diff --git "a/checkpoint-4000/trainer_state.json" "b/checkpoint-4000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-4000/trainer_state.json"
@@ -0,0 +1,28033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.35555555555555557,
+  "eval_steps": 500,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8.888888888888889e-05,
+      "grad_norm": 231.0,
+      "learning_rate": 5.9259259259259265e-09,
+      "loss": 2.8262,
+      "step": 1
+    },
+    {
+      "epoch": 0.00017777777777777779,
+      "grad_norm": 201.0,
+      "learning_rate": 1.1851851851851853e-08,
+      "loss": 2.9179,
+      "step": 2
+    },
+    {
+      "epoch": 0.0002666666666666667,
+      "grad_norm": 202.0,
+      "learning_rate": 1.777777777777778e-08,
+      "loss": 2.9485,
+      "step": 3
+    },
+    {
+      "epoch": 0.00035555555555555557,
+      "grad_norm": 228.0,
+      "learning_rate": 2.3703703703703706e-08,
+      "loss": 2.8092,
+      "step": 4
+    },
+    {
+      "epoch": 0.00044444444444444447,
+      "grad_norm": 222.0,
+      "learning_rate": 2.9629629629629632e-08,
+      "loss": 2.7296,
+      "step": 5
+    },
+    {
+      "epoch": 0.0005333333333333334,
+      "grad_norm": 205.0,
+      "learning_rate": 3.555555555555556e-08,
+      "loss": 3.0451,
+      "step": 6
+    },
+    {
+      "epoch": 0.0006222222222222223,
+      "grad_norm": 248.0,
+      "learning_rate": 4.148148148148148e-08,
+      "loss": 3.0115,
+      "step": 7
+    },
+    {
+      "epoch": 0.0007111111111111111,
+      "grad_norm": 227.0,
+      "learning_rate": 4.740740740740741e-08,
+      "loss": 2.988,
+      "step": 8
+    },
+    {
+      "epoch": 0.0008,
+      "grad_norm": 354.0,
+      "learning_rate": 5.3333333333333334e-08,
+      "loss": 2.6447,
+      "step": 9
+    },
+    {
+      "epoch": 0.0008888888888888889,
+      "grad_norm": 358.0,
+      "learning_rate": 5.9259259259259263e-08,
+      "loss": 2.8176,
+      "step": 10
+    },
+    {
+      "epoch": 0.0009777777777777777,
+      "grad_norm": 322.0,
+      "learning_rate": 6.518518518518519e-08,
+      "loss": 2.9887,
+      "step": 11
+    },
+    {
+      "epoch": 0.0010666666666666667,
+      "grad_norm": 210.0,
+      "learning_rate": 7.111111111111112e-08,
+      "loss": 2.8252,
+      "step": 12
+    },
+    {
+      "epoch": 0.0011555555555555555,
+      "grad_norm": 240.0,
+      "learning_rate": 7.703703703703705e-08,
+      "loss": 2.6588,
+      "step": 13
+    },
+    {
+      "epoch": 0.0012444444444444445,
+      "grad_norm": 158.0,
+      "learning_rate": 8.296296296296297e-08,
+      "loss": 2.8137,
+      "step": 14
+    },
+    {
+      "epoch": 0.0013333333333333333,
+      "grad_norm": 231.0,
+      "learning_rate": 8.88888888888889e-08,
+      "loss": 2.7542,
+      "step": 15
+    },
+    {
+      "epoch": 0.0014222222222222223,
+      "grad_norm": 390.0,
+      "learning_rate": 9.481481481481482e-08,
+      "loss": 2.6801,
+      "step": 16
+    },
+    {
+      "epoch": 0.001511111111111111,
+      "grad_norm": 338.0,
+      "learning_rate": 1.0074074074074075e-07,
+      "loss": 2.8202,
+      "step": 17
+    },
+    {
+      "epoch": 0.0016,
+      "grad_norm": 408.0,
+      "learning_rate": 1.0666666666666667e-07,
+      "loss": 2.5446,
+      "step": 18
+    },
+    {
+      "epoch": 0.0016888888888888889,
+      "grad_norm": 348.0,
+      "learning_rate": 1.125925925925926e-07,
+      "loss": 2.9478,
+      "step": 19
+    },
+    {
+      "epoch": 0.0017777777777777779,
+      "grad_norm": 306.0,
+      "learning_rate": 1.1851851851851853e-07,
+      "loss": 2.9068,
+      "step": 20
+    },
+    {
+      "epoch": 0.0018666666666666666,
+      "grad_norm": 410.0,
+      "learning_rate": 1.2444444444444446e-07,
+      "loss": 2.6607,
+      "step": 21
+    },
+    {
+      "epoch": 0.0019555555555555554,
+      "grad_norm": 524.0,
+      "learning_rate": 1.3037037037037038e-07,
+      "loss": 2.8778,
+      "step": 22
+    },
+    {
+      "epoch": 0.0020444444444444447,
+      "grad_norm": 248.0,
+      "learning_rate": 1.3629629629629631e-07,
+      "loss": 2.8313,
+      "step": 23
+    },
+    {
+      "epoch": 0.0021333333333333334,
+      "grad_norm": 370.0,
+      "learning_rate": 1.4222222222222224e-07,
+      "loss": 2.6689,
+      "step": 24
+    },
+    {
+      "epoch": 0.0022222222222222222,
+      "grad_norm": 300.0,
+      "learning_rate": 1.4814814814814817e-07,
+      "loss": 2.7834,
+      "step": 25
+    },
+    {
