
+      "epoch": 0.03,
+      "loss_char": 0.7392,
+      "loss_token": 1.6184,
+      "lr": "2.00e-07",
+      "norm": 5.5041,
+      "step": 1388
+    },
+    {
+      "acc_char": 0.3064,
+      "acc_token": 0.6982,
+      "epoch": 0.03,
+      "loss_char": 0.6812,
+      "loss_token": 1.6384,
+      "lr": "2.00e-07",
+      "norm": 6.0128,
+      "step": 1389
+    },
+    {
+      "acc_char": 0.2931,
+      "acc_token": 0.6757,
+      "epoch": 0.03,
+      "loss_char": 0.7242,
+      "loss_token": 1.5323,
+      "lr": "2.00e-07",
+      "norm": 5.6286,
+      "step": 1390
+    },
+    {
+      "acc_char": 0.3113,
+      "acc_token": 0.6916,
+      "epoch": 0.03,
+      "loss_char": 0.7159,
+      "loss_token": 1.6158,
+      "lr": "2.00e-07",
+      "norm": 5.7847,
+      "step": 1391
+    },
+    {
+      "acc_char": 0.3085,
+      "acc_token": 0.7056,
+      "epoch": 0.03,
+      "loss_char": 0.6436,
+      "loss_token": 1.476,
+      "lr": "2.00e-07",
+      "norm": 5.8416,
+      "step": 1392
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6734,
+      "epoch": 0.03,
+      "loss_char": 0.7523,
+      "loss_token": 1.5721,
+      "lr": "2.00e-07",
+      "norm": 5.6675,
+      "step": 1393
+    },
+    {
+      "acc_char": 0.3048,
+      "acc_token": 0.6848,
+      "epoch": 0.03,
+      "loss_char": 0.7287,
+      "loss_token": 1.5632,
+      "lr": "2.00e-07",
+      "norm": 5.4626,
+      "step": 1394
+    },
+    {
+      "acc_char": 0.3077,
+      "acc_token": 0.6963,
+      "epoch": 0.03,
+      "loss_char": 0.6829,
+      "loss_token": 1.667,
+      "lr": "2.00e-07",
+      "norm": 5.6939,
+      "step": 1395
+    },
+    {
+      "acc_char": 0.3158,
+      "acc_token": 0.7124,
+      "epoch": 0.03,
+      "loss_char": 0.6541,
+      "loss_token": 1.6637,
+      "lr": "2.00e-07",
+      "norm": 5.609,
+      "step": 1396
+    },
+    {
+      "acc_char": 0.3019,
+      "acc_token": 0.6927,
+      "epoch": 0.03,
+      "loss_char": 0.6862,
+      "loss_token": 1.5703,
+      "lr": "2.00e-07",
+      "norm": 5.6961,
+      "step": 1397
+    },
+    {
+      "acc_char": 0.3018,
+      "acc_token": 0.6779,
+      "epoch": 0.03,
+      "loss_char": 0.723,
+      "loss_token": 1.6526,
+      "lr": "2.00e-07",
+      "norm": 5.8147,
+      "step": 1398
+    },
+    {
+      "acc_char": 0.2908,
+      "acc_token": 0.6306,
+      "epoch": 0.03,
+      "loss_char": 0.8343,
+      "loss_token": 1.6166,
+      "lr": "2.00e-07",
+      "norm": 5.3458,
+      "step": 1399
+    },
+    {
+      "acc_char": 0.3062,
+      "acc_token": 0.7036,
+      "epoch": 0.03,
+      "loss_char": 0.6465,
+      "loss_token": 1.5848,
+      "lr": "2.00e-07",
+      "norm": 5.6703,
+      "step": 1400
+    },
+    {
+      "acc_char": 0.3115,
+      "acc_token": 0.7026,
+      "epoch": 0.03,
+      "loss_char": 0.6719,
+      "loss_token": 1.5331,
+      "lr": "2.00e-07",
+      "norm": 5.415,
+      "step": 1401
+    },
+    {
+      "acc_char": 0.305,
+      "acc_token": 0.6994,
+      "epoch": 0.03,
+      "loss_char": 0.6684,
+      "loss_token": 1.6298,
+      "lr": "2.00e-07",
+      "norm": 5.7916,
+      "step": 1402
+    },
+    {
+      "acc_char": 0.3037,
+      "acc_token": 0.6807,
+      "epoch": 0.03,
+      "loss_char": 0.746,
+      "loss_token": 1.5851,
+      "lr": "2.00e-07",
+      "norm": 5.7043,
+      "step": 1403
+    },
+    {
+      "acc_char": 0.3037,
+      "acc_token": 0.6985,
+      "epoch": 0.03,
+      "loss_char": 0.6662,
+      "loss_token": 1.5652,
+      "lr": "2.00e-07",
+      "norm": 5.7722,
+      "step": 1404
+    },
+    {
+      "acc_char": 0.3007,
+      "acc_token": 0.6844,
+      "epoch": 0.03,
+      "loss_char": 0.7001,
+      "loss_token": 1.5728,
+      "lr": "2.00e-07",
+      "norm": 5.6028,
+      "step": 1405
+    },
+    {
+      "acc_char": 0.292,
+      "acc_token": 0.6601,
+      "epoch": 0.03,
+      "loss_char": 0.8013,
+      "loss_token": 1.6926,
+      "lr": "2.00e-07",
+      "norm": 6.0242,
+      "step": 1406
+    },
+    {
+      "acc_char": 0.3058,
+      "acc_token": 0.698,
+      "epoch": 0.03,
+      "loss_char": 0.6768,
+      "loss_token": 1.5712,
+      "lr": "2.00e-07",
+      "norm": 5.611,
+      "step": 1407
+    },
+    {
+      "acc_char": 0.3038,
+      "acc_token": 0.6904,
+      "epoch": 0.03,
+      "loss_char": 0.6899,
+      "loss_token": 1.624,
+      "lr": "2.00e-07",
+      "norm": 6.0308,
+      "step": 1408
+    },
+    {
+      "acc_char": 0.2939,
+      "acc_token": 0.669,
+      "epoch": 0.03,
+      "loss_char": 0.748,
+      "loss_token": 1.6023,
+      "lr": "2.00e-07",
+      "norm": 5.6687,
+      "step": 1409
+    },
+    {
+      "acc_char": 0.2921,
+      "acc_token": 0.669,
+      "epoch": 0.03,
+      "loss_char": 0.7351,
+      "loss_token": 1.6203,
+      "lr": "2.00e-07",
+      "norm": 5.5791,
+      "step": 1410
+    },
+    {
+      "acc_char": 0.2854,
+      "acc_token": 0.6869,
+      "epoch": 0.03,
+      "loss_char": 0.6622,
+      "loss_token": 1.6635,
+      "lr": "2.00e-07",
+      "norm": 7.0393,
+      "step": 1411
+    },
+    {
+      "acc_char": 0.3082,
+      "acc_token": 0.6844,
+      "epoch": 0.03,
+      "loss_char": 0.7358,
+      "loss_token": 1.6045,
+      "lr": "2.00e-07",
+      "norm": 5.7435,
+      "step": 1412
+    },
+    {
+      "acc_char": 0.3065,
+      "acc_token": 0.7006,
+      "epoch": 0.03,
+      "loss_char": 0.6737,
+      "loss_token": 1.5908,
+      "lr": "2.00e-07",
+      "norm": 5.6507,
+      "step": 1413
+    },
+    {
+      "acc_char": 0.3041,
+      "acc_token": 0.6997,
+      "epoch": 0.03,
+      "loss_char": 0.6748,
+      "loss_token": 1.5888,
+      "lr": "2.00e-07",
+      "norm": 5.6674,
+      "step": 1414
+    },
+    {
+      "acc_char": 0.3062,
+      "acc_token": 0.6914,
+      "epoch": 0.03,
+      "loss_char": 0.7032,
+      "loss_token": 1.5616,
+      "lr": "2.00e-07",
+      "norm": 5.8995,
+      "step": 1415
+    },
+    {
+      "acc_char": 0.291,
+      "acc_token": 0.6312,
+      "epoch": 0.03,
+      "loss_char": 0.86,
+      "loss_token": 1.7505,
+      "lr": "2.00e-07",
+      "norm": 5.5878,
+      "step": 1416
+    },
+    {
+      "acc_char": 0.3097,
+      "acc_token": 0.6963,
+      "epoch": 0.03,
+      "loss_char": 0.672,
+      "loss_token": 1.5904,
+      "lr": "2.00e-07",
+      "norm": 5.6347,
+      "step": 1417
+    },
+    {
+      "acc_char": 0.3116,
+      "acc_token": 0.6902,
+      "epoch": 0.03,
+      "loss_char": 0.7182,
+      "loss_token": 1.572,
+      "lr": "2.00e-07",
+      "norm": 5.7473,
+      "step": 1418
+    },
+    {
+      "acc_char": 0.3138,
+      "acc_token": 0.7077,
+      "epoch": 0.03,
+      "loss_char": 0.677,
+      "loss_token": 1.5683,
+      "lr": "2.00e-07",
+      "norm": 5.97,
+      "step": 1419
+    },
+    {
+      "acc_char": 0.3232,
+      "acc_token": 0.7237,
+      "epoch": 0.03,
+      "loss_char": 0.6442,
+      "loss_token": 1.585,
+      "lr": "2.00e-07",
+      "norm": 6.0055,
+      "step": 1420
+    },
+    {
+      "acc_char": 0.3097,
+      "acc_token": 0.6964,
+      "epoch": 0.03,
+      "loss_char": 0.6874,
+      "loss_token": 1.5741,
+      "lr": "2.00e-07",
+      "norm": 5.5152,
+      "step": 1421
+    },
+    {
+      "acc_char": 0.3156,
+      "acc_token": 0.7126,
+      "epoch": 0.03,
+      "loss_char": 0.6461,
+      "loss_token": 1.4811,
+      "lr": "2.00e-07",
+      "norm": 5.4544,
+      "step": 1422
+    },
+    {
+      "acc_char": 0.3079,
+      "acc_token": 0.704,
+      "epoch": 0.03,
+      "loss_char": 0.6918,
+      "loss_token": 1.6122,
+      "lr": "2.00e-07",
+      "norm": 5.8757,
+      "step": 1423
+    },
+    {
+      "acc_char": 0.3054,
+      "acc_token": 0.6862,
+      "epoch": 0.03,
+      "loss_char": 0.7286,
+      "loss_token": 1.595,
+      "lr": "2.00e-07",
+      "norm": 5.8165,
+      "step": 1424
+    },
+    {
+      "acc_char": 0.3133,
+      "acc_token": 0.6962,
+      "epoch": 0.03,
+      "loss_char": 0.6887,
+      "loss_token": 1.6077,
+      "lr": "2.00e-07",
+      "norm": 6.1779,
+      "step": 1425
+    },
+    {
+      "acc_char": 0.312,
+      "acc_token": 0.7053,
+      "epoch": 0.03,
+      "loss_char": 0.6846,
+      "loss_token": 1.5866,
+      "lr": "2.00e-07",
+      "norm": 5.7713,
+      "step": 1426
+    },
+    {
+      "acc_char": 0.2996,
+      "acc_token": 0.6674,
+      "epoch": 0.03,
+      "loss_char": 0.787,
+      "loss_token": 1.6678,
+      "lr": "2.00e-07",
+      "norm": 5.5794,
+      "step": 1427
+    },
+    {
+      "acc_char": 0.2895,
+      "acc_token": 0.6244,
+      "epoch": 0.03,
+      "loss_char": 0.8862,
+      "loss_token": 1.8093,
+      "lr": "2.00e-07",
+      "norm": 5.699,
+      "step": 1428
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6784,
+      "epoch": 0.03,
+      "loss_char": 0.7447,
+      "loss_token": 1.6302,
+      "lr": "2.00e-07",
+      "norm": 5.6967,
+      "step": 1429
+    },
+    {
+      "acc_char": 0.3022,
+      "acc_token": 0.6747,
+      "epoch": 0.03,
+      "loss_char": 0.7457,
+      "loss_token": 1.6382,
+      "lr": "2.00e-07",
+      "norm": 5.7382,
+      "step": 1430
+    },
+    {
+      "acc_char": 0.2978,
+      "acc_token": 0.6589,
+      "epoch": 0.03,
+      "loss_char": 0.8084,
+      "loss_token": 1.668,
+      "lr": "2.00e-07",
+      "norm": 6.0445,
+      "step": 1431
+    },
+    {
+      "acc_char": 0.3157,
+      "acc_token": 0.7066,
+      "epoch": 0.03,
+      "loss_char": 0.6527,
+      "loss_token": 1.5359,
+      "lr": "2.00e-07",
+      "norm": 5.8243,
+      "step": 1432
+    },
+    {
+      "acc_char": 0.312,
+      "acc_token": 0.6989,
+      "epoch": 0.03,
+      "loss_char": 0.6841,
+      "loss_token": 1.4395,
+      "lr": "2.00e-07",
+      "norm": 5.8681,
+      "step": 1433
+    },
+    {
+      "acc_char": 0.3123,
+      "acc_token": 0.7054,
+      "epoch": 0.03,
+      "loss_char": 0.6706,
+      "loss_token": 1.5895,
+      "lr": "2.00e-07",
+      "norm": 6.5931,
+      "step": 1434
+    },
+    {
+      "acc_char": 0.3129,
+      "acc_token": 0.7109,
+      "epoch": 0.03,
+      "loss_char": 0.642,
+      "loss_token": 1.5843,
+      "lr": "2.00e-07",
+      "norm": 5.7947,
+      "step": 1435
+    },
+    {
+      "acc_char": 0.3108,
+      "acc_token": 0.7093,
+      "epoch": 0.03,
+      "loss_char": 0.646,
+      "loss_token": 1.552,
+      "lr": "2.00e-07",
+      "norm": 5.4622,
+      "step": 1436
+    },
+    {
+      "acc_char": 0.2952,
+      "acc_token": 0.6775,
+      "epoch": 0.03,
+      "loss_char": 0.7064,
+      "loss_token": 1.6145,
+      "lr": "2.00e-07",
+      "norm": 5.5288,
+      "step": 1437
+    },
+    {
+      "acc_char": 0.3134,
+      "acc_token": 0.6921,
+      "epoch": 0.03,
+      "loss_char": 0.6998,
+      "loss_token": 1.5238,
+      "lr": "2.00e-07",
+      "norm": 5.704,
+      "step": 1438
+    },
+    {
+      "acc_char": 0.2999,
+      "acc_token": 0.6891,
+      "epoch": 0.03,
+      "loss_char": 0.6862,
+      "loss_token": 1.6609,
+      "lr": "2.00e-07",
+      "norm": 5.6822,
+      "step": 1439
+    },