+      "epoch": 0.002311111111111111,
+      "grad_norm": 216.0,
+      "learning_rate": 1.540740740740741e-07,
+      "loss": 2.6983,
+      "step": 26
+    },
+    {
+      "epoch": 0.0024,
+      "grad_norm": 266.0,
+      "learning_rate": 1.6e-07,
+      "loss": 3.0001,
+      "step": 27
+    },
+    {
+      "epoch": 0.002488888888888889,
+      "grad_norm": 628.0,
+      "learning_rate": 1.6592592592592593e-07,
+      "loss": 2.846,
+      "step": 28
+    },
+    {
+      "epoch": 0.002577777777777778,
+      "grad_norm": 414.0,
+      "learning_rate": 1.7185185185185186e-07,
+      "loss": 2.7031,
+      "step": 29
+    },
+    {
+      "epoch": 0.0026666666666666666,
+      "grad_norm": 588.0,
+      "learning_rate": 1.777777777777778e-07,
+      "loss": 2.8667,
+      "step": 30
+    },
+    {
+      "epoch": 0.0027555555555555554,
+      "grad_norm": 253.0,
+      "learning_rate": 1.8370370370370372e-07,
+      "loss": 2.7814,
+      "step": 31
+    },
+    {
+      "epoch": 0.0028444444444444446,
+      "grad_norm": 430.0,
+      "learning_rate": 1.8962962962962965e-07,
+      "loss": 2.7724,
+      "step": 32
+    },
+    {
+      "epoch": 0.0029333333333333334,
+      "grad_norm": 314.0,
+      "learning_rate": 1.9555555555555558e-07,
+      "loss": 2.6213,
+      "step": 33
+    },
+    {
+      "epoch": 0.003022222222222222,
+      "grad_norm": 540.0,
+      "learning_rate": 2.014814814814815e-07,
+      "loss": 2.6998,
+      "step": 34
+    },
+    {
+      "epoch": 0.003111111111111111,
+      "grad_norm": 928.0,
+      "learning_rate": 2.074074074074074e-07,
+      "loss": 2.6746,
+      "step": 35
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 564.0,
+      "learning_rate": 2.1333333333333334e-07,
+      "loss": 2.8693,
+      "step": 36
+    },
+    {
+      "epoch": 0.003288888888888889,
+      "grad_norm": 752.0,
+      "learning_rate": 2.1925925925925927e-07,
+      "loss": 2.7987,
+      "step": 37
+    },
+    {
+      "epoch": 0.0033777777777777777,
+      "grad_norm": 788.0,
+      "learning_rate": 2.251851851851852e-07,
+      "loss": 2.7889,
+      "step": 38
+    },
+    {
+      "epoch": 0.0034666666666666665,
+      "grad_norm": 356.0,
+      "learning_rate": 2.3111111111111112e-07,
+      "loss": 3.028,
+      "step": 39
+    },
+    {
+      "epoch": 0.0035555555555555557,
+      "grad_norm": 516.0,
+      "learning_rate": 2.3703703703703705e-07,
+      "loss": 2.7597,
+      "step": 40
+    },
+    {
+      "epoch": 0.0036444444444444445,
+      "grad_norm": 740.0,
+      "learning_rate": 2.4296296296296296e-07,
+      "loss": 2.648,
+      "step": 41
+    },
+    {
+      "epoch": 0.0037333333333333333,
+      "grad_norm": 258.0,
+      "learning_rate": 2.488888888888889e-07,
+      "loss": 2.8626,
+      "step": 42
+    },
+    {
+      "epoch": 0.003822222222222222,
+      "grad_norm": 456.0,
+      "learning_rate": 2.548148148148148e-07,
+      "loss": 2.8472,
+      "step": 43
+    },
+    {
+      "epoch": 0.003911111111111111,
+      "grad_norm": 458.0,
+      "learning_rate": 2.6074074074074077e-07,
+      "loss": 2.8322,
+      "step": 44
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 414.0,
+      "learning_rate": 2.666666666666667e-07,
+      "loss": 2.9897,
+      "step": 45
+    },
+    {
+      "epoch": 0.004088888888888889,
+      "grad_norm": 235.0,
+      "learning_rate": 2.7259259259259263e-07,
+      "loss": 2.7264,
+      "step": 46
+    },
+    {
+      "epoch": 0.004177777777777778,
+      "grad_norm": 202.0,
+      "learning_rate": 2.7851851851851853e-07,
+      "loss": 2.7972,
+      "step": 47
+    },
+    {
+      "epoch": 0.004266666666666667,
+      "grad_norm": 280.0,
+      "learning_rate": 2.844444444444445e-07,
+      "loss": 2.512,
+      "step": 48
+    },
+    {
+      "epoch": 0.004355555555555555,
+      "grad_norm": 668.0,
+      "learning_rate": 2.903703703703704e-07,
+      "loss": 2.6697,
+      "step": 49
+    },
+    {
+      "epoch": 0.0044444444444444444,
+      "grad_norm": 306.0,
+      "learning_rate": 2.9629629629629634e-07,
+      "loss": 2.9175,
+      "step": 50
+    },
+    {
+      "epoch": 0.004533333333333334,
+      "grad_norm": 256.0,
+      "learning_rate": 3.0222222222222225e-07,
+      "loss": 2.7045,
+      "step": 51
+    },
+    {
+      "epoch": 0.004622222222222222,
+      "grad_norm": 364.0,
+      "learning_rate": 3.081481481481482e-07,
+      "loss": 2.7193,
+      "step": 52
+    },
+    {
+      "epoch": 0.004711111111111111,
+      "grad_norm": 214.0,
+      "learning_rate": 3.1407407407407405e-07,
+      "loss": 2.79,
+      "step": 53
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 249.0,
+      "learning_rate": 3.2e-07,
+      "loss": 2.8762,
+      "step": 54
+    },
+    {
+      "epoch": 0.004888888888888889,
+      "grad_norm": 460.0,
+      "learning_rate": 3.259259259259259e-07,
+      "loss": 3.0873,
+      "step": 55
+    },
+    {
+      "epoch": 0.004977777777777778,
+      "grad_norm": 450.0,
+      "learning_rate": 3.3185185185185186e-07,
+      "loss": 2.6331,
+      "step": 56
+    },
+    {
+      "epoch": 0.005066666666666666,
+      "grad_norm": 552.0,
+      "learning_rate": 3.3777777777777777e-07,
+      "loss": 2.6765,
+      "step": 57
+    },
+    {
+      "epoch": 0.005155555555555556,
+      "grad_norm": 700.0,
+      "learning_rate": 3.437037037037037e-07,
+      "loss": 2.7971,
+      "step": 58
+    },
+    {
+      "epoch": 0.005244444444444445,
+      "grad_norm": 436.0,
+      "learning_rate": 3.496296296296296e-07,
+      "loss": 2.9912,
+      "step": 59
+    },
+    {
+      "epoch": 0.005333333333333333,
+      "grad_norm": 342.0,
+      "learning_rate": 3.555555555555556e-07,
+      "loss": 2.7947,
+      "step": 60
+    },
+    {
+      "epoch": 0.005422222222222222,
+      "grad_norm": 239.0,
+      "learning_rate": 3.6148148148148154e-07,
+      "loss": 2.994,
+      "step": 61
+    },
+    {
+      "epoch": 0.005511111111111111,
+      "grad_norm": 217.0,
+      "learning_rate": 3.6740740740740744e-07,
+      "loss": 3.1475,
+      "step": 62
+    },
+    {
+      "epoch": 0.0056,
+      "grad_norm": 231.0,
+      "learning_rate": 3.733333333333334e-07,
+      "loss": 2.6752,
+      "step": 63
+    },
+    {
+      "epoch": 0.005688888888888889,
+      "grad_norm": 304.0,
+      "learning_rate": 3.792592592592593e-07,
+      "loss": 2.7891,
+      "step": 64
+    },
+    {
+      "epoch": 0.0057777777777777775,
+      "grad_norm": 161.0,
+      "learning_rate": 3.8518518518518525e-07,
+      "loss": 2.9989,
+      "step": 65
+    },
+    {
+      "epoch": 0.005866666666666667,
+      "grad_norm": 172.0,
+      "learning_rate": 3.9111111111111115e-07,
+      "loss": 2.6968,
+      "step": 66
+    },
+    {
+      "epoch": 0.005955555555555556,
+      "grad_norm": 194.0,
+      "learning_rate": 3.970370370370371e-07,
+      "loss": 2.7227,
+      "step": 67
+    },
+    {
+      "epoch": 0.006044444444444444,
+      "grad_norm": 284.0,
+      "learning_rate": 4.02962962962963e-07,
+      "loss": 2.7829,
+      "step": 68
+    },
+    {
+      "epoch": 0.0061333333333333335,
+      "grad_norm": 107.5,
+      "learning_rate": 4.0888888888888897e-07,
+      "loss": 3.139,
+      "step": 69
+    },
+    {
+      "epoch": 0.006222222222222222,
+      "grad_norm": 344.0,
+      "learning_rate": 4.148148148148148e-07,
+      "loss": 2.8163,
+      "step": 70
+    },
+    {
+      "epoch": 0.006311111111111111,
+      "grad_norm": 154.0,
+      "learning_rate": 4.2074074074074077e-07,
+      "loss": 2.9013,
+      "step": 71
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 306.0,
+      "learning_rate": 4.266666666666667e-07,
+      "loss": 2.6297,
+      "step": 72
+    },
+    {
+      "epoch": 0.006488888888888889,
+      "grad_norm": 241.0,
+      "learning_rate": 4.3259259259259263e-07,
+      "loss": 3.0675,
+      "step": 73
+    },
+    {
+      "epoch": 0.006577777777777778,
+      "grad_norm": 352.0,
+      "learning_rate": 4.3851851851851853e-07,
+      "loss": 3.0723,
+      "step": 74
+    },
+    {
+      "epoch": 0.006666666666666667,
+      "grad_norm": 418.0,
+      "learning_rate": 4.444444444444445e-07,
+      "loss": 2.6478,
+      "step": 75
+    },
+    {
+      "epoch": 0.0067555555555555554,
+      "grad_norm": 125.0,
+      "learning_rate": 4.503703703703704e-07,
+      "loss": 3.5321,
+      "step": 76
+    },
+    {
+      "epoch": 0.006844444444444445,
+      "grad_norm": 520.0,
+      "learning_rate": 4.5629629629629635e-07,
+      "loss": 3.2329,
+      "step": 77
+    },
+    {
+      "epoch": 0.006933333333333333,
+      "grad_norm": 288.0,
+      "learning_rate": 4.6222222222222225e-07,
+      "loss": 2.8868,
+      "step": 78
+    },
+    {
+      "epoch": 0.007022222222222222,
+      "grad_norm": 235.0,
+      "learning_rate": 4.681481481481482e-07,
+      "loss": 2.8696,
+      "step": 79
+    },
+    {
+      "epoch": 0.0071111111111111115,