+    {
+      "acc_char": 0.3086,
+      "acc_token": 0.7125,
+      "epoch": 0.03,
+      "loss_char": 0.6314,
+      "loss_token": 1.5158,
+      "lr": "2.00e-07",
+      "norm": 5.4196,
+      "step": 1440
+    },
+    {
+      "acc_char": 0.3088,
+      "acc_token": 0.7033,
+      "epoch": 0.03,
+      "loss_char": 0.6497,
+      "loss_token": 1.5208,
+      "lr": "2.00e-07",
+      "norm": 5.9641,
+      "step": 1441
+    },
+    {
+      "acc_char": 0.3007,
+      "acc_token": 0.6797,
+      "epoch": 0.03,
+      "loss_char": 0.733,
+      "loss_token": 1.6274,
+      "lr": "2.00e-07",
+      "norm": 5.7354,
+      "step": 1442
+    },
+    {
+      "acc_char": 0.3165,
+      "acc_token": 0.7008,
+      "epoch": 0.03,
+      "loss_char": 0.6918,
+      "loss_token": 1.591,
+      "lr": "2.00e-07",
+      "norm": 5.808,
+      "step": 1443
+    },
+    {
+      "acc_char": 0.3011,
+      "acc_token": 0.674,
+      "epoch": 0.03,
+      "loss_char": 0.7647,
+      "loss_token": 1.7132,
+      "lr": "2.00e-07",
+      "norm": 5.7308,
+      "step": 1444
+    },
+    {
+      "acc_char": 0.3117,
+      "acc_token": 0.7057,
+      "epoch": 0.03,
+      "loss_char": 0.6518,
+      "loss_token": 1.6234,
+      "lr": "2.00e-07",
+      "norm": 5.8338,
+      "step": 1445
+    },
+    {
+      "acc_char": 0.3148,
+      "acc_token": 0.7049,
+      "epoch": 0.03,
+      "loss_char": 0.7024,
+      "loss_token": 1.5629,
+      "lr": "2.00e-07",
+      "norm": 5.7902,
+      "step": 1446
+    },
+    {
+      "acc_char": 0.3071,
+      "acc_token": 0.6901,
+      "epoch": 0.03,
+      "loss_char": 0.699,
+      "loss_token": 1.6437,
+      "lr": "2.00e-07",
+      "norm": 5.9924,
+      "step": 1447
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6833,
+      "epoch": 0.03,
+      "loss_char": 0.7037,
+      "loss_token": 1.5695,
+      "lr": "2.00e-07",
+      "norm": 5.6272,
+      "step": 1448
+    },
+    {
+      "acc_char": 0.2949,
+      "acc_token": 0.6685,
+      "epoch": 0.03,
+      "loss_char": 0.739,
+      "loss_token": 1.7014,
+      "lr": "2.00e-07",
+      "norm": 5.7195,
+      "step": 1449
+    },
+    {
+      "acc_char": 0.2872,
+      "acc_token": 0.6562,
+      "epoch": 0.03,
+      "loss_char": 0.7832,
+      "loss_token": 1.7406,
+      "lr": "2.00e-07",
+      "norm": 5.8079,
+      "step": 1450
+    },
+    {
+      "acc_char": 0.318,
+      "acc_token": 0.7229,
+      "epoch": 0.03,
+      "loss_char": 0.6029,
+      "loss_token": 1.4753,
+      "lr": "2.00e-07",
+      "norm": 5.6095,
+      "step": 1451
+    },
+    {
+      "acc_char": 0.3115,
+      "acc_token": 0.6985,
+      "epoch": 0.03,
+      "loss_char": 0.6745,
+      "loss_token": 1.5394,
+      "lr": "2.00e-07",
+      "norm": 5.5028,
+      "step": 1452
+    },
+    {
+      "acc_char": 0.2884,
+      "acc_token": 0.6616,
+      "epoch": 0.03,
+      "loss_char": 0.7668,
+      "loss_token": 1.6874,
+      "lr": "2.00e-07",
+      "norm": 5.9784,
+      "step": 1453
+    },
+    {
+      "acc_char": 0.3044,
+      "acc_token": 0.6927,
+      "epoch": 0.03,
+      "loss_char": 0.6746,
+      "loss_token": 1.6322,
+      "lr": "2.00e-07",
+      "norm": 5.9423,
+      "step": 1454
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.6784,
+      "epoch": 0.03,
+      "loss_char": 0.7432,
+      "loss_token": 1.5928,
+      "lr": "2.00e-07",
+      "norm": 5.9516,
+      "step": 1455
+    },
+    {
+      "acc_char": 0.3164,
+      "acc_token": 0.7029,
+      "epoch": 0.03,
+      "loss_char": 0.6968,
+      "loss_token": 1.5889,
+      "lr": "2.00e-07",
+      "norm": 5.5862,
+      "step": 1456
+    },
+    {
+      "acc_char": 0.3036,
+      "acc_token": 0.6886,
+      "epoch": 0.03,
+      "loss_char": 0.7043,
+      "loss_token": 1.7186,
+      "lr": "2.00e-07",
+      "norm": 5.8658,
+      "step": 1457
+    },
+    {
+      "acc_char": 0.3089,
+      "acc_token": 0.6924,
+      "epoch": 0.03,
+      "loss_char": 0.697,
+      "loss_token": 1.5579,
+      "lr": "2.00e-07",
+      "norm": 5.526,
+      "step": 1458
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6933,
+      "epoch": 0.03,
+      "loss_char": 0.6822,
+      "loss_token": 1.5683,
+      "lr": "2.00e-07",
+      "norm": 5.565,
+      "step": 1459
+    },
+    {
+      "acc_char": 0.3016,
+      "acc_token": 0.6906,
+      "epoch": 0.03,
+      "loss_char": 0.6766,
+      "loss_token": 1.6086,
+      "lr": "2.00e-07",
+      "norm": 5.7796,
+      "step": 1460
+    },
+    {
+      "acc_char": 0.2909,
+      "acc_token": 0.6303,
+      "epoch": 0.03,
+      "loss_char": 0.8754,
+      "loss_token": 1.7548,
+      "lr": "2.00e-07",
+      "norm": 5.8089,
+      "step": 1461
+    },
+    {
+      "acc_char": 0.3114,
+      "acc_token": 0.6996,
+      "epoch": 0.03,
+      "loss_char": 0.6875,
+      "loss_token": 1.634,
+      "lr": "2.00e-07",
+      "norm": 5.7081,
+      "step": 1462
+    },
+    {
+      "acc_char": 0.3136,
+      "acc_token": 0.6977,
+      "epoch": 0.03,
+      "loss_char": 0.7016,
+      "loss_token": 1.6071,
+      "lr": "2.00e-07",
+      "norm": 5.6163,
+      "step": 1463
+    },
+    {
+      "acc_char": 0.3071,
+      "acc_token": 0.6808,
+      "epoch": 0.03,
+      "loss_char": 0.7488,
+      "loss_token": 1.6655,
+      "lr": "2.00e-07",
+      "norm": 5.9807,
+      "step": 1464
+    },
+    {
+      "acc_char": 0.292,
+      "acc_token": 0.6687,
+      "epoch": 0.03,
+      "loss_char": 0.7498,
+      "loss_token": 1.6668,
+      "lr": "2.00e-07",
+      "norm": 5.646,
+      "step": 1465
+    },
+    {
+      "acc_char": 0.3112,
+      "acc_token": 0.6972,
+      "epoch": 0.03,
+      "loss_char": 0.6942,
+      "loss_token": 1.6347,
+      "lr": "2.00e-07",
+      "norm": 5.8645,
+      "step": 1466
+    },
+    {
+      "acc_char": 0.3131,
+      "acc_token": 0.709,
+      "epoch": 0.03,
+      "loss_char": 0.6439,
+      "loss_token": 1.5205,
+      "lr": "2.00e-07",
+      "norm": 5.8664,
+      "step": 1467
+    },
+    {
+      "acc_char": 0.3017,
+      "acc_token": 0.6935,
+      "epoch": 0.03,
+      "loss_char": 0.6633,
+      "loss_token": 1.5259,
+      "lr": "2.00e-07",
+      "norm": 5.7257,
+      "step": 1468
+    },
+    {
+      "acc_char": 0.3088,
+      "acc_token": 0.7001,
+      "epoch": 0.03,
+      "loss_char": 0.679,
+      "loss_token": 1.5258,
+      "lr": "2.00e-07",
+      "norm": 5.5526,
+      "step": 1469
+    },
+    {
+      "acc_char": 0.3113,
+      "acc_token": 0.6895,
+      "epoch": 0.03,
+      "loss_char": 0.7152,
+      "loss_token": 1.5858,
+      "lr": "2.00e-07",
+      "norm": 5.8972,
+      "step": 1470
+    },
+    {
+      "acc_char": 0.3095,
+      "acc_token": 0.6881,
+      "epoch": 0.03,
+      "loss_char": 0.7051,
+      "loss_token": 1.6035,
+      "lr": "2.00e-07",
+      "norm": 5.7331,
+      "step": 1471
+    },
+    {
+      "acc_char": 0.2944,
+      "acc_token": 0.6664,
+      "epoch": 0.03,
+      "loss_char": 0.7404,
+      "loss_token": 1.5585,
+      "lr": "2.00e-07",
+      "norm": 5.8637,
+      "step": 1472
+    },
+    {
+      "acc_char": 0.2941,
+      "acc_token": 0.681,
+      "epoch": 0.03,
+      "loss_char": 0.721,
+      "loss_token": 1.61,
+      "lr": "2.00e-07",
+      "norm": 5.8771,
+      "step": 1473
+    },
+    {
+      "acc_char": 0.3011,
+      "acc_token": 0.6886,
+      "epoch": 0.03,
+      "loss_char": 0.7031,
+      "loss_token": 1.7011,
+      "lr": "2.00e-07",
+      "norm": 5.6953,
+      "step": 1474
+    },
+    {
+      "acc_char": 0.2966,
+      "acc_token": 0.684,
+      "epoch": 0.03,
+      "loss_char": 0.6892,
+      "loss_token": 1.5817,
+      "lr": "2.00e-07",
+      "norm": 5.8247,
+      "step": 1475
+    },
+    {
+      "acc_char": 0.2882,
+      "acc_token": 0.6594,
+      "epoch": 0.03,
+      "loss_char": 0.7986,
+      "loss_token": 1.7027,
+      "lr": "2.00e-07",
+      "norm": 5.6554,
+      "step": 1476
+    },
+    {
+      "acc_char": 0.2976,
+      "acc_token": 0.6816,
+      "epoch": 0.03,
+      "loss_char": 0.7055,
+      "loss_token": 1.5926,
+      "lr": "2.00e-07",
+      "norm": 5.6155,
+      "step": 1477
+    },
+    {
+      "acc_char": 0.3091,
+      "acc_token": 0.6927,
+      "epoch": 0.03,
+      "loss_char": 0.6936,
+      "loss_token": 1.6087,
+      "lr": "2.00e-07",
+      "norm": 5.8617,
+      "step": 1478
+    },
+    {
+      "acc_char": 0.3095,
+      "acc_token": 0.6873,
+      "epoch": 0.03,
+      "loss_char": 0.7059,
+      "loss_token": 1.583,
+      "lr": "2.00e-07",
+      "norm": 5.9507,
+      "step": 1479
+    },
+    {
+      "acc_char": 0.2857,
+      "acc_token": 0.611,
+      "epoch": 0.03,
+      "loss_char": 0.9413,
+      "loss_token": 1.7928,
+      "lr": "2.00e-07",
+      "norm": 5.3356,
+      "step": 1480
+    },
+    {
+      "acc_char": 0.315,
+      "acc_token": 0.706,
+      "epoch": 0.03,
+      "loss_char": 0.6653,
+      "loss_token": 1.557,
+      "lr": "2.00e-07",
+      "norm": 5.6644,
+      "step": 1481
+    },
+    {
+      "acc_char": 0.3029,
+      "acc_token": 0.6934,
+      "epoch": 0.03,
+      "loss_char": 0.7113,
+      "loss_token": 1.6224,
+      "lr": "2.00e-07",
+      "norm": 5.8906,
+      "step": 1482
+    },
+    {
+      "acc_char": 0.3124,
+      "acc_token": 0.7088,
+      "epoch": 0.03,
+      "loss_char": 0.6479,
+      "loss_token": 1.44,
+      "lr": "2.00e-07",
+      "norm": 5.4433,
+      "step": 1483
+    },
+    {
+      "acc_char": 0.2985,
+      "acc_token": 0.6853,
+      "epoch": 0.03,
+      "loss_char": 0.6856,
+      "loss_token": 1.5717,
+      "lr": "2.00e-07",
+      "norm": 5.6576,
+      "step": 1484
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6793,
+      "epoch": 0.03,
+      "loss_char": 0.7508,
+      "loss_token": 1.5706,
+      "lr": "2.00e-07",
+      "norm": 5.9455,
+      "step": 1485
+    },
+    {
+      "acc_char": 0.3116,
+      "acc_token": 0.7013,
+      "epoch": 0.03,
+      "loss_char": 0.6822,
+      "loss_token": 1.5799,
+      "lr": "2.00e-07",
+      "norm": 5.6926,
+      "step": 1486
+    },
+    {
+      "acc_char": 0.3101,
+      "acc_token": 0.6957,
+      "epoch": 0.03,
+      "loss_char": 0.6928,
+      "loss_token": 1.5266,
+      "lr": "2.00e-07",
+      "norm": 5.6095,
+      "step": 1487
+    },
+    {
+      "acc_char": 0.2985,
+      "acc_token": 0.679,
+      "epoch": 0.03,
+      "loss_char": 0.7325,
+      "loss_token": 1.7235,
+      "lr": "2.00e-07",
+      "norm": 6.2878,
+      "step": 1488
+    },
+    {
+      "acc_char": 0.3033,
+      "acc_token": 0.6928,
+      "epoch": 0.03,
+      "loss_char": 0.6938,
+      "loss_token": 1.5439,
+      "lr": "2.00e-07",
+      "norm": 5.8769,
+      "step": 1489
+    },
+    {
+      "acc_char": 0.2956,
+      "acc_token": 0.6612,
+      "epoch": 0.03,
+      "loss_char": 0.8068,
+      "loss_token": 1.7251,
+      "lr": "2.00e-07",
+      "norm": 5.9183,
+      "step": 1490
+    },
+    {
+      "acc_char": 0.3088,
+      "acc_token": 0.6941,
+      "epoch": 0.03,
+      "loss_char": 0.691,
+      "loss_token": 1.5389,
+      "lr": "2.00e-07",