+      "grad_norm": 520.0,
+      "learning_rate": 4.740740740740741e-07,
+      "loss": 2.7416,
+      "step": 80
+    },
+    {
+      "epoch": 0.0072,
+      "grad_norm": 576.0,
+      "learning_rate": 4.800000000000001e-07,
+      "loss": 2.6907,
+      "step": 81
+    },
+    {
+      "epoch": 0.007288888888888889,
+      "grad_norm": 380.0,
+      "learning_rate": 4.859259259259259e-07,
+      "loss": 2.6807,
+      "step": 82
+    },
+    {
+      "epoch": 0.007377777777777777,
+      "grad_norm": 296.0,
+      "learning_rate": 4.918518518518519e-07,
+      "loss": 2.5397,
+      "step": 83
+    },
+    {
+      "epoch": 0.007466666666666667,
+      "grad_norm": 262.0,
+      "learning_rate": 4.977777777777778e-07,
+      "loss": 2.9621,
+      "step": 84
+    },
+    {
+      "epoch": 0.007555555555555556,
+      "grad_norm": 384.0,
+      "learning_rate": 5.037037037037038e-07,
+      "loss": 3.0418,
+      "step": 85
+    },
+    {
+      "epoch": 0.007644444444444444,
+      "grad_norm": 398.0,
+      "learning_rate": 5.096296296296296e-07,
+      "loss": 2.5786,
+      "step": 86
+    },
+    {
+      "epoch": 0.007733333333333333,
+      "grad_norm": 482.0,
+      "learning_rate": 5.155555555555556e-07,
+      "loss": 2.5717,
+      "step": 87
+    },
+    {
+      "epoch": 0.007822222222222222,
+      "grad_norm": 476.0,
+      "learning_rate": 5.214814814814815e-07,
+      "loss": 2.9471,
+      "step": 88
+    },
+    {
+      "epoch": 0.007911111111111112,
+      "grad_norm": 198.0,
+      "learning_rate": 5.274074074074075e-07,
+      "loss": 2.7742,
+      "step": 89
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 237.0,
+      "learning_rate": 5.333333333333335e-07,
+      "loss": 2.6615,
+      "step": 90
+    },
+    {
+      "epoch": 0.008088888888888889,
+      "grad_norm": 199.0,
+      "learning_rate": 5.392592592592593e-07,
+      "loss": 2.9306,
+      "step": 91
+    },
+    {
+      "epoch": 0.008177777777777779,
+      "grad_norm": 396.0,
+      "learning_rate": 5.451851851851853e-07,
+      "loss": 2.9596,
+      "step": 92
+    },
+    {
+      "epoch": 0.008266666666666667,
+      "grad_norm": 202.0,
+      "learning_rate": 5.511111111111111e-07,
+      "loss": 2.9567,
+      "step": 93
+    },
+    {
+      "epoch": 0.008355555555555555,
+      "grad_norm": 195.0,
+      "learning_rate": 5.570370370370371e-07,
+      "loss": 2.9,
+      "step": 94
+    },
+    {
+      "epoch": 0.008444444444444444,
+      "grad_norm": 552.0,
+      "learning_rate": 5.62962962962963e-07,
+      "loss": 3.0706,
+      "step": 95
+    },
+    {
+      "epoch": 0.008533333333333334,
+      "grad_norm": 508.0,
+      "learning_rate": 5.68888888888889e-07,
+      "loss": 2.6964,
+      "step": 96
+    },
+    {
+      "epoch": 0.008622222222222222,
+      "grad_norm": 157.0,
+      "learning_rate": 5.748148148148148e-07,
+      "loss": 2.6942,
+      "step": 97
+    },
+    {
+      "epoch": 0.00871111111111111,
+      "grad_norm": 416.0,
+      "learning_rate": 5.807407407407408e-07,
+      "loss": 2.8071,
+      "step": 98
+    },
+    {
+      "epoch": 0.0088,
+      "grad_norm": 286.0,
+      "learning_rate": 5.866666666666667e-07,
+      "loss": 3.1441,
+      "step": 99
+    },
+    {
+      "epoch": 0.008888888888888889,
+      "grad_norm": 206.0,
+      "learning_rate": 5.925925925925927e-07,
+      "loss": 2.551,
+      "step": 100
+    },
+    {
+      "epoch": 0.008977777777777777,
+      "grad_norm": 440.0,
+      "learning_rate": 5.985185185185185e-07,
+      "loss": 2.878,
+      "step": 101
+    },
+    {
+      "epoch": 0.009066666666666667,
+      "grad_norm": 135.0,
+      "learning_rate": 6.044444444444445e-07,
+      "loss": 2.9306,
+      "step": 102
+    },
+    {
+      "epoch": 0.009155555555555556,
+      "grad_norm": 568.0,
+      "learning_rate": 6.103703703703704e-07,
+      "loss": 2.6823,
+      "step": 103
+    },
+    {
+      "epoch": 0.009244444444444444,
+      "grad_norm": 510.0,
+      "learning_rate": 6.162962962962964e-07,
+      "loss": 2.9133,
+      "step": 104
+    },
+    {
+      "epoch": 0.009333333333333334,
+      "grad_norm": 178.0,
+      "learning_rate": 6.222222222222223e-07,
+      "loss": 2.8718,
+      "step": 105
+    },
+    {
+      "epoch": 0.009422222222222222,
+      "grad_norm": 368.0,
+      "learning_rate": 6.281481481481481e-07,
+      "loss": 2.5801,
+      "step": 106
+    },
+    {
+      "epoch": 0.00951111111111111,
+      "grad_norm": 146.0,
+      "learning_rate": 6.340740740740742e-07,
+      "loss": 2.6506,
+      "step": 107
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 139.0,
+      "learning_rate": 6.4e-07,
+      "loss": 2.6614,
+      "step": 108
+    },
+    {
+      "epoch": 0.00968888888888889,
+      "grad_norm": 144.0,
+      "learning_rate": 6.45925925925926e-07,
+      "loss": 2.7694,
+      "step": 109
+    },
+    {
+      "epoch": 0.009777777777777778,
+      "grad_norm": 185.0,
+      "learning_rate": 6.518518518518518e-07,
+      "loss": 2.5577,
+      "step": 110
+    },
+    {
+      "epoch": 0.009866666666666666,
+      "grad_norm": 342.0,
+      "learning_rate": 6.577777777777779e-07,
+      "loss": 2.8403,
+      "step": 111
+    },
+    {
+      "epoch": 0.009955555555555556,
+      "grad_norm": 312.0,
+      "learning_rate": 6.637037037037037e-07,
+      "loss": 2.5352,
+      "step": 112
+    },
+    {
+      "epoch": 0.010044444444444444,
+      "grad_norm": 264.0,
+      "learning_rate": 6.696296296296297e-07,
+      "loss": 2.7432,
+      "step": 113
+    },
+    {
+      "epoch": 0.010133333333333333,
+      "grad_norm": 224.0,
+      "learning_rate": 6.755555555555555e-07,
+      "loss": 2.869,
+      "step": 114
+    },
+    {
+      "epoch": 0.010222222222222223,
+      "grad_norm": 510.0,
+      "learning_rate": 6.814814814814816e-07,
+      "loss": 2.6895,
+      "step": 115
+    },
+    {
+      "epoch": 0.010311111111111111,
+      "grad_norm": 592.0,
+      "learning_rate": 6.874074074074074e-07,
+      "loss": 2.8418,
+      "step": 116
+    },
+    {
+      "epoch": 0.0104,
+      "grad_norm": 576.0,
+      "learning_rate": 6.933333333333334e-07,
+      "loss": 2.5461,
+      "step": 117
+    },
+    {
+      "epoch": 0.01048888888888889,
+      "grad_norm": 820.0,
+      "learning_rate": 6.992592592592593e-07,
+      "loss": 2.6697,
+      "step": 118
+    },
+    {
+      "epoch": 0.010577777777777778,
+      "grad_norm": 540.0,
+      "learning_rate": 7.051851851851853e-07,
+      "loss": 2.9918,
+      "step": 119
+    },
+    {
+      "epoch": 0.010666666666666666,
+      "grad_norm": 688.0,
+      "learning_rate": 7.111111111111112e-07,
+      "loss": 2.4519,
+      "step": 120
+    },
+    {
+      "epoch": 0.010755555555555556,
+      "grad_norm": 564.0,
+      "learning_rate": 7.17037037037037e-07,
+      "loss": 2.1969,
+      "step": 121
+    },
+    {
+      "epoch": 0.010844444444444445,
+      "grad_norm": 580.0,
+      "learning_rate": 7.229629629629631e-07,
+      "loss": 2.463,
+      "step": 122
+    },
+    {
+      "epoch": 0.010933333333333333,
+      "grad_norm": 696.0,
+      "learning_rate": 7.28888888888889e-07,
+      "loss": 2.5667,
+      "step": 123
+    },
+    {
+      "epoch": 0.011022222222222221,
+      "grad_norm": 452.0,
+      "learning_rate": 7.348148148148149e-07,
+      "loss": 2.942,
+      "step": 124
+    },
+    {
+      "epoch": 0.011111111111111112,
+      "grad_norm": 1576.0,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 2.1558,
+      "step": 125
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 1464.0,
+      "learning_rate": 7.466666666666668e-07,
+      "loss": 2.563,
+      "step": 126
+    },
+    {
+      "epoch": 0.011288888888888888,
+      "grad_norm": 1496.0,
+      "learning_rate": 7.525925925925926e-07,
+      "loss": 2.6966,
+      "step": 127
+    },
+    {
+      "epoch": 0.011377777777777778,
+      "grad_norm": 1520.0,
+      "learning_rate": 7.585185185185186e-07,
+      "loss": 2.4643,
+      "step": 128
+    },
+    {
+      "epoch": 0.011466666666666667,
+      "grad_norm": 956.0,
+      "learning_rate": 7.644444444444444e-07,
+      "loss": 2.3735,
+      "step": 129
+    },
+    {
+      "epoch": 0.011555555555555555,
+      "grad_norm": 1712.0,
+      "learning_rate": 7.703703703703705e-07,
+      "loss": 2.4782,
+      "step": 130
+    },
+    {
+      "epoch": 0.011644444444444445,
+      "grad_norm": 1728.0,
+      "learning_rate": 7.762962962962964e-07,
+      "loss": 2.8266,
+      "step": 131
+    },
+    {
+      "epoch": 0.011733333333333333,
+      "grad_norm": 1392.0,
+      "learning_rate": 7.822222222222223e-07,
+      "loss": 2.4784,
+      "step": 132
+    },
+    {
+      "epoch": 0.011822222222222222,
+      "grad_norm": 1128.0,
+      "learning_rate": 7.881481481481482e-07,
+      "loss": 2.3841,
+      "step": 133
+    },
+    {
+      "epoch": 0.011911111111111112,
+      "grad_norm": 334.0,