+      "norm": 5.8215,
+      "step": 1491
+    },
+    {
+      "acc_char": 0.3048,
+      "acc_token": 0.6917,
+      "epoch": 0.03,
+      "loss_char": 0.6979,
+      "loss_token": 1.6155,
+      "lr": "2.00e-07",
+      "norm": 5.684,
+      "step": 1492
+    },
+    {
+      "acc_char": 0.3073,
+      "acc_token": 0.6831,
+      "epoch": 0.03,
+      "loss_char": 0.7344,
+      "loss_token": 1.6261,
+      "lr": "2.00e-07",
+      "norm": 5.7121,
+      "step": 1493
+    },
+    {
+      "acc_char": 0.2876,
+      "acc_token": 0.6484,
+      "epoch": 0.03,
+      "loss_char": 0.8171,
+      "loss_token": 1.7074,
+      "lr": "2.00e-07",
+      "norm": 5.8721,
+      "step": 1494
+    },
+    {
+      "acc_char": 0.3122,
+      "acc_token": 0.7034,
+      "epoch": 0.03,
+      "loss_char": 0.6991,
+      "loss_token": 1.5556,
+      "lr": "2.00e-07",
+      "norm": 5.6013,
+      "step": 1495
+    },
+    {
+      "acc_char": 0.3105,
+      "acc_token": 0.6988,
+      "epoch": 0.03,
+      "loss_char": 0.6859,
+      "loss_token": 1.5507,
+      "lr": "2.00e-07",
+      "norm": 5.7857,
+      "step": 1496
+    },
+    {
+      "acc_char": 0.3001,
+      "acc_token": 0.6766,
+      "epoch": 0.03,
+      "loss_char": 0.747,
+      "loss_token": 1.6438,
+      "lr": "2.00e-07",
+      "norm": 5.6768,
+      "step": 1497
+    },
+    {
+      "acc_char": 0.3066,
+      "acc_token": 0.7027,
+      "epoch": 0.03,
+      "loss_char": 0.6664,
+      "loss_token": 1.6072,
+      "lr": "2.00e-07",
+      "norm": 5.7578,
+      "step": 1498
+    },
+    {
+      "acc_char": 0.2913,
+      "acc_token": 0.67,
+      "epoch": 0.03,
+      "loss_char": 0.7487,
+      "loss_token": 1.7215,
+      "lr": "2.00e-07",
+      "norm": 6.554,
+      "step": 1499
+    },
+    {
+      "acc_char": 0.292,
+      "acc_token": 0.6318,
+      "epoch": 0.03,
+      "loss_char": 0.8849,
+      "loss_token": 1.7365,
+      "lr": "2.00e-07",
+      "norm": 5.7328,
+      "step": 1500
+    },
+    {
+      "acc_char": 0.295,
+      "acc_token": 0.6673,
+      "epoch": 0.03,
+      "loss_char": 0.7515,
+      "loss_token": 1.6953,
+      "lr": "2.00e-07",
+      "norm": 5.9586,
+      "step": 1501
+    },
+    {
+      "acc_char": 0.3103,
+      "acc_token": 0.7079,
+      "epoch": 0.03,
+      "loss_char": 0.6526,
+      "loss_token": 1.4509,
+      "lr": "2.00e-07",
+      "norm": 5.6547,
+      "step": 1502
+    },
+    {
+      "acc_char": 0.3149,
+      "acc_token": 0.7111,
+      "epoch": 0.03,
+      "loss_char": 0.6496,
+      "loss_token": 1.5897,
+      "lr": "2.00e-07",
+      "norm": 5.7801,
+      "step": 1503
+    },
+    {
+      "acc_char": 0.2957,
+      "acc_token": 0.6702,
+      "epoch": 0.03,
+      "loss_char": 0.7507,
+      "loss_token": 1.6831,
+      "lr": "2.00e-07",
+      "norm": 5.8254,
+      "step": 1504
+    },
+    {
+      "acc_char": 0.3132,
+      "acc_token": 0.7008,
+      "epoch": 0.03,
+      "loss_char": 0.682,
+      "loss_token": 1.5586,
+      "lr": "2.00e-07",
+      "norm": 5.5422,
+      "step": 1505
+    },
+    {
+      "acc_char": 0.2954,
+      "acc_token": 0.6786,
+      "epoch": 0.03,
+      "loss_char": 0.7245,
+      "loss_token": 1.6014,
+      "lr": "2.00e-07",
+      "norm": 5.6137,
+      "step": 1506
+    },
+    {
+      "acc_char": 0.3009,
+      "acc_token": 0.6757,
+      "epoch": 0.03,
+      "loss_char": 0.7528,
+      "loss_token": 1.6957,
+      "lr": "2.00e-07",
+      "norm": 6.0229,
+      "step": 1507
+    },
+    {
+      "acc_char": 0.3042,
+      "acc_token": 0.6788,
+      "epoch": 0.03,
+      "loss_char": 0.7415,
+      "loss_token": 1.6613,
+      "lr": "2.00e-07",
+      "norm": 5.6879,
+      "step": 1508
+    },
+    {
+      "acc_char": 0.3037,
+      "acc_token": 0.6762,
+      "epoch": 0.03,
+      "loss_char": 0.762,
+      "loss_token": 1.6386,
+      "lr": "2.00e-07",
+      "norm": 5.9348,
+      "step": 1509
+    },
+    {
+      "acc_char": 0.3078,
+      "acc_token": 0.69,
+      "epoch": 0.03,
+      "loss_char": 0.7001,
+      "loss_token": 1.5491,
+      "lr": "2.00e-07",
+      "norm": 6.2251,
+      "step": 1510
+    },
+    {
+      "acc_char": 0.3051,
+      "acc_token": 0.6926,
+      "epoch": 0.03,
+      "loss_char": 0.6873,
+      "loss_token": 1.4929,
+      "lr": "2.00e-07",
+      "norm": 5.5336,
+      "step": 1511
+    },
+    {
+      "acc_char": 0.3049,
+      "acc_token": 0.688,
+      "epoch": 0.03,
+      "loss_char": 0.7237,
+      "loss_token": 1.6694,
+      "lr": "2.00e-07",
+      "norm": 6.0466,
+      "step": 1512
+    },
+    {
+      "acc_char": 0.3071,
+      "acc_token": 0.6709,
+      "epoch": 0.03,
+      "loss_char": 0.7927,
+      "loss_token": 1.6294,
+      "lr": "2.00e-07",
+      "norm": 6.2413,
+      "step": 1513
+    },
+    {
+      "acc_char": 0.3108,
+      "acc_token": 0.6995,
+      "epoch": 0.03,
+      "loss_char": 0.6674,
+      "loss_token": 1.5762,
+      "lr": "2.00e-07",
+      "norm": 5.6092,
+      "step": 1514
+    },
+    {
+      "acc_char": 0.2982,
+      "acc_token": 0.6776,
+      "epoch": 0.03,
+      "loss_char": 0.7238,
+      "loss_token": 1.5251,
+      "lr": "2.00e-07",
+      "norm": 5.4862,
+      "step": 1515
+    },
+    {
+      "acc_char": 0.3253,
+      "acc_token": 0.7275,
+      "epoch": 0.03,
+      "loss_char": 0.6361,
+      "loss_token": 1.4564,
+      "lr": "2.00e-07",
+      "norm": 5.9106,
+      "step": 1516
+    },
+    {
+      "acc_char": 0.2983,
+      "acc_token": 0.6783,
+      "epoch": 0.03,
+      "loss_char": 0.725,
+      "loss_token": 1.5952,
+      "lr": "2.00e-07",
+      "norm": 5.4654,
+      "step": 1517
+    },
+    {
+      "acc_char": 0.3083,
+      "acc_token": 0.6961,
+      "epoch": 0.03,
+      "loss_char": 0.6991,
+      "loss_token": 1.637,
+      "lr": "2.00e-07",
+      "norm": 5.9309,
+      "step": 1518
+    },
+    {
+      "acc_char": 0.3237,
+      "acc_token": 0.736,
+      "epoch": 0.03,
+      "loss_char": 0.5821,
+      "loss_token": 1.4387,
+      "lr": "2.00e-07",
+      "norm": 9.0917,
+      "step": 1519
+    },
+    {
+      "acc_char": 0.3144,
+      "acc_token": 0.7185,
+      "epoch": 0.03,
+      "loss_char": 0.6176,
+      "loss_token": 1.4751,
+      "lr": "2.00e-07",
+      "norm": 5.3616,
+      "step": 1520
+    },
+    {
+      "acc_char": 0.302,
+      "acc_token": 0.6792,
+      "epoch": 0.03,
+      "loss_char": 0.7392,
+      "loss_token": 1.6167,
+      "lr": "2.00e-07",
+      "norm": 5.9091,
+      "step": 1521
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6809,
+      "epoch": 0.03,
+      "loss_char": 0.7184,
+      "loss_token": 1.6441,
+      "lr": "2.00e-07",
+      "norm": 6.1267,
+      "step": 1522
+    },
+    {
+      "acc_char": 0.3028,
+      "acc_token": 0.6993,
+      "epoch": 0.03,
+      "loss_char": 0.659,
+      "loss_token": 1.5538,
+      "lr": "2.00e-07",
+      "norm": 5.7089,
+      "step": 1523
+    },
+    {
+      "acc_char": 0.3049,
+      "acc_token": 0.6934,
+      "epoch": 0.03,
+      "loss_char": 0.7043,
+      "loss_token": 1.6431,
+      "lr": "2.00e-07",
+      "norm": 5.7937,
+      "step": 1524
+    },
+    {
+      "acc_char": 0.3015,
+      "acc_token": 0.6784,
+      "epoch": 0.03,
+      "loss_char": 0.7425,
+      "loss_token": 1.5777,
+      "lr": "2.00e-07",
+      "norm": 5.8933,
+      "step": 1525
+    },
+    {
+      "acc_char": 0.3037,
+      "acc_token": 0.6474,
+      "epoch": 0.03,
+      "loss_char": 0.8172,
+      "loss_token": 1.7262,
+      "lr": "2.00e-07",
+      "norm": 6.004,
+      "step": 1526
+    },
+    {
+      "acc_char": 0.3026,
+      "acc_token": 0.684,
+      "epoch": 0.03,
+      "loss_char": 0.7037,
+      "loss_token": 1.5507,
+      "lr": "2.00e-07",
+      "norm": 5.5811,
+      "step": 1527
+    },
+    {
+      "acc_char": 0.2999,
+      "acc_token": 0.6867,
+      "epoch": 0.03,
+      "loss_char": 0.7148,
+      "loss_token": 1.6341,
+      "lr": "2.00e-07",
+      "norm": 5.7757,
+      "step": 1528
+    },
+    {
+      "acc_char": 0.3044,
+      "acc_token": 0.6911,
+      "epoch": 0.03,
+      "loss_char": 0.7137,
+      "loss_token": 1.5792,
+      "lr": "2.00e-07",
+      "norm": 5.6105,
+      "step": 1529
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.6929,
+      "epoch": 0.03,
+      "loss_char": 0.6726,
+      "loss_token": 1.5453,
+      "lr": "2.00e-07",
+      "norm": 5.6457,
+      "step": 1530
+    },
+    {
+      "acc_char": 0.2814,
+      "acc_token": 0.6126,
+      "epoch": 0.03,
+      "loss_char": 0.8906,
+      "loss_token": 1.6564,
+      "lr": "2.00e-07",
+      "norm": 5.5373,
+      "step": 1531
+    },
+    {
+      "acc_char": 0.3132,
+      "acc_token": 0.7041,
+      "epoch": 0.03,
+      "loss_char": 0.6804,
+      "loss_token": 1.544,
+      "lr": "2.00e-07",
+      "norm": 5.454,
+      "step": 1532
+    },
+    {
+      "acc_char": 0.2948,
+      "acc_token": 0.6682,
+      "epoch": 0.03,
+      "loss_char": 0.7727,
+      "loss_token": 1.6557,
+      "lr": "2.00e-07",
+      "norm": 5.9132,
+      "step": 1533
+    },
+    {
+      "acc_char": 0.3199,
+      "acc_token": 0.7156,
+      "epoch": 0.03,
+      "loss_char": 0.6603,
+      "loss_token": 1.6086,
+      "lr": "2.00e-07",
+      "norm": 5.5924,
+      "step": 1534
+    },
+    {
+      "acc_char": 0.2951,
+      "acc_token": 0.6527,
+      "epoch": 0.03,
+      "loss_char": 0.8307,
+      "loss_token": 1.7452,
+      "lr": "2.00e-07",
+      "norm": 5.8703,
+      "step": 1535
+    },
+    {
+      "acc_char": 0.2731,
+      "acc_token": 0.6037,
+      "epoch": 0.03,
+      "loss_char": 0.9329,
+      "loss_token": 1.8911,
+      "lr": "2.00e-07",
+      "norm": 5.598,
+      "step": 1536
+    },
+    {
+      "acc_char": 0.308,
+      "acc_token": 0.6992,
+      "epoch": 0.03,
+      "loss_char": 0.6665,
+      "loss_token": 1.5613,
+      "lr": "2.00e-07",
+      "norm": 5.7335,
+      "step": 1537
+    },
+    {
+      "acc_char": 0.2874,
+      "acc_token": 0.6679,
+      "epoch": 0.03,
+      "loss_char": 0.7361,
+      "loss_token": 1.7051,
+      "lr": "2.00e-07",
+      "norm": 5.9876,
+      "step": 1538
+    },
+    {
+      "acc_char": 0.3093,
+      "acc_token": 0.6969,
+      "epoch": 0.03,
+      "loss_char": 0.6927,
+      "loss_token": 1.5823,
+      "lr": "2.00e-07",
+      "norm": 5.6095,
+      "step": 1539
+    },
+    {
+      "acc_char": 0.2997,
+      "acc_token": 0.6724,
+      "epoch": 0.03,
+      "loss_char": 0.7626,
+      "loss_token": 1.5533,
+      "lr": "2.00e-07",
+      "norm": 5.8144,
+      "step": 1540
+    },
+    {
+      "acc_char": 0.284,
+      "acc_token": 0.6538,
+      "epoch": 0.03,
+      "loss_char": 0.7933,
+      "loss_token": 1.7849,
+      "lr": "2.00e-07",
+      "norm": 5.8935,
+      "step": 1541
+    },
+    {
+      "acc_char": 0.3049,
+      "acc_token": 0.7073,
+      "epoch": 0.03,
+      "loss_char": 0.6401,
+      "loss_token": 1.554,
+      "lr": "2.00e-07",
+      "norm": 6.5277,
+      "step": 1542
+    },
+    {
+      "acc_char": 0.2852,
+      "acc_token": 0.631,
+      "epoch": 0.03,