+      "learning_rate": 7.940740740740742e-07,
+      "loss": 2.2875,
+      "step": 134
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 952.0,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.3121,
+      "step": 135
+    },
+    {
+      "epoch": 0.012088888888888889,
+      "grad_norm": 832.0,
+      "learning_rate": 8.05925925925926e-07,
+      "loss": 2.4132,
+      "step": 136
+    },
+    {
+      "epoch": 0.012177777777777777,
+      "grad_norm": 668.0,
+      "learning_rate": 8.118518518518519e-07,
+      "loss": 2.6992,
+      "step": 137
+    },
+    {
+      "epoch": 0.012266666666666667,
+      "grad_norm": 1072.0,
+      "learning_rate": 8.177777777777779e-07,
+      "loss": 2.3798,
+      "step": 138
+    },
+    {
+      "epoch": 0.012355555555555555,
+      "grad_norm": 636.0,
+      "learning_rate": 8.237037037037038e-07,
+      "loss": 2.2285,
+      "step": 139
+    },
+    {
+      "epoch": 0.012444444444444444,
+      "grad_norm": 908.0,
+      "learning_rate": 8.296296296296296e-07,
+      "loss": 2.1809,
+      "step": 140
+    },
+    {
+      "epoch": 0.012533333333333334,
+      "grad_norm": 322.0,
+      "learning_rate": 8.355555555555556e-07,
+      "loss": 2.2422,
+      "step": 141
+    },
+    {
+      "epoch": 0.012622222222222222,
+      "grad_norm": 556.0,
+      "learning_rate": 8.414814814814815e-07,
+      "loss": 2.2545,
+      "step": 142
+    },
+    {
+      "epoch": 0.01271111111111111,
+      "grad_norm": 1808.0,
+      "learning_rate": 8.474074074074075e-07,
+      "loss": 2.2822,
+      "step": 143
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 936.0,
+      "learning_rate": 8.533333333333334e-07,
+      "loss": 2.8311,
+      "step": 144
+    },
+    {
+      "epoch": 0.012888888888888889,
+      "grad_norm": 1400.0,
+      "learning_rate": 8.592592592592593e-07,
+      "loss": 2.2761,
+      "step": 145
+    },
+    {
+      "epoch": 0.012977777777777777,
+      "grad_norm": 1296.0,
+      "learning_rate": 8.651851851851853e-07,
+      "loss": 2.1615,
+      "step": 146
+    },
+    {
+      "epoch": 0.013066666666666667,
+      "grad_norm": 1072.0,
+      "learning_rate": 8.711111111111112e-07,
+      "loss": 1.903,
+      "step": 147
+    },
+    {
+      "epoch": 0.013155555555555556,
+      "grad_norm": 536.0,
+      "learning_rate": 8.770370370370371e-07,
+      "loss": 2.8517,
+      "step": 148
+    },
+    {
+      "epoch": 0.013244444444444444,
+      "grad_norm": 486.0,
+      "learning_rate": 8.829629629629629e-07,
+      "loss": 2.2429,
+      "step": 149
+    },
+    {
+      "epoch": 0.013333333333333334,
+      "grad_norm": 524.0,
+      "learning_rate": 8.88888888888889e-07,
+      "loss": 2.5065,
+      "step": 150
+    },
+    {
+      "epoch": 0.013422222222222223,
+      "grad_norm": 968.0,
+      "learning_rate": 8.948148148148149e-07,
+      "loss": 1.9435,
+      "step": 151
+    },
+    {
+      "epoch": 0.013511111111111111,
+      "grad_norm": 996.0,
+      "learning_rate": 9.007407407407408e-07,
+      "loss": 2.5237,
+      "step": 152
+    },
+    {
+      "epoch": 0.0136,
+      "grad_norm": 796.0,
+      "learning_rate": 9.066666666666668e-07,
+      "loss": 2.0421,
+      "step": 153
+    },
+    {
+      "epoch": 0.01368888888888889,
+      "grad_norm": 1392.0,
+      "learning_rate": 9.125925925925927e-07,
+      "loss": 2.1644,
+      "step": 154
+    },
+    {
+      "epoch": 0.013777777777777778,
+      "grad_norm": 676.0,
+      "learning_rate": 9.185185185185185e-07,
+      "loss": 2.1881,
+      "step": 155
+    },
+    {
+      "epoch": 0.013866666666666666,
+      "grad_norm": 1216.0,
+      "learning_rate": 9.244444444444445e-07,
+      "loss": 2.1719,
+      "step": 156
+    },
+    {
+      "epoch": 0.013955555555555556,
+      "grad_norm": 348.0,
+      "learning_rate": 9.303703703703705e-07,
+      "loss": 2.077,
+      "step": 157
+    },
+    {
+      "epoch": 0.014044444444444444,
+      "grad_norm": 672.0,
+      "learning_rate": 9.362962962962964e-07,
+      "loss": 1.9108,
+      "step": 158
+    },
+    {
+      "epoch": 0.014133333333333333,
+      "grad_norm": 508.0,
+      "learning_rate": 9.422222222222223e-07,
+      "loss": 2.7049,
+      "step": 159
+    },
+    {
+      "epoch": 0.014222222222222223,
+      "grad_norm": 1160.0,
+      "learning_rate": 9.481481481481482e-07,
+      "loss": 2.1949,
+      "step": 160
+    },
+    {
+      "epoch": 0.014311111111111111,
+      "grad_norm": 1008.0,
+      "learning_rate": 9.540740740740742e-07,