+      "loss_char": 0.874,
+      "loss_token": 1.7801,
+      "lr": "2.00e-07",
+      "norm": 5.8398,
+      "step": 1543
+    },
+    {
+      "acc_char": 0.2913,
+      "acc_token": 0.6557,
+      "epoch": 0.03,
+      "loss_char": 0.8232,
+      "loss_token": 1.6764,
+      "lr": "2.00e-07",
+      "norm": 6.0703,
+      "step": 1544
+    },
+    {
+      "acc_char": 0.2922,
+      "acc_token": 0.6589,
+      "epoch": 0.03,
+      "loss_char": 0.7863,
+      "loss_token": 1.7354,
+      "lr": "2.00e-07",
+      "norm": 5.9921,
+      "step": 1545
+    },
+    {
+      "acc_char": 0.2999,
+      "acc_token": 0.6869,
+      "epoch": 0.03,
+      "loss_char": 0.71,
+      "loss_token": 1.5729,
+      "lr": "2.00e-07",
+      "norm": 5.8846,
+      "step": 1546
+    },
+    {
+      "acc_char": 0.3036,
+      "acc_token": 0.6931,
+      "epoch": 0.03,
+      "loss_char": 0.6835,
+      "loss_token": 1.4914,
+      "lr": "2.00e-07",
+      "norm": 5.4592,
+      "step": 1547
+    },
+    {
+      "acc_char": 0.3007,
+      "acc_token": 0.6818,
+      "epoch": 0.03,
+      "loss_char": 0.7137,
+      "loss_token": 1.5716,
+      "lr": "2.00e-07",
+      "norm": 5.5647,
+      "step": 1548
+    },
+    {
+      "acc_char": 0.3113,
+      "acc_token": 0.7023,
+      "epoch": 0.03,
+      "loss_char": 0.6746,
+      "loss_token": 1.6481,
+      "lr": "2.00e-07",
+      "norm": 6.1887,
+      "step": 1549
+    },
+    {
+      "acc_char": 0.3199,
+      "acc_token": 0.7231,
+      "epoch": 0.03,
+      "loss_char": 0.6157,
+      "loss_token": 1.4955,
+      "lr": "2.00e-07",
+      "norm": 5.2796,
+      "step": 1550
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.6919,
+      "epoch": 0.03,
+      "loss_char": 0.6797,
+      "loss_token": 1.5701,
+      "lr": "2.00e-07",
+      "norm": 5.3542,
+      "step": 1551
+    },
+    {
+      "acc_char": 0.2986,
+      "acc_token": 0.6709,
+      "epoch": 0.03,
+      "loss_char": 0.7678,
+      "loss_token": 1.6893,
+      "lr": "2.00e-07",
+      "norm": 6.0462,
+      "step": 1552
+    },
+    {
+      "acc_char": 0.3057,
+      "acc_token": 0.6952,
+      "epoch": 0.03,
+      "loss_char": 0.6986,
+      "loss_token": 1.6515,
+      "lr": "2.00e-07",
+      "norm": 6.3134,
+      "step": 1553
+    },
+    {
+      "acc_char": 0.3197,
+      "acc_token": 0.7216,
+      "epoch": 0.03,
+      "loss_char": 0.6201,
+      "loss_token": 1.4663,
+      "lr": "2.00e-07",
+      "norm": 5.6139,
+      "step": 1554
+    },
+    {
+      "acc_char": 0.2872,
+      "acc_token": 0.6617,
+      "epoch": 0.03,
+      "loss_char": 0.7866,
+      "loss_token": 1.7042,
+      "lr": "2.00e-07",
+      "norm": 5.8264,
+      "step": 1555
+    },
+    {
+      "acc_char": 0.2946,
+      "acc_token": 0.6686,
+      "epoch": 0.03,
+      "loss_char": 0.7846,
+      "loss_token": 1.677,
+      "lr": "2.00e-07",
+      "norm": 5.8586,
+      "step": 1556
+    },
+    {
+      "acc_char": 0.2852,
+      "acc_token": 0.6372,
+      "epoch": 0.03,
+      "loss_char": 0.8567,
+      "loss_token": 1.756,
+      "lr": "2.00e-07",
+      "norm": 5.9905,
+      "step": 1557
+    },
+    {
+      "acc_char": 0.2986,
+      "acc_token": 0.6659,
+      "epoch": 0.03,
+      "loss_char": 0.785,
+      "loss_token": 1.5579,
+      "lr": "2.00e-07",
+      "norm": 5.7136,
+      "step": 1558
+    },
+    {
+      "acc_char": 0.3083,
+      "acc_token": 0.7062,
+      "epoch": 0.03,
+      "loss_char": 0.643,
+      "loss_token": 1.4669,
+      "lr": "2.00e-07",
+      "norm": 5.4612,
+      "step": 1559
+    },
+    {
+      "acc_char": 0.2992,
+      "acc_token": 0.6764,
+      "epoch": 0.03,
+      "loss_char": 0.7477,
+      "loss_token": 1.7046,
+      "lr": "2.00e-07",
+      "norm": 6.3699,
+      "step": 1560
+    },
+    {
+      "acc_char": 0.2895,
+      "acc_token": 0.6508,
+      "epoch": 0.03,
+      "loss_char": 0.8214,
+      "loss_token": 1.7717,
+      "lr": "2.00e-07",
+      "norm": 6.0916,
+      "step": 1561
+    },
+    {
+      "acc_char": 0.2966,
+      "acc_token": 0.674,
+      "epoch": 0.03,
+      "loss_char": 0.7445,
+      "loss_token": 1.6343,
+      "lr": "2.00e-07",
+      "norm": 6.1589,
+      "step": 1562
+    },
+    {
+      "acc_char": 0.2937,
+      "acc_token": 0.669,
+      "epoch": 0.03,
+      "loss_char": 0.7472,
+      "loss_token": 1.6864,
+      "lr": "2.00e-07",
+      "norm": 5.7867,
+      "step": 1563
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6748,
+      "epoch": 0.03,
+      "loss_char": 0.742,
+      "loss_token": 1.6648,
+      "lr": "2.00e-07",
+      "norm": 5.9749,
+      "step": 1564
+    },
+    {
+      "acc_char": 0.2934,
+      "acc_token": 0.6705,
+      "epoch": 0.03,
+      "loss_char": 0.7405,
+      "loss_token": 1.6838,
+      "lr": "2.00e-07",
+      "norm": 5.7158,
+      "step": 1565
+    },
+    {
+      "acc_char": 0.2926,
+      "acc_token": 0.6674,
+      "epoch": 0.03,
+      "loss_char": 0.7529,
+      "loss_token": 1.746,
+      "lr": "2.00e-07",
+      "norm": 5.8629,
+      "step": 1566
+    },
+    {
+      "acc_char": 0.2944,
+      "acc_token": 0.6342,
+      "epoch": 0.03,
+      "loss_char": 0.8539,
+      "loss_token": 1.7368,
+      "lr": "2.00e-07",
+      "norm": 5.8781,
+      "step": 1567
+    },
+    {
+      "acc_char": 0.2776,
+      "acc_token": 0.6914,
+      "epoch": 0.03,
+      "loss_char": 0.6223,
+      "loss_token": 1.6307,
+      "lr": "2.00e-07",
+      "norm": 5.8752,
+      "step": 1568
+    },
+    {
+      "acc_char": 0.3047,
+      "acc_token": 0.6875,
+      "epoch": 0.03,
+      "loss_char": 0.7133,
+      "loss_token": 1.6326,
+      "lr": "2.00e-07",
+      "norm": 5.7235,
+      "step": 1569
+    },
+    {
+      "acc_char": 0.3014,
+      "acc_token": 0.6858,
+      "epoch": 0.03,
+      "loss_char": 0.7244,
+      "loss_token": 1.6255,
+      "lr": "2.00e-07",
+      "norm": 5.8664,
+      "step": 1570
+    },
+    {
+      "acc_char": 0.2994,
+      "acc_token": 0.6859,
+      "epoch": 0.03,
+      "loss_char": 0.686,
+      "loss_token": 1.6009,
+      "lr": "2.00e-07",
+      "norm": 5.8634,
+      "step": 1571
+    },
+    {
+      "acc_char": 0.3086,
+      "acc_token": 0.6875,
+      "epoch": 0.03,
+      "loss_char": 0.7275,
+      "loss_token": 1.577,
+      "lr": "2.00e-07",
+      "norm": 5.6534,
+      "step": 1572
+    },
+    {
+      "acc_char": 0.3029,
+      "acc_token": 0.6868,
+      "epoch": 0.03,
+      "loss_char": 0.6953,
+      "loss_token": 1.5916,
+      "lr": "2.00e-07",
+      "norm": 5.8533,
+      "step": 1573
+    },
+    {
+      "acc_char": 0.2905,
+      "acc_token": 0.6677,
+      "epoch": 0.03,
+      "loss_char": 0.7498,
+      "loss_token": 1.6433,
+      "lr": "2.00e-07",
+      "norm": 5.6664,
+      "step": 1574
+    },
+    {
+      "acc_char": 0.3043,
+      "acc_token": 0.6911,
+      "epoch": 0.03,
+      "loss_char": 0.689,
+      "loss_token": 1.6164,
+      "lr": "2.00e-07",
+      "norm": 5.9362,
+      "step": 1575
+    },
+    {
+      "acc_char": 0.3025,
+      "acc_token": 0.6621,
+      "epoch": 0.03,
+      "loss_char": 0.7666,
+      "loss_token": 1.728,
+      "lr": "2.00e-07",
+      "norm": 5.61,
+      "step": 1576
+    },
+    {
+      "acc_char": 0.2988,
+      "acc_token": 0.6836,
+      "epoch": 0.03,
+      "loss_char": 0.6929,
+      "loss_token": 1.5914,
+      "lr": "2.00e-07",
+      "norm": 5.441,
+      "step": 1577
+    },
+    {
+      "acc_char": 0.3046,
+      "acc_token": 0.6872,
+      "epoch": 0.03,
+      "loss_char": 0.7199,
+      "loss_token": 1.5617,
+      "lr": "2.00e-07",
+      "norm": 5.874,
+      "step": 1578
+    },
+    {
+      "acc_char": 0.3127,
+      "acc_token": 0.7107,
+      "epoch": 0.03,
+      "loss_char": 0.6479,
+      "loss_token": 1.5476,
+      "lr": "2.00e-07",
+      "norm": 5.5485,
+      "step": 1579
+    },
+    {
+      "acc_char": 0.3122,
+      "acc_token": 0.7101,
+      "epoch": 0.03,
+      "loss_char": 0.6448,
+      "loss_token": 1.5271,
+      "lr": "2.00e-07",
+      "norm": 5.7229,
+      "step": 1580
+    },
+    {
+      "acc_char": 0.3166,
+      "acc_token": 0.7051,
+      "epoch": 0.03,
+      "loss_char": 0.6688,
+      "loss_token": 1.5672,
+      "lr": "2.00e-07",
+      "norm": 6.229,
+      "step": 1581
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.6724,
+      "epoch": 0.03,
+      "loss_char": 0.7677,
+      "loss_token": 1.7375,
+      "lr": "2.00e-07",
+      "norm": 5.9405,
+      "step": 1582
+    },
+    {
+      "acc_char": 0.305,
+      "acc_token": 0.6874,
+      "epoch": 0.03,
+      "loss_char": 0.7192,
+      "loss_token": 1.6619,
+      "lr": "2.00e-07",
+      "norm": 5.5951,
+      "step": 1583
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6898,
+      "epoch": 0.03,
+      "loss_char": 0.6772,
+      "loss_token": 1.5391,
+      "lr": "2.00e-07",
+      "norm": 5.6114,
+      "step": 1584
+    },
+    {
+      "acc_char": 0.2955,
+      "acc_token": 0.6787,
+      "epoch": 0.03,
+      "loss_char": 0.7257,
+      "loss_token": 1.6871,
+      "lr": "2.00e-07",
+      "norm": 5.8745,
+      "step": 1585
+    },
+    {
+      "acc_char": 0.3106,
+      "acc_token": 0.6899,
+      "epoch": 0.03,
+      "loss_char": 0.7315,
+      "loss_token": 1.641,
+      "lr": "2.00e-07",
+      "norm": 5.6256,
+      "step": 1586
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6919,
+      "epoch": 0.03,
+      "loss_char": 0.7055,
+      "loss_token": 1.6004,
+      "lr": "2.00e-07",
+      "norm": 5.8087,
+      "step": 1587
+    },
+    {
+      "acc_char": 0.2998,
+      "acc_token": 0.6839,
+      "epoch": 0.03,
+      "loss_char": 0.6994,
+      "loss_token": 1.6181,
+      "lr": "2.00e-07",
+      "norm": 6.0405,
+      "step": 1588
+    },
+    {
+      "acc_char": 0.2947,
+      "acc_token": 0.6892,
+      "epoch": 0.03,
+      "loss_char": 0.6667,
+      "loss_token": 1.5679,
+      "lr": "2.00e-07",
+      "norm": 5.5971,
+      "step": 1589
+    },
+    {
+      "acc_char": 0.3153,
+      "acc_token": 0.7029,
+      "epoch": 0.03,
+      "loss_char": 0.6714,
+      "loss_token": 1.5788,
+      "lr": "2.00e-07",
+      "norm": 5.8712,
+      "step": 1590
+    },
+    {
+      "acc_char": 0.2986,
+      "acc_token": 0.6734,
+      "epoch": 0.03,
+      "loss_char": 0.746,
+      "loss_token": 1.5374,
+      "lr": "2.00e-07",
+      "norm": 5.7618,
+      "step": 1591
+    },
+    {
+      "acc_char": 0.2929,
+      "acc_token": 0.6699,
+      "epoch": 0.03,
+      "loss_char": 0.7309,
+      "loss_token": 1.6542,
+      "lr": "2.00e-07",
+      "norm": 5.9127,
+      "step": 1592
+    },
+    {
+      "acc_char": 0.3077,
+      "acc_token": 0.7082,
+      "epoch": 0.03,
+      "loss_char": 0.6441,
+      "loss_token": 1.565,
+      "lr": "2.00e-07",
+      "norm": 5.8163,
+      "step": 1593
+    },
+    {
+      "acc_char": 0.3114,
+      "acc_token": 0.7076,
+      "epoch": 0.03,
+      "loss_char": 0.6438,
+      "loss_token": 1.5501,
+      "lr": "2.00e-07",
+      "norm": 5.9952,
+      "step": 1594
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6792,
+      "epoch": 0.03,
+      "loss_char": 0.7387,
+      "loss_token": 1.6415,
+      "lr": "2.00e-07",
+      "norm": 6.0267,
+      "step": 1595
+    },
+    {
+      "acc_char": 0.3012,
+      "acc_token": 0.6793,
+      "epoch": 0.03,
+      "loss_char": 0.7423,
+      "loss_token": 1.6092,
+      "lr": "2.00e-07",
+      "norm": 5.8914,
+      "step": 1596
+    },
+    {
+      "acc_char": 0.2812,
+      "acc_token": 0.6123,
+      "epoch": 0.03,
+      "loss_char": 0.9265,
+      "loss_token": 1.8237,
+      "lr": "2.00e-07",
+      "norm": 6.0108,
+      "step": 1597
+    },
+    {
+      "acc_char": 0.3074,
+      "acc_token": 0.6941,
+      "epoch": 0.03,
+      "loss_char": 0.7013,
+      "loss_token": 1.5543,
+      "lr": "2.00e-07",
+      "norm": 5.7529,
+      "step": 1598
+    },
+    {
+      "acc_char": 0.3193,
+      "acc_token": 0.7049,
+      "epoch": 0.03,
+      "loss_char": 0.6746,
+      "loss_token": 1.5883,
+      "lr": "2.00e-07",
+      "norm": 5.7992,
+      "step": 1599
+    },
+    {
+      "acc_char": 0.2794,
+      "acc_token": 0.6196,
+      "epoch": 0.03,
+      "loss_char": 0.8698,
+      "loss_token": 1.8613,
+      "lr": "2.00e-07",
+      "norm": 5.6176,
+      "step": 1600
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6837,
+      "epoch": 0.03,
+      "loss_char": 0.716,
+      "loss_token": 1.6654,
+      "lr": "2.00e-07",
+      "norm": 5.8047,
+      "step": 1601
+    },
+    {
+      "acc_char": 0.2936,
+      "acc_token": 0.6689,
+      "epoch": 0.03,
+      "loss_char": 0.7596,
+      "loss_token": 1.5968,
+      "lr": "2.00e-07",
+      "norm": 5.7253,
+      "step": 1602
+    },
+    {
+      "acc_char": 0.307,
+      "acc_token": 0.6971,
+      "epoch": 0.03,
+      "loss_char": 0.7045,
+      "loss_token": 1.5782,
+      "lr": "2.00e-07",
+      "norm": 5.7435,
+      "step": 1603
+    },
+    {
+      "acc_char": 0.3112,
+      "acc_token": 0.6984,
+      "epoch": 0.03,
+      "loss_char": 0.6758,
+      "loss_token": 1.5436,
+      "lr": "2.00e-07",
+      "norm": 5.5545,
+      "step": 1604
+    },
+    {
+      "acc_char": 0.3087,
+      "acc_token": 0.6902,
+      "epoch": 0.03,
+      "loss_char": 0.7368,
+      "loss_token": 1.6031,
+      "lr": "2.00e-07",
+      "norm": 5.8911,
+      "step": 1605
+    },
+    {
+      "acc_char": 0.3206,
+      "acc_token": 0.7121,
+      "epoch": 0.03,
+      "loss_char": 0.6525,
+      "loss_token": 1.4743,
+      "lr": "2.00e-07",
+      "norm": 5.6098,
+      "step": 1606
+    },
+    {
+      "acc_char": 0.2956,
+      "acc_token": 0.6733,
+      "epoch": 0.03,
+      "loss_char": 0.7465,
+      "loss_token": 1.6542,
+      "lr": "2.00e-07",
+      "norm": 5.6449,
+      "step": 1607
+    },
+    {
+      "acc_char": 0.3006,
+      "acc_token": 0.6817,
+      "epoch": 0.03,
+      "loss_char": 0.7201,
+      "loss_token": 1.6398,
+      "lr": "2.00e-07",
+      "norm": 6.1526,
+      "step": 1608
+    },
+    {
+      "acc_char": 0.32,
+      "acc_token": 0.7127,
+      "epoch": 0.03,
+      "loss_char": 0.6491,
+      "loss_token": 1.4844,
+      "lr": "2.00e-07",
+      "norm": 5.6564,
+      "step": 1609
+    },
+    {
+      "acc_char": 0.2977,
+      "acc_token": 0.6769,
+      "epoch": 0.03,
+      "loss_char": 0.7228,
+      "loss_token": 1.6341,
+      "lr": "2.00e-07",
+      "norm": 6.3097,
+      "step": 1610
+    },
+    {
+      "acc_char": 0.3051,
+      "acc_token": 0.685,
+      "epoch": 0.03,
+      "loss_char": 0.7089,
+      "loss_token": 1.6848,
+      "lr": "2.00e-07",
+      "norm": 5.7969,
+      "step": 1611
+    },
+    {
+      "acc_char": 0.3016,
+      "acc_token": 0.6773,
+      "epoch": 0.03,
+      "loss_char": 0.7574,
+      "loss_token": 1.6114,
+      "lr": "2.00e-07",
+      "norm": 5.7741,
+      "step": 1612
+    },
+    {
+      "acc_char": 0.3078,
+      "acc_token": 0.6986,
+      "epoch": 0.03,
+      "loss_char": 0.6857,
+      "loss_token": 1.5755,
+      "lr": "2.00e-07",
+      "norm": 5.4162,
+      "step": 1613
+    },
+    {
+      "acc_char": 0.2967,
+      "acc_token": 0.6676,
+      "epoch": 0.03,
+      "loss_char": 0.781,
+      "loss_token": 1.6191,
+      "lr": "2.00e-07",
+      "norm": 6.1566,
+      "step": 1614
+    },
+    {
+      "acc_char": 0.3072,
+      "acc_token": 0.7054,
+      "epoch": 0.03,
+      "loss_char": 0.6538,
+      "loss_token": 1.5423,
+      "lr": "2.00e-07",
+      "norm": 6.3931,
+      "step": 1615
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.6881,
+      "epoch": 0.03,
+      "loss_char": 0.7145,
+      "loss_token": 1.5765,
+      "lr": "2.00e-07",
+      "norm": 5.7099,
+      "step": 1616
+    },
+    {
+      "acc_char": 0.298,
+      "acc_token": 0.6861,
+      "epoch": 0.03,
+      "loss_char": 0.7009,
+      "loss_token": 1.6293,
+      "lr": "2.00e-07",
+      "norm": 5.4639,
+      "step": 1617
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6804,
+      "epoch": 0.03,
+      "loss_char": 0.7174,
+      "loss_token": 1.6531,
+      "lr": "2.00e-07",
+      "norm": 5.9676,
+      "step": 1618
+    },
+    {
+      "acc_char": 0.3102,
+      "acc_token": 0.7056,
+      "epoch": 0.03,
+      "loss_char": 0.647,
+      "loss_token": 1.4693,
+      "lr": "2.00e-07",
+      "norm": 5.5077,
+      "step": 1619
+    },
+    {
+      "acc_char": 0.3146,
+      "acc_token": 0.7076,
+      "epoch": 0.03,
+      "loss_char": 0.6477,
+      "loss_token": 1.5071,
+      "lr": "2.00e-07",
+      "norm": 5.8307,
+      "step": 1620
+    },
+    {
+      "acc_char": 0.323,
+      "acc_token": 0.7129,
+      "epoch": 0.03,
+      "loss_char": 0.6701,
+      "loss_token": 1.462,
+      "lr": "2.00e-07",
+      "norm": 5.6707,
+      "step": 1621
+    },
+    {
+      "acc_char": 0.3075,
+      "acc_token": 0.6937,
+      "epoch": 0.03,
+      "loss_char": 0.6996,
+      "loss_token": 1.6832,
+      "lr": "2.00e-07",
+      "norm": 5.8797,
+      "step": 1622
+    },
+    {
+      "acc_char": 0.3074,
+      "acc_token": 0.6768,
+      "epoch": 0.03,
+      "loss_char": 0.7479,
+      "loss_token": 1.6044,
+      "lr": "2.00e-07",
+      "norm": 6.0979,
+      "step": 1623
+    },
+    {
+      "acc_char": 0.3096,
+      "acc_token": 0.6896,
+      "epoch": 0.03,
+      "loss_char": 0.7126,
+      "loss_token": 1.5891,
+      "lr": "2.00e-07",
+      "norm": 5.9975,
+      "step": 1624
+    },
+    {
+      "acc_char": 0.3054,
+      "acc_token": 0.7098,
+      "epoch": 0.03,
+      "loss_char": 0.6366,
+      "loss_token": 1.4998,
+      "lr": "2.00e-07",
+      "norm": 5.5291,
+      "step": 1625
+    },
+    {
+      "acc_char": 0.2923,
+      "acc_token": 0.6795,
+      "epoch": 0.03,
+      "loss_char": 0.6891,
+      "loss_token": 1.6086,
+      "lr": "2.00e-07",
+      "norm": 6.0158,
+      "step": 1626
+    },
+    {
+      "acc_char": 0.3,
+      "acc_token": 0.6793,
+      "epoch": 0.03,
+      "loss_char": 0.7358,
+      "loss_token": 1.6231,
+      "lr": "2.00e-07",
+      "norm": 5.588,
+      "step": 1627
+    },
+    {
+      "acc_char": 0.2866,
+      "acc_token": 0.6444,
+      "epoch": 0.03,
+      "loss_char": 0.8436,
+      "loss_token": 1.7327,
+      "lr": "2.00e-07",
+      "norm": 5.6852,
+      "step": 1628
+    },
+    {
+      "acc_char": 0.3153,
+      "acc_token": 0.718,
+      "epoch": 0.03,
+      "loss_char": 0.6243,
+      "loss_token": 1.4671,
+      "lr": "2.00e-07",
+      "norm": 5.329,
+      "step": 1629
+    },
+    {
+      "acc_char": 0.298,
+      "acc_token": 0.6713,
+      "epoch": 0.03,
+      "loss_char": 0.7667,
+      "loss_token": 1.6396,
+      "lr": "2.00e-07",
+      "norm": 5.5255,
+      "step": 1630
+    },
+    {
+      "acc_char": 0.3222,
+      "acc_token": 0.7327,
+      "epoch": 0.03,
+      "loss_char": 0.6019,
+      "loss_token": 1.478,
+      "lr": "2.00e-07",
+      "norm": 5.5881,
+      "step": 1631
+    },
+    {
+      "acc_char": 0.2903,
+      "acc_token": 0.6636,
+      "epoch": 0.03,
+      "loss_char": 0.7665,
+      "loss_token": 1.7035,
+      "lr": "2.00e-07",
+      "norm": 5.9326,
+      "step": 1632
+    },
+    {
+      "acc_char": 0.3145,
+      "acc_token": 0.7065,
+      "epoch": 0.03,
+      "loss_char": 0.6586,
+      "loss_token": 1.4825,
+      "lr": "2.00e-07",
+      "norm": 5.404,
+      "step": 1633
+    },
+    {
+      "acc_char": 0.2927,
+      "acc_token": 0.6748,
+      "epoch": 0.03,
+      "loss_char": 0.7274,
+      "loss_token": 1.6455,
+      "lr": "2.00e-07",
+      "norm": 5.5322,
+      "step": 1634
+    },
+    {
+      "acc_char": 0.2951,
+      "acc_token": 0.6765,
+      "epoch": 0.03,
+      "loss_char": 0.7247,
+      "loss_token": 1.6789,
+      "lr": "2.00e-07",
+      "norm": 5.9366,
+      "step": 1635
+    },
+    {
+      "acc_char": 0.3179,
+      "acc_token": 0.7132,
+      "epoch": 0.03,
+      "loss_char": 0.651,
+      "loss_token": 1.5661,
+      "lr": "2.00e-07",
+      "norm": 5.6878,
+      "step": 1636
+    },
+    {
+      "acc_char": 0.2999,
+      "acc_token": 0.6914,
+      "epoch": 0.03,
+      "loss_char": 0.6779,
+      "loss_token": 1.5841,
+      "lr": "2.00e-07",
+      "norm": 5.5705,
+      "step": 1637
+    },
+    {
+      "acc_char": 0.2868,
+      "acc_token": 0.6586,
+      "epoch": 0.03,
+      "loss_char": 0.7848,
+      "loss_token": 1.7319,
+      "lr": "2.00e-07",
+      "norm": 5.9417,
+      "step": 1638
+    },
+    {
+      "acc_char": 0.3076,
+      "acc_token": 0.7015,
+      "epoch": 0.03,
+      "loss_char": 0.664,
+      "loss_token": 1.5467,
+      "lr": "2.00e-07",
+      "norm": 5.6836,
+      "step": 1639
+    },
+    {
+      "acc_char": 0.2993,
+      "acc_token": 0.6912,
+      "epoch": 0.03,
+      "loss_char": 0.6707,
+      "loss_token": 1.5604,
+      "lr": "2.00e-07",
+      "norm": 5.6171,
+      "step": 1640
+    },
+    {
+      "acc_char": 0.3036,
+      "acc_token": 0.6803,
+      "epoch": 0.03,
+      "loss_char": 0.7486,
+      "loss_token": 1.6083,
+      "lr": "2.00e-07",
+      "norm": 5.6892,
+      "step": 1641
+    },
+    {
+      "acc_char": 0.2826,
+      "acc_token": 0.6196,
+      "epoch": 0.03,
+      "loss_char": 0.864,
+      "loss_token": 1.7846,
+      "lr": "2.00e-07",
+      "norm": 5.726,
+      "step": 1642
+    },
+    {
+      "acc_char": 0.2922,
+      "acc_token": 0.6715,
+      "epoch": 0.03,
+      "loss_char": 0.7413,
+      "loss_token": 1.622,
+      "lr": "2.00e-07",
+      "norm": 5.8253,
+      "step": 1643
+    },
+    {
+      "acc_char": 0.3087,
+      "acc_token": 0.7041,
+      "epoch": 0.03,
+      "loss_char": 0.6732,
+      "loss_token": 1.6123,
+      "lr": "2.00e-07",
+      "norm": 5.7101,
+      "step": 1644
+    },
+    {
+      "acc_char": 0.2978,
+      "acc_token": 0.6763,
+      "epoch": 0.03,
+      "loss_char": 0.7507,
+      "loss_token": 1.6697,
+      "lr": "2.00e-07",
+      "norm": 6.2216,
+      "step": 1645
+    },
+    {
+      "acc_char": 0.3017,
+      "acc_token": 0.6764,
+      "epoch": 0.03,
+      "loss_char": 0.7482,
+      "loss_token": 1.6147,
+      "lr": "2.00e-07",
+      "norm": 5.9259,
+      "step": 1646
+    },