+      "loss": 2.0846,
+      "step": 161
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 1688.0,
+      "learning_rate": 9.600000000000001e-07,
+      "loss": 2.4231,
+      "step": 162
+    },
+    {
+      "epoch": 0.01448888888888889,
+      "grad_norm": 1472.0,
+      "learning_rate": 9.65925925925926e-07,
+      "loss": 2.3051,
+      "step": 163
+    },
+    {
+      "epoch": 0.014577777777777778,
+      "grad_norm": 972.0,
+      "learning_rate": 9.718518518518518e-07,
+      "loss": 2.087,
+      "step": 164
+    },
+    {
+      "epoch": 0.014666666666666666,
+      "grad_norm": 528.0,
+      "learning_rate": 9.77777777777778e-07,
+      "loss": 2.4555,
+      "step": 165
+    },
+    {
+      "epoch": 0.014755555555555555,
+      "grad_norm": 1560.0,
+      "learning_rate": 9.837037037037037e-07,
+      "loss": 2.0087,
+      "step": 166
+    },
+    {
+      "epoch": 0.014844444444444445,
+      "grad_norm": 1664.0,
+      "learning_rate": 9.896296296296297e-07,
+      "loss": 2.2001,
+      "step": 167
+    },
+    {
+      "epoch": 0.014933333333333333,
+      "grad_norm": 636.0,
+      "learning_rate": 9.955555555555556e-07,
+      "loss": 2.0916,
+      "step": 168
+    },
+    {
+      "epoch": 0.015022222222222222,
+      "grad_norm": 536.0,
+      "learning_rate": 1.0014814814814816e-06,
+      "loss": 2.4077,
+      "step": 169
+    },
+    {
+      "epoch": 0.015111111111111112,
+      "grad_norm": 1336.0,
+      "learning_rate": 1.0074074074074076e-06,
+      "loss": 1.957,
+      "step": 170
+    },
+    {
+      "epoch": 0.0152,
+      "grad_norm": 430.0,
+      "learning_rate": 1.0133333333333333e-06,
+      "loss": 2.0978,
+      "step": 171
+    },
+    {
+      "epoch": 0.015288888888888888,
+      "grad_norm": 270.0,
+      "learning_rate": 1.0192592592592593e-06,
+      "loss": 2.169,
+      "step": 172
+    },
+    {
+      "epoch": 0.015377777777777778,
+      "grad_norm": 207.0,
+      "learning_rate": 1.0251851851851852e-06,
+      "loss": 2.3478,
+      "step": 173
+    },
+    {
+      "epoch": 0.015466666666666667,
+      "grad_norm": 564.0,
+      "learning_rate": 1.0311111111111112e-06,
+      "loss": 2.3526,
+      "step": 174
+    },
+    {
+      "epoch": 0.015555555555555555,
+      "grad_norm": 916.0,
+      "learning_rate": 1.0370370370370371e-06,
+      "loss": 1.8312,
+      "step": 175
+    },
+    {
+      "epoch": 0.015644444444444443,
+      "grad_norm": 1752.0,
+      "learning_rate": 1.042962962962963e-06,
+      "loss": 2.4186,
+      "step": 176
+    },
+    {
+      "epoch": 0.015733333333333332,
+      "grad_norm": 1232.0,
+      "learning_rate": 1.048888888888889e-06,
+      "loss": 2.3607,
+      "step": 177
+    },
+    {
+      "epoch": 0.015822222222222224,
+      "grad_norm": 1472.0,
+      "learning_rate": 1.054814814814815e-06,
+      "loss": 2.421,
+      "step": 178
+    },
+    {
+      "epoch": 0.015911111111111112,
+      "grad_norm": 1608.0,
+      "learning_rate": 1.0607407407407407e-06,
+      "loss": 2.5158,
+      "step": 179
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 1400.0,
+      "learning_rate": 1.066666666666667e-06,
+      "loss": 1.903,
+      "step": 180
+    },
+    {
+      "epoch": 0.01608888888888889,
+      "grad_norm": 318.0,
+      "learning_rate": 1.0725925925925926e-06,
+      "loss": 2.0623,
+      "step": 181
+    },
+    {
+      "epoch": 0.016177777777777777,
+      "grad_norm": 1224.0,
+      "learning_rate": 1.0785185185185186e-06,
+      "loss": 1.9954,
+      "step": 182
+    },
+    {
+      "epoch": 0.016266666666666665,
+      "grad_norm": 428.0,
+      "learning_rate": 1.0844444444444446e-06,
+      "loss": 2.3117,
+      "step": 183
+    },
+    {
+      "epoch": 0.016355555555555557,
+      "grad_norm": 616.0,
+      "learning_rate": 1.0903703703703705e-06,
+      "loss": 2.4414,
+      "step": 184
+    },
+    {
+      "epoch": 0.016444444444444446,
+      "grad_norm": 696.0,
+      "learning_rate": 1.0962962962962965e-06,
+      "loss": 2.1727,
+      "step": 185
+    },
+    {
+      "epoch": 0.016533333333333334,
+      "grad_norm": 454.0,
+      "learning_rate": 1.1022222222222222e-06,
+      "loss": 2.2616,
+      "step": 186
+    },
+    {
+      "epoch": 0.016622222222222222,
+      "grad_norm": 1560.0,
+      "learning_rate": 1.1081481481481482e-06,
+      "loss": 1.829,
+      "step": 187
+    },
+    {
+      "epoch": 0.01671111111111111,
+      "grad_norm": 1120.0,