+    {
+      "acc_char": 0.305,
+      "acc_token": 0.6866,
+      "epoch": 0.03,
+      "loss_char": 0.7198,
+      "loss_token": 1.612,
+      "lr": "2.00e-07",
+      "norm": 5.7423,
+      "step": 1647
+    },
+    {
+      "acc_char": 0.3004,
+      "acc_token": 0.6849,
+      "epoch": 0.03,
+      "loss_char": 0.6971,
+      "loss_token": 1.479,
+      "lr": "2.00e-07",
+      "norm": 5.5494,
+      "step": 1648
+    },
+    {
+      "acc_char": 0.3019,
+      "acc_token": 0.691,
+      "epoch": 0.03,
+      "loss_char": 0.6878,
+      "loss_token": 1.5946,
+      "lr": "2.00e-07",
+      "norm": 5.4234,
+      "step": 1649
+    },
+    {
+      "acc_char": 0.3052,
+      "acc_token": 0.682,
+      "epoch": 0.03,
+      "loss_char": 0.7504,
+      "loss_token": 1.6363,
+      "lr": "2.00e-07",
+      "norm": 5.7636,
+      "step": 1650
+    },
+    {
+      "acc_char": 0.317,
+      "acc_token": 0.6991,
+      "epoch": 0.03,
+      "loss_char": 0.7028,
+      "loss_token": 1.553,
+      "lr": "2.00e-07",
+      "norm": 5.845,
+      "step": 1651
+    },
+    {
+      "acc_char": 0.2978,
+      "acc_token": 0.6823,
+      "epoch": 0.03,
+      "loss_char": 0.7021,
+      "loss_token": 1.6016,
+      "lr": "2.00e-07",
+      "norm": 5.9888,
+      "step": 1652
+    },
+    {
+      "acc_char": 0.3052,
+      "acc_token": 0.6953,
+      "epoch": 0.03,
+      "loss_char": 0.6907,
+      "loss_token": 1.5925,
+      "lr": "2.00e-07",
+      "norm": 5.5029,
+      "step": 1653
+    },
+    {
+      "acc_char": 0.2973,
+      "acc_token": 0.6836,
+      "epoch": 0.03,
+      "loss_char": 0.6947,
+      "loss_token": 1.573,
+      "lr": "2.00e-07",
+      "norm": 5.7139,
+      "step": 1654
+    },
+    {
+      "acc_char": 0.2818,
+      "acc_token": 0.6206,
+      "epoch": 0.03,
+      "loss_char": 0.8751,
+      "loss_token": 1.7996,
+      "lr": "2.00e-07",
+      "norm": 5.4943,
+      "step": 1655
+    },
+    {
+      "acc_char": 0.319,
+      "acc_token": 0.7293,
+      "epoch": 0.03,
+      "loss_char": 0.6127,
+      "loss_token": 1.5183,
+      "lr": "2.00e-07",
+      "norm": 5.6807,
+      "step": 1656
+    },
+    {
+      "acc_char": 0.305,
+      "acc_token": 0.6877,
+      "epoch": 0.03,
+      "loss_char": 0.7121,
+      "loss_token": 1.6186,
+      "lr": "2.00e-07",
+      "norm": 5.7593,
+      "step": 1657
+    },
+    {
+      "acc_char": 0.2833,
+      "acc_token": 0.6182,
+      "epoch": 0.03,
+      "loss_char": 0.8656,
+      "loss_token": 1.7297,
+      "lr": "2.00e-07",
+      "norm": 5.6634,
+      "step": 1658
+    },
+    {
+      "acc_char": 0.3079,
+      "acc_token": 0.6799,
+      "epoch": 0.03,
+      "loss_char": 0.7352,
+      "loss_token": 1.6045,
+      "lr": "2.00e-07",
+      "norm": 5.8658,
+      "step": 1659
+    },
+    {
+      "acc_char": 0.3183,
+      "acc_token": 0.7286,
+      "epoch": 0.03,
+      "loss_char": 0.5886,
+      "loss_token": 1.4656,
+      "lr": "2.00e-07",
+      "norm": 5.3731,
+      "step": 1660
+    },
+    {
+      "acc_char": 0.3095,
+      "acc_token": 0.7071,
+      "epoch": 0.03,
+      "loss_char": 0.6553,
+      "loss_token": 1.6271,
+      "lr": "2.00e-07",
+      "norm": 5.7404,
+      "step": 1661
+    },
+    {
+      "acc_char": 0.3166,
+      "acc_token": 0.7046,
+      "epoch": 0.03,
+      "loss_char": 0.6912,
+      "loss_token": 1.5677,
+      "lr": "2.00e-07",
+      "norm": 5.5271,
+      "step": 1662
+    },
+    {
+      "acc_char": 0.2971,
+      "acc_token": 0.671,
+      "epoch": 0.03,
+      "loss_char": 0.7658,
+      "loss_token": 1.6013,
+      "lr": "2.00e-07",
+      "norm": 5.7884,
+      "step": 1663
+    },
+    {
+      "acc_char": 0.3086,
+      "acc_token": 0.6928,
+      "epoch": 0.03,
+      "loss_char": 0.713,
+      "loss_token": 1.6149,
+      "lr": "2.00e-07",
+      "norm": 5.6196,
+      "step": 1664
+    },
+    {
+      "acc_char": 0.308,
+      "acc_token": 0.6924,
+      "epoch": 0.03,
+      "loss_char": 0.6979,
+      "loss_token": 1.5536,
+      "lr": "2.00e-07",
+      "norm": 5.8624,
+      "step": 1665
+    },
+    {
+      "acc_char": 0.3089,
+      "acc_token": 0.7017,
+      "epoch": 0.03,
+      "loss_char": 0.6717,
+      "loss_token": 1.6092,
+      "lr": "2.00e-07",
+      "norm": 5.6943,
+      "step": 1666
+    },
+    {
+      "acc_char": 0.2943,
+      "acc_token": 0.6652,
+      "epoch": 0.03,
+      "loss_char": 0.7602,
+      "loss_token": 1.7535,
+      "lr": "2.00e-07",
+      "norm": 6.2278,
+      "step": 1667
+    },
+    {
+      "acc_char": 0.3104,
+      "acc_token": 0.7119,
+      "epoch": 0.03,
+      "loss_char": 0.636,
+      "loss_token": 1.421,
+      "lr": "2.00e-07",
+      "norm": 5.2749,
+      "step": 1668
+    },
+    {
+      "acc_char": 0.3014,
+      "acc_token": 0.6839,
+      "epoch": 0.03,
+      "loss_char": 0.706,
+      "loss_token": 1.5401,
+      "lr": "2.00e-07",
+      "norm": 5.5522,
+      "step": 1669
+    },
+    {
+      "acc_char": 0.3068,
+      "acc_token": 0.6969,
+      "epoch": 0.03,
+      "loss_char": 0.6767,
+      "loss_token": 1.5882,
+      "lr": "2.00e-07",
+      "norm": 5.5253,
+      "step": 1670
+    },
+    {
+      "acc_char": 0.2855,
+      "acc_token": 0.6567,
+      "epoch": 0.03,
+      "loss_char": 0.7808,
+      "loss_token": 1.6454,
+      "lr": "2.00e-07",
+      "norm": 5.9135,
+      "step": 1671
+    },
+    {
+      "acc_char": 0.3075,
+      "acc_token": 0.6892,
+      "epoch": 0.03,
+      "loss_char": 0.7201,
+      "loss_token": 1.6653,
+      "lr": "2.00e-07",
+      "norm": 5.6643,
+      "step": 1672
+    },
+    {
+      "acc_char": 0.2983,
+      "acc_token": 0.686,
+      "epoch": 0.03,
+      "loss_char": 0.7036,
+      "loss_token": 1.4937,
+      "lr": "2.00e-07",
+      "norm": 5.5387,
+      "step": 1673
+    },
+    {
+      "acc_char": 0.3125,
+      "acc_token": 0.7078,
+      "epoch": 0.03,
+      "loss_char": 0.6312,
+      "loss_token": 1.4955,
+      "lr": "2.00e-07",
+      "norm": 5.6384,
+      "step": 1674
+    },
+    {
+      "acc_char": 0.2953,
+      "acc_token": 0.6475,
+      "epoch": 0.03,
+      "loss_char": 0.8527,
+      "loss_token": 1.7366,
+      "lr": "2.00e-07",
+      "norm": 5.8287,
+      "step": 1675
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.7003,
+      "epoch": 0.03,
+      "loss_char": 0.6829,
+      "loss_token": 1.5926,
+      "lr": "2.00e-07",
+      "norm": 6.3705,
+      "step": 1676
+    },
+    {
+      "acc_char": 0.3088,
+      "acc_token": 0.7072,
+      "epoch": 0.03,
+      "loss_char": 0.6492,
+      "loss_token": 1.5889,
+      "lr": "2.00e-07",
+      "norm": 5.6321,
+      "step": 1677
+    },
+    {
+      "acc_char": 0.3039,
+      "acc_token": 0.6827,
+      "epoch": 0.03,
+      "loss_char": 0.7191,
+      "loss_token": 1.639,
+      "lr": "2.00e-07",
+      "norm": 5.8795,
+      "step": 1678
+    },
+    {
+      "acc_char": 0.29,
+      "acc_token": 0.6521,
+      "epoch": 0.03,
+      "loss_char": 0.8078,
+      "loss_token": 1.6645,
+      "lr": "2.00e-07",
+      "norm": 5.6345,
+      "step": 1679
+    },
+    {
+      "acc_char": 0.291,
+      "acc_token": 0.6557,
+      "epoch": 0.03,
+      "loss_char": 0.8089,
+      "loss_token": 1.7202,
+      "lr": "2.00e-07",
+      "norm": 6.1896,
+      "step": 1680
+    },
+    {
+      "acc_char": 0.2941,
+      "acc_token": 0.6743,
+      "epoch": 0.03,
+      "loss_char": 0.7159,
+      "loss_token": 1.5983,
+      "lr": "2.00e-07",
+      "norm": 5.5012,
+      "step": 1681
+    },
+    {
+      "acc_char": 0.3052,
+      "acc_token": 0.6911,
+      "epoch": 0.03,
+      "loss_char": 0.7331,
+      "loss_token": 1.6407,
+      "lr": "2.00e-07",
+      "norm": 5.7705,
+      "step": 1682
+    },
+    {
+      "acc_char": 0.2974,
+      "acc_token": 0.6622,
+      "epoch": 0.03,
+      "loss_char": 0.8001,
+      "loss_token": 1.7329,
+      "lr": "2.00e-07",
+      "norm": 6.0946,
+      "step": 1683
+    },
+    {
+      "acc_char": 0.3164,
+      "acc_token": 0.7132,
+      "epoch": 0.03,
+      "loss_char": 0.6316,
+      "loss_token": 1.4631,
+      "lr": "2.00e-07",
+      "norm": 5.8817,
+      "step": 1684
+    },
+    {
+      "acc_char": 0.3026,
+      "acc_token": 0.6784,
+      "epoch": 0.03,
+      "loss_char": 0.747,
+      "loss_token": 1.6259,
+      "lr": "2.00e-07",
+      "norm": 5.7106,
+      "step": 1685
+    },
+    {
+      "acc_char": 0.3283,
+      "acc_token": 0.735,
+      "epoch": 0.03,
+      "loss_char": 0.598,
+      "loss_token": 1.4537,
+      "lr": "2.00e-07",
+      "norm": 5.9056,
+      "step": 1686
+    },
+    {
+      "acc_char": 0.3011,
+      "acc_token": 0.6876,
+      "epoch": 0.03,
+      "loss_char": 0.6903,
+      "loss_token": 1.608,
+      "lr": "2.00e-07",
+      "norm": 5.6,
+      "step": 1687
+    },
+    {
+      "acc_char": 0.3081,
+      "acc_token": 0.6884,
+      "epoch": 0.03,
+      "loss_char": 0.7476,
+      "loss_token": 1.7288,
+      "lr": "2.00e-07",
+      "norm": 5.8679,
+      "step": 1688
+    },
+    {
+      "acc_char": 0.3007,
+      "acc_token": 0.6842,
+      "epoch": 0.03,
+      "loss_char": 0.7168,
+      "loss_token": 1.5752,
+      "lr": "2.00e-07",
+      "norm": 5.681,
+      "step": 1689
+    },
+    {
+      "acc_char": 0.3368,
+      "acc_token": 0.7385,
+      "epoch": 0.03,
+      "loss_char": 0.5935,
+      "loss_token": 1.556,
+      "lr": "2.00e-07",
+      "norm": 5.8472,
+      "step": 1690
+    },
+    {
+      "acc_char": 0.2926,
+      "acc_token": 0.6757,
+      "epoch": 0.03,
+      "loss_char": 0.7092,
+      "loss_token": 1.627,
+      "lr": "2.00e-07",
+      "norm": 5.8508,
+      "step": 1691
+    },
+    {
+      "acc_char": 0.2891,
+      "acc_token": 0.6339,
+      "epoch": 0.03,
+      "loss_char": 0.8447,
+      "loss_token": 1.7228,
+      "lr": "2.00e-07",
+      "norm": 5.39,
+      "step": 1692
+    },
+    {
+      "acc_char": 0.2962,
+      "acc_token": 0.6806,
+      "epoch": 0.03,
+      "loss_char": 0.7107,
+      "loss_token": 1.6122,
+      "lr": "2.00e-07",
+      "norm": 5.5625,
+      "step": 1693
+    },
+    {
+      "acc_char": 0.2935,
+      "acc_token": 0.6781,
+      "epoch": 0.03,
+      "loss_char": 0.7368,
+      "loss_token": 1.6023,
+      "lr": "2.00e-07",
+      "norm": 5.7586,
+      "step": 1694
+    },
+    {
+      "acc_char": 0.2888,
+      "acc_token": 0.618,
+      "epoch": 0.03,
+      "loss_char": 0.914,
+      "loss_token": 1.7759,
+      "lr": "2.00e-07",
+      "norm": 5.5458,
+      "step": 1695
+    },
+    {
+      "acc_char": 0.2986,
+      "acc_token": 0.6778,
+      "epoch": 0.03,
+      "loss_char": 0.7617,
+      "loss_token": 1.6123,
+      "lr": "2.00e-07",
+      "norm": 5.6041,
+      "step": 1696
+    },
+    {
+      "acc_char": 0.3014,
+      "acc_token": 0.6804,
+      "epoch": 0.03,
+      "loss_char": 0.7763,
+      "loss_token": 1.5783,
+      "lr": "2.00e-07",
+      "norm": 6.4606,
+      "step": 1697
+    },
+    {
+      "acc_char": 0.2949,
+      "acc_token": 0.6779,
+      "epoch": 0.03,
+      "loss_char": 0.7234,
+      "loss_token": 1.6868,
+      "lr": "2.00e-07",
+      "norm": 5.8553,
+      "step": 1698
+    },
+    {
+      "acc_char": 0.3016,
+      "acc_token": 0.6914,
+      "epoch": 0.03,
+      "loss_char": 0.6839,
+      "loss_token": 1.6502,
+      "lr": "2.00e-07",
+      "norm": 5.9714,
+      "step": 1699
+    },
+    {
+      "acc_char": 0.3106,
+      "acc_token": 0.7059,
+      "epoch": 0.03,
+      "loss_char": 0.6671,
+      "loss_token": 1.5837,
+      "lr": "2.00e-07",
+      "norm": 5.6659,
+      "step": 1700
+    },
+    {
+      "acc_char": 0.2735,
+      "acc_token": 0.6861,
+      "epoch": 0.03,
+      "loss_char": 0.6165,
+      "loss_token": 1.5865,
+      "lr": "2.00e-07",
+      "norm": 6.0211,
+      "step": 1701
+    },
+    {
+      "acc_char": 0.3034,
+      "acc_token": 0.6779,
+      "epoch": 0.03,
+      "loss_char": 0.7463,
+      "loss_token": 1.6613,
+      "lr": "2.00e-07",
+      "norm": 5.7788,
+      "step": 1702
+    },
+    {
+      "acc_char": 0.3097,
+      "acc_token": 0.6844,
+      "epoch": 0.03,
+      "loss_char": 0.7017,
+      "loss_token": 1.6179,
+      "lr": "2.00e-07",
+      "norm": 5.996,
+      "step": 1703
+    },
+    {
+      "acc_char": 0.2974,
+      "acc_token": 0.6822,
+      "epoch": 0.03,
+      "loss_char": 0.7026,
+      "loss_token": 1.7367,
+      "lr": "2.00e-07",
+      "norm": 6.6624,
+      "step": 1704
+    },
+    {
+      "acc_char": 0.3138,
+      "acc_token": 0.6773,
+      "epoch": 0.03,
+      "loss_char": 0.7662,
+      "loss_token": 1.6222,
+      "lr": "2.00e-07",
+      "norm": 5.5022,
+      "step": 1705
+    },
+    {
+      "acc_char": 0.2954,
+      "acc_token": 0.663,
+      "epoch": 0.03,
+      "loss_char": 0.7875,
+      "loss_token": 1.5797,
+      "lr": "2.00e-07",
+      "norm": 5.453,
+      "step": 1706
+    },
+    {
+      "acc_char": 0.3159,
+      "acc_token": 0.6957,
+      "epoch": 0.03,
+      "loss_char": 0.7143,
+      "loss_token": 1.5202,
+      "lr": "2.00e-07",
+      "norm": 6.0906,
+      "step": 1707
+    },
+    {
+      "acc_char": 0.314,
+      "acc_token": 0.7008,
+      "epoch": 0.03,
+      "loss_char": 0.6894,
+      "loss_token": 1.5779,
+      "lr": "2.00e-07",
+      "norm": 5.8956,
+      "step": 1708
+    },
+    {
+      "acc_char": 0.3105,
+      "acc_token": 0.6938,
+      "epoch": 0.03,
+      "loss_char": 0.7217,
+      "loss_token": 1.588,
+      "lr": "2.00e-07",
+      "norm": 5.8675,
+      "step": 1709
+    },
+    {
+      "acc_char": 0.3103,
+      "acc_token": 0.6963,
+      "epoch": 0.03,
+      "loss_char": 0.6995,
+      "loss_token": 1.6254,
+      "lr": "2.00e-07",
+      "norm": 5.8286,
+      "step": 1710
+    },
+    {
+      "acc_char": 0.3003,
+      "acc_token": 0.6787,
+      "epoch": 0.03,
+      "loss_char": 0.7472,
+      "loss_token": 1.6933,
+      "lr": "2.00e-07",
+      "norm": 5.8765,
+      "step": 1711
+    },
+    {
+      "acc_char": 0.3151,
+      "acc_token": 0.7065,
+      "epoch": 0.03,
+      "loss_char": 0.667,
+      "loss_token": 1.536,
+      "lr": "2.00e-07",
+      "norm": 5.7391,
+      "step": 1712
+    },
+    {
+      "acc_char": 0.3097,
+      "acc_token": 0.699,
+      "epoch": 0.03,
+      "loss_char": 0.6717,
+      "loss_token": 1.5737,
+      "lr": "2.00e-07",
+      "norm": 5.6138,
+      "step": 1713
+    },
+    {
+      "acc_char": 0.316,
+      "acc_token": 0.7113,
+      "epoch": 0.03,
+      "loss_char": 0.6721,
+      "loss_token": 1.5796,
+      "lr": "2.00e-07",
+      "norm": 5.8443,
+      "step": 1714
+    },
+    {
+      "acc_char": 0.305,
+      "acc_token": 0.6889,
+      "epoch": 0.03,
+      "loss_char": 0.724,
+      "loss_token": 1.5748,
+      "lr": "2.00e-07",
+      "norm": 5.5035,
+      "step": 1715
+    },
+    {
+      "acc_char": 0.3032,
+      "acc_token": 0.7013,
+      "epoch": 0.03,
+      "loss_char": 0.6446,
+      "loss_token": 1.5459,
+      "lr": "2.00e-07",
+      "norm": 5.642,
+      "step": 1716
+    },
+    {
+      "acc_char": 0.2952,
+      "acc_token": 0.688,
+      "epoch": 0.03,
+      "loss_char": 0.6734,
+      "loss_token": 1.5776,
+      "lr": "2.00e-07",
+      "norm": 5.6175,
+      "step": 1717
+    },
+    {
+      "acc_char": 0.2981,
+      "acc_token": 0.6796,
+      "epoch": 0.03,
+      "loss_char": 0.7212,
+      "loss_token": 1.6105,
+      "lr": "2.00e-07",
+      "norm": 6.1429,
+      "step": 1718
+    },
+    {
+      "acc_char": 0.2959,
+      "acc_token": 0.6775,
+      "epoch": 0.03,
+      "loss_char": 0.7264,
+      "loss_token": 1.6338,
+      "lr": "2.00e-07",
+      "norm": 5.5765,
+      "step": 1719
+    },
+    {
+      "acc_char": 0.3074,
+      "acc_token": 0.6851,
+      "epoch": 0.03,
+      "loss_char": 0.7234,
+      "loss_token": 1.612,
+      "lr": "2.00e-07",
+      "norm": 5.7202,
+      "step": 1720
+    },
+    {
+      "acc_char": 0.2906,
+      "acc_token": 0.6583,
+      "epoch": 0.03,
+      "loss_char": 0.8232,
+      "loss_token": 1.7554,
+      "lr": "2.00e-07",
+      "norm": 6.091,
+      "step": 1721
+    },
+    {
+      "acc_char": 0.3159,
+      "acc_token": 0.7077,
+      "epoch": 0.03,
+      "loss_char": 0.6617,
+      "loss_token": 1.4657,
+      "lr": "2.00e-07",
+      "norm": 5.3041,
+      "step": 1722
+    },
+    {
+      "acc_char": 0.2902,
+      "acc_token": 0.6618,
+      "epoch": 0.03,
+      "loss_char": 0.7708,
+      "loss_token": 1.6204,
+      "lr": "2.00e-07",
+      "norm": 5.605,
+      "step": 1723
+    },
+    {
+      "acc_char": 0.3085,
+      "acc_token": 0.6796,
+      "epoch": 0.03,
+      "loss_char": 0.7645,
+      "loss_token": 1.612,
+      "lr": "2.00e-07",
+      "norm": 5.7922,
+      "step": 1724
+    },
+    {
+      "acc_char": 0.2971,
+      "acc_token": 0.6404,
+      "epoch": 0.03,
+      "loss_char": 0.8225,
+      "loss_token": 1.6631,
+      "lr": "2.00e-07",
+      "norm": 5.5715,
+      "step": 1725
+    },
+    {
+      "acc_char": 0.319,
+      "acc_token": 0.6897,
+      "epoch": 0.03,
+      "loss_char": 0.6891,
+      "loss_token": 1.5501,
+      "lr": "2.00e-07",
+      "norm": 5.5443,
+      "step": 1726
+    },
+    {
+      "acc_char": 0.3164,
+      "acc_token": 0.7073,
+      "epoch": 0.03,
+      "loss_char": 0.6799,
+      "loss_token": 1.592,
+      "lr": "2.00e-07",
+      "norm": 5.5647,
+      "step": 1727
+    },
+    {
+      "acc_char": 0.3133,
+      "acc_token": 0.7051,
+      "epoch": 0.03,
+      "loss_char": 0.661,
+      "loss_token": 1.6084,
+      "lr": "2.00e-07",
+      "norm": 5.8548,
+      "step": 1728
+    },
+    {
+      "acc_char": 0.3017,
+      "acc_token": 0.6912,
+      "epoch": 0.03,
+      "loss_char": 0.694,
+      "loss_token": 1.7379,
+      "lr": "2.00e-07",
+      "norm": 6.0479,
+      "step": 1729
+    },
+    {
+      "acc_char": 0.3022,
+      "acc_token": 0.6878,
+      "epoch": 0.03,
+      "loss_char": 0.699,
+      "loss_token": 1.6654,
+      "lr": "2.00e-07",
+      "norm": 5.7858,
+      "step": 1730
+    },
+    {
+      "acc_char": 0.3061,
+      "acc_token": 0.6872,
+      "epoch": 0.03,
+      "loss_char": 0.7225,
+      "loss_token": 1.5633,
+      "lr": "2.00e-07",
+      "norm": 5.5348,
+      "step": 1731
+    },
+    {
+      "acc_char": 0.2839,
+      "acc_token": 0.6558,
+      "epoch": 0.03,
+      "loss_char": 0.7938,
+      "loss_token": 1.6926,
+      "lr": "2.00e-07",
+      "norm": 6.0785,
+      "step": 1732
+    },
+    {
+      "acc_char": 0.2934,
+      "acc_token": 0.6419,
+      "epoch": 0.03,
+      "loss_char": 0.8494,
+      "loss_token": 1.7951,
+      "lr": "2.00e-07",
+      "norm": 5.7669,
+      "step": 1733
+    },
+    {
+      "acc_char": 0.2941,
+      "acc_token": 0.6791,
+      "epoch": 0.03,
+      "loss_char": 0.7158,
+      "loss_token": 1.6679,
+      "lr": "2.00e-07",
+      "norm": 5.7,
+      "step": 1734
+    },
+    {
+      "acc_char": 0.3115,
+      "acc_token": 0.6936,
+      "epoch": 0.03,
+      "loss_char": 0.7035,
+      "loss_token": 1.5711,
+      "lr": "2.00e-07",
+      "norm": 5.7193,
+      "step": 1735
+    },
+    {
+      "acc_char": 0.3071,
+      "acc_token": 0.7071,
+      "epoch": 0.03,
+      "loss_char": 0.6319,
+      "loss_token": 1.519,
+      "lr": "2.00e-07",
+      "norm": 5.7064,
+      "step": 1736
+    },
+    {
+      "acc_char": 0.3034,
+      "acc_token": 0.6798,
+      "epoch": 0.03,
+      "loss_char": 0.7287,
+      "loss_token": 1.5719,
+      "lr": "2.00e-07",
+      "norm": 5.7227,
+      "step": 1737
+    },
+    {
+      "acc_char": 0.2943,
+      "acc_token": 0.6784,
+      "epoch": 0.03,
+      "loss_char": 0.728,
+      "loss_token": 1.6606,
+      "lr": "2.00e-07",
+      "norm": 5.7986,
+      "step": 1738
+    },
+    {
+      "acc_char": 0.3183,
+      "acc_token": 0.6948,
+      "epoch": 0.03,
+      "loss_char": 0.7256,
+      "loss_token": 1.5931,
+      "lr": "2.00e-07",
+      "norm": 5.6669,
+      "step": 1739
+    },
+    {
+      "acc_char": 0.3087,
+      "acc_token": 0.6671,
+      "epoch": 0.03,
+      "loss_char": 0.78,
+      "loss_token": 1.7101,
+      "lr": "2.00e-07",
+      "norm": 5.8135,
+      "step": 1740
+    },
+    {
+      "acc_char": 0.3024,
+      "acc_token": 0.6934,
+      "epoch": 0.03,
+      "loss_char": 0.6551,
+      "loss_token": 1.5156,
+      "lr": "2.00e-07",
+      "norm": 6.3682,
+      "step": 1741
+    },
+    {
+      "acc_char": 0.3054,
+      "acc_token": 0.6736,
+      "epoch": 0.03,
+      "loss_char": 0.7574,
+      "loss_token": 1.5606,
+      "lr": "2.00e-07",
+      "norm": 5.8203,
+      "step": 1742
+    },
+    {
+      "acc_char": 0.307,
+      "acc_token": 0.6946,
+      "epoch": 0.03,
+      "loss_char": 0.6862,
+      "loss_token": 1.5647,
+      "lr": "2.00e-07",
+      "norm": 5.7305,
+      "step": 1743
+    },
+    {
+      "acc_char": 0.2871,
+      "acc_token": 0.6231,
+      "epoch": 0.03,
+      "loss_char": 0.884,
+      "loss_token": 1.819,
+      "lr": "2.00e-07",
+      "norm": 5.9354,
+      "step": 1744
+    },
+    {
+      "acc_char": 0.2858,
+      "acc_token": 0.6322,
+      "epoch": 0.03,
+      "loss_char": 0.8419,
+      "loss_token": 1.6682,
+      "lr": "2.00e-07",
+      "norm": 5.5717,
+      "step": 1745
+    },
+    {
+      "acc_char": 0.2926,
+      "acc_token": 0.6643,
+      "epoch": 0.03,
+      "loss_char": 0.7732,
+      "loss_token": 1.718,
+      "lr": "2.00e-07",
+      "norm": 5.7846,
+      "step": 1746
+    },
+    {
+      "acc_char": 0.2998,
+      "acc_token": 0.6812,
+      "epoch": 0.03,
+      "loss_char": 0.7343,
+      "loss_token": 1.6174,
+      "lr": "2.00e-07",
+      "norm": 5.7623,
+      "step": 1747
+    },
+    {
+      "acc_char": 0.2984,
+      "acc_token": 0.6937,
+      "epoch": 0.03,
+      "loss_char": 0.6639,
+      "loss_token": 1.5517,
+      "lr": "2.00e-07",
+      "norm": 5.6115,
+      "step": 1748
+    },
+    {
+      "acc_char": 0.3132,
+      "acc_token": 0.6951,
+      "epoch": 0.03,
+      "loss_char": 0.708,
+      "loss_token": 1.5821,
+      "lr": "2.00e-07",
+      "norm": 5.8594,
+      "step": 1749
+    },
+    {
+      "acc_char": 0.2966,
+      "acc_token": 0.6765,
+      "epoch": 0.03,
+      "loss_char": 0.7249,
+      "loss_token": 1.6193,
+      "lr": "2.00e-07",
+      "norm": 5.9276,
+      "step": 1750
+    },
+    {
+      "acc_char": 0.2956,
+      "acc_token": 0.6666,
+      "epoch": 0.03,
+      "loss_char": 0.7695,
+      "loss_token": 1.6098,
+      "lr": "2.00e-07",
+      "norm": 5.9244,
+      "step": 1751
+    },
+    {
+      "acc_char": 0.2856,
+      "acc_token": 0.6745,
+      "epoch": 0.03,
+      "loss_char": 0.7153,
+      "loss_token": 1.6257,
+      "lr": "2.00e-07",
+      "norm": 6.2479,
+      "step": 1752
+    },
+    {
+      "acc_char": 0.3144,
+      "acc_token": 0.716,
+      "epoch": 0.03,
+      "loss_char": 0.6248,
+      "loss_token": 1.4773,
+      "lr": "2.00e-07",
+      "norm": 5.328,
+      "step": 1753
+    },
+    {
+      "acc_char": 0.2995,
+      "acc_token": 0.6832,
+      "epoch": 0.03,
+      "loss_char": 0.706,
+      "loss_token": 1.6197,
+      "lr": "2.00e-07",
+      "norm": 5.9709,
+      "step": 1754
+    },
+    {
+      "acc_char": 0.2874,
+      "acc_token": 0.6328,
+      "epoch": 0.03,
+      "loss_char": 0.8373,
+      "loss_token": 1.7249,
+      "lr": "2.00e-07",
+      "norm": 5.7176,
+      "step": 1755
+    },
+    {
+      "acc_char": 0.2957,
+      "acc_token": 0.6702,
+      "epoch": 0.03,
+      "loss_char": 0.7423,
+      "loss_token": 1.6452,
+      "lr": "2.00e-07",
+      "norm": 5.7704,
+      "step": 1756
+    },
+    {
+      "acc_char": 0.2989,
+      "acc_token": 0.6776,
+      "epoch": 0.03,
+      "loss_char": 0.7342,
+      "loss_token": 1.6687,
+      "lr": "2.00e-07",
+      "norm": 5.7341,
+      "step": 1757
+    },
+    {
+      "acc_char": 0.3055,
+      "acc_token": 0.7035,
+      "epoch": 0.03,
+      "loss_char": 0.6279,
+      "loss_token": 1.5377,
+      "lr": "2.00e-07",
+      "norm": 5.5632,
+      "step": 1758
+    },
+    {
+      "acc_char": 0.2955,
+      "acc_token": 0.6865,
+      "epoch": 0.03,
+      "loss_char": 0.6719,
+      "loss_token": 1.6061,
+      "lr": "2.00e-07",
+      "norm": 5.8618,
+      "step": 1759
+    },
+    {
+      "acc_char": 0.2821,
+      "acc_token": 0.6584,
+      "epoch": 0.03,
+      "loss_char": 0.7972,
+      "loss_token": 1.7979,
+      "lr": "1.99e-07",
+      "norm": 6.072,
+      "step": 1760
+    },
+    {
+      "acc_char": 0.3088,
+      "acc_token": 0.6928,
+      "epoch": 0.03,
+      "loss_char": 0.6977,
+      "loss_token": 1.5895,
+      "lr": "1.99e-07",
+      "norm": 6.2279,
+      "step": 1761
+    },
+    {
+      "acc_char": 0.3033,
+      "acc_token": 0.6922,
+      "epoch": 0.03,
+      "loss_char": 0.6863,
+      "loss_token": 1.6013,
+      "lr": "1.99e-07",
+      "norm": 5.6444,
+      "step": 1762
+    },
+    {
+      "acc_char": 0.2986,
+      "acc_token": 0.6751,
+      "epoch": 0.03,
+      "loss_char": 0.7423,
+      "loss_token": 1.6667,
+      "lr": "1.99e-07",
+      "norm": 5.7379,
+      "step": 1763
+    },
+    {
+      "acc_char": 0.297,
+      "acc_token": 0.6774,
+      "epoch": 0.03,
+      "loss_char": 0.7249,
+      "loss_token": 1.6749,
+      "lr": "1.99e-07",
+      "norm": 6.1533,
+      "step": 1764
+    },
+    {
+      "acc_char": 0.2982,
+      "acc_token": 0.6805,
+      "epoch": 0.03,
+      "loss_char": 0.7077,
+      "loss_token": 1.5948,
+      "lr": "1.99e-07",
+      "norm": 5.8114,
+      "step": 1765
+    },
+    {
+      "acc_char": 0.3067,
+      "acc_token": 0.6859,
+      "epoch": 0.03,
+      "loss_char": 0.7297,
+      "loss_token": 1.6133,
+      "lr": "1.99e-07",
+      "norm": 5.5519,
+      "step": 1766
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6767,
+      "epoch": 0.03,
+      "loss_char": 0.7591,
+      "loss_token": 1.6389,
+      "lr": "1.99e-07",
+      "norm": 5.7607,
+      "step": 1767
+    },
+    {
+      "acc_char": 0.3089,
+      "acc_token": 0.7008,
+      "epoch": 0.03,
+      "loss_char": 0.6708,
+      "loss_token": 1.6271,
+      "lr": "1.99e-07",
+      "norm": 5.9782,
+      "step": 1768
+    },
+    {
+      "acc_char": 0.2963,
+      "acc_token": 0.6712,
+      "epoch": 0.03,
+      "loss_char": 0.7529,
+      "loss_token": 1.6849,
+      "lr": "1.99e-07",
+      "norm": 6.1694,
+      "step": 1769
+    },
+    {
+      "acc_char": 0.2967,
+      "acc_token": 0.6822,
+      "epoch": 0.03,
+      "loss_char": 0.7043,
+      "loss_token": 1.525,
+      "lr": "1.99e-07",
+      "norm": 5.579,
+      "step": 1770
+    },
+    {
+      "acc_char": 0.3024,
+      "acc_token": 0.6872,
+      "epoch": 0.03,
+      "loss_char": 0.7056,
+      "loss_token": 1.5849,
+      "lr": "1.99e-07",
+      "norm": 5.9316,
+      "step": 1771
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.6997,
+      "epoch": 0.03,
+      "loss_char": 0.6743,
+      "loss_token": 1.5539,
+      "lr": "1.99e-07",
+      "norm": 5.6747,
+      "step": 1772
+    },
+    {
+      "acc_char": 0.3042,
+      "acc_token": 0.679,
+      "epoch": 0.03,
+      "loss_char": 0.7526,
+      "loss_token": 1.6509,
+      "lr": "1.99e-07",
+      "norm": 5.8773,
+      "step": 1773
+    },
+    {
+      "acc_char": 0.3031,
+      "acc_token": 0.6836,
+      "epoch": 0.03,
+      "loss_char": 0.7267,
+      "loss_token": 1.6467,
+      "lr": "1.99e-07",
+      "norm": 5.8952,
+      "step": 1774
+    },
+    {
+      "acc_char": 0.3017,
+      "acc_token": 0.6808,
+      "epoch": 0.03,
+      "loss_char": 0.7203,
+      "loss_token": 1.6047,
+      "lr": "1.99e-07",
+      "norm": 5.8058,
+      "step": 1775
+    },
+    {
+      "acc_char": 0.303,
+      "acc_token": 0.6802,
+      "epoch": 0.03,
+      "loss_char": 0.7312,
+      "loss_token": 1.6448,
+      "lr": "1.99e-07",
+      "norm": 6.1805,
+      "step": 1776
+    },
+    {
+      "acc_char": 0.3019,
+      "acc_token": 0.6815,
+      "epoch": 0.03,
+      "loss_char": 0.7149,
+      "loss_token": 1.6754,
+      "lr": "1.99e-07",
+      "norm": 6.1277,
+      "step": 1777
+    },
+    {
+      "acc_char": 0.2831,
+      "acc_token": 0.6658,
+      "epoch": 0.03,
+      "loss_char": 0.7458,
+      "loss_token": 1.6643,
+      "lr": "1.99e-07",
+      "norm": 5.6969,
+      "step": 1778
+    },
+    {
+      "acc_char": 0.3091,
+      "acc_token": 0.6912,
+      "epoch": 0.03,
+      "loss_char": 0.7104,
+      "loss_token": 1.5921,
+      "lr": "1.99e-07",
+      "norm": 6.0038,
+      "step": 1779
+    },
+    {
+      "acc_char": 0.2966,
+      "acc_token": 0.6718,
+      "epoch": 0.03,
+      "loss_char": 0.7499,
+      "loss_token": 1.6861,
+      "lr": "1.99e-07",
+      "norm": 5.9677,
+      "step": 1780
+    },
+    {
+      "acc_char": 0.2973,
+      "acc_token": 0.6715,
+      "epoch": 0.03,
+      "loss_char": 0.7577,
+      "loss_token": 1.675,
+      "lr": "1.99e-07",
+      "norm": 5.8496,
+      "step": 1781
+    },
+    {
+      "acc_char": 0.3088,
+      "acc_token": 0.6957,
+      "epoch": 0.03,
+      "loss_char": 0.6928,
+      "loss_token": 1.4824,
+      "lr": "1.99e-07",
+      "norm": 5.617,
+      "step": 1782
+    },
+    {
+      "acc_char": 0.3093,
+      "acc_token": 0.6729,
+      "epoch": 0.03,
+      "loss_char": 0.7746,
+      "loss_token": 1.6802,
+      "lr": "1.99e-07",
+      "norm": 5.9807,
+      "step": 1783
+    },
+    {
+      "acc_char": 0.3068,
+      "acc_token": 0.6768,
+      "epoch": 0.03,
+      "loss_char": 0.7679,
+      "loss_token": 1.6124,
+      "lr": "1.99e-07",
+      "norm": 5.9877,
+      "step": 1784
+    },
+    {
+      "acc_char": 0.3244,
+      "acc_token": 0.7239,
+      "epoch": 0.03,
+      "loss_char": 0.6387,
+      "loss_token": 1.4429,
+      "lr": "1.99e-07",
+      "norm": 5.3048,
+      "step": 1785
+    },
+    {
+      "acc_char": 0.3015,
+      "acc_token": 0.6626,
+      "epoch": 0.03,
+      "loss_char": 0.8047,
+      "loss_token": 1.6889,
+      "lr": "1.99e-07",
+      "norm": 5.7797,
+      "step": 1786
+    },
+    {
+      "acc_char": 0.2968,
+      "acc_token": 0.6674,
+      "epoch": 0.03,
+      "loss_char": 0.7591,
+      "loss_token": 1.6156,
+      "lr": "1.99e-07",
+      "norm": 5.6749,
+      "step": 1787
+    },
+    {
+      "acc_char": 0.2884,
+      "acc_token": 0.6264,
+      "epoch": 0.03,
+      "loss_char": 0.8927,
+      "loss_token": 1.7878,
+      "lr": "1.99e-07",
+      "norm": 5.9712,
+      "step": 1788
+    },
+    {
+      "acc_char": 0.3074,
+      "acc_token": 0.7018,
+      "epoch": 0.03,
+      "loss_char": 0.6909,
+      "loss_token": 1.5815,
+      "lr": "1.99e-07",
+      "norm": 5.7451,
+      "step": 1789
+    },
+    {
+      "acc_char": 0.2977,
+      "acc_token": 0.6778,
+      "epoch": 0.03,
+      "loss_char": 0.7358,
+      "loss_token": 1.6576,
+      "lr": "1.99e-07",
+      "norm": 6.4111,
+      "step": 1790
+    },
+    {
+      "acc_char": 0.3044,
+      "acc_token": 0.6762,
+      "epoch": 0.03,
+      "loss_char": 0.7493,
+      "loss_token": 1.6757,
+      "lr": "1.99e-07",
+      "norm": 5.9979,
+      "step": 1791
+    },
+    {
+      "acc_char": 0.2726,
+      "acc_token": 0.5943,
+      "epoch": 0.03,
+      "loss_char": 0.9682,
+      "loss_token": 1.8345,
+      "lr": "1.99e-07",
+      "norm": 5.644,
+      "step": 1792
+    },
+    {
+      "acc_char": 0.3091,
+      "acc_token": 0.7004,
+      "epoch": 0.03,
+      "loss_char": 0.6841,
+      "loss_token": 1.5326,
+      "lr": "1.99e-07",
+      "norm": 5.5105,
+      "step": 1793
+    },
+    {
+      "acc_char": 0.3115,
+      "acc_token": 0.7085,
+      "epoch": 0.03,
+      "loss_char": 0.6296,
+      "loss_token": 1.5025,
+      "lr": "1.99e-07",
+      "norm": 5.435,
+      "step": 1794
+    },
+    {
+      "acc_char": 0.2962,
+      "acc_token": 0.6761,
+      "epoch": 0.03,
+      "loss_char": 0.7292,
+      "loss_token": 1.6037,
+      "lr": "1.99e-07",
+      "norm": 5.7251,
+      "step": 1795
+    },
+    {
+      "acc_char": 0.3096,
+      "acc_token": 0.7037,
+      "epoch": 0.03,
+      "loss_char": 0.662,
+      "loss_token": 1.6494,
+      "lr": "1.99e-07",
+      "norm": 6.0807,
+      "step": 1796
+    },
+    {
+      "acc_char": 0.3063,
+      "acc_token": 0.6889,
+      "epoch": 0.03,
+      "loss_char": 0.7193,
+      "loss_token": 1.5482,
+      "lr": "1.99e-07",
+      "norm": 5.6579,
+      "step": 1797
+    },
+    {
+      "acc_char": 0.2851,
+      "acc_token": 0.6324,
+      "epoch": 0.03,
+      "loss_char": 0.8989,
+      "loss_token": 1.8132,
+      "lr": "1.99e-07",
+      "norm": 5.7828,
+      "step": 1798
+    },
+    {
+      "acc_char": 0.3053,
+      "acc_token": 0.6844,
+      "epoch": 0.03,
+      "loss_char": 0.7401,
+      "loss_token": 1.6022,
+      "lr": "1.99e-07",
+      "norm": 5.6144,
+      "step": 1799
+    },
+    {
+      "acc_char": 0.3087,
+      "acc_token": 0.69,
+      "epoch": 0.03,
+      "loss_char": 0.7005,
+      "loss_token": 1.538,
+      "lr": "1.99e-07",
+      "norm": 5.6592,
+      "step": 1800
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 54626,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "total_flos": 93278636605440.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}