+      "learning_rate": 1.1140740740740741e-06,
+      "loss": 2.1582,
+      "step": 188
+    },
+    {
+      "epoch": 0.0168,
+      "grad_norm": 1512.0,
+      "learning_rate": 1.12e-06,
+      "loss": 1.9578,
+      "step": 189
+    },
+    {
+      "epoch": 0.016888888888888887,
+      "grad_norm": 1408.0,
+      "learning_rate": 1.125925925925926e-06,
+      "loss": 2.0031,
+      "step": 190
+    },
+    {
+      "epoch": 0.01697777777777778,
+      "grad_norm": 560.0,
+      "learning_rate": 1.131851851851852e-06,
+      "loss": 1.8243,
+      "step": 191
+    },
+    {
+      "epoch": 0.017066666666666667,
+      "grad_norm": 564.0,
+      "learning_rate": 1.137777777777778e-06,
+      "loss": 2.0516,
+      "step": 192
+    },
+    {
+      "epoch": 0.017155555555555556,
+      "grad_norm": 253.0,
+      "learning_rate": 1.143703703703704e-06,
+      "loss": 1.8532,
+      "step": 193
+    },
+    {
+      "epoch": 0.017244444444444444,
+      "grad_norm": 788.0,
+      "learning_rate": 1.1496296296296296e-06,
+      "loss": 1.8967,
+      "step": 194
+    },
+    {
+      "epoch": 0.017333333333333333,
+      "grad_norm": 306.0,
+      "learning_rate": 1.1555555555555556e-06,
+      "loss": 1.927,
+      "step": 195
+    },
+    {
+      "epoch": 0.01742222222222222,
+      "grad_norm": 219.0,
+      "learning_rate": 1.1614814814814816e-06,
+      "loss": 2.3675,
+      "step": 196
+    },
+    {
+      "epoch": 0.017511111111111113,
+      "grad_norm": 206.0,
+      "learning_rate": 1.1674074074074075e-06,
+      "loss": 2.6033,
+      "step": 197
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 474.0,
+      "learning_rate": 1.1733333333333335e-06,
+      "loss": 1.9506,
+      "step": 198
+    },
+    {
+      "epoch": 0.01768888888888889,
+      "grad_norm": 268.0,
+      "learning_rate": 1.1792592592592592e-06,
+      "loss": 1.7833,
+      "step": 199
+    },
+    {
+      "epoch": 0.017777777777777778,
+      "grad_norm": 182.0,
+      "learning_rate": 1.1851851851851854e-06,
+      "loss": 2.4446,
+      "step": 200
+    },
+    {
+      "epoch": 0.017866666666666666,
+      "grad_norm": 1608.0,
+      "learning_rate": 1.1911111111111111e-06,
+      "loss": 2.0552,
+      "step": 201
+    },
+    {
+      "epoch": 0.017955555555555554,
+      "grad_norm": 1576.0,
+      "learning_rate": 1.197037037037037e-06,
+      "loss": 2.2308,
+      "step": 202
+    },
+    {
+      "epoch": 0.018044444444444443,
+      "grad_norm": 1608.0,
+      "learning_rate": 1.202962962962963e-06,
+      "loss": 2.2934,
+      "step": 203
+    },
+    {
+      "epoch": 0.018133333333333335,
+      "grad_norm": 2040.0,
+      "learning_rate": 1.208888888888889e-06,
+      "loss": 2.1001,
+      "step": 204
+    },
+    {
+      "epoch": 0.018222222222222223,
+      "grad_norm": 392.0,
+      "learning_rate": 1.214814814814815e-06,
+      "loss": 1.9445,
+      "step": 205
+    },
+    {
+      "epoch": 0.01831111111111111,
+      "grad_norm": 264.0,
+      "learning_rate": 1.220740740740741e-06,
+      "loss": 1.8039,
+      "step": 206
+    },
+    {
+      "epoch": 0.0184,
+      "grad_norm": 2024.0,
+      "learning_rate": 1.2266666666666666e-06,
+      "loss": 2.1559,
+      "step": 207
+    },
+    {
+      "epoch": 0.018488888888888888,
+      "grad_norm": 812.0,
+      "learning_rate": 1.2325925925925928e-06,
+      "loss": 1.9572,
+      "step": 208
+    },
+    {
+      "epoch": 0.018577777777777776,
+      "grad_norm": 332.0,
+      "learning_rate": 1.2385185185185186e-06,
+      "loss": 2.346,
+      "step": 209
+    },
+    {
+      "epoch": 0.018666666666666668,
+      "grad_norm": 608.0,
+      "learning_rate": 1.2444444444444445e-06,
+      "loss": 1.9277,
+      "step": 210
+    },
+    {
+      "epoch": 0.018755555555555557,
+      "grad_norm": 836.0,
+      "learning_rate": 1.2503703703703705e-06,
+      "loss": 2.2991,
+      "step": 211
+    },
+    {
+      "epoch": 0.018844444444444445,
+      "grad_norm": 143.0,
+      "learning_rate": 1.2562962962962962e-06,
+      "loss": 1.9679,
+      "step": 212
+    },
+    {
+      "epoch": 0.018933333333333333,
+      "grad_norm": 204.0,
+      "learning_rate": 1.2622222222222224e-06,
+      "loss": 2.2486,
+      "step": 213
+    },
+    {
+      "epoch": 0.01902222222222222,
+      "grad_norm": 460.0,
+      "learning_rate": 1.2681481481481483e-06,
+      "loss": 1.6202,
+      "step": 214
+    },
+    {
+      "epoch": 0.01911111111111111,
+      "grad_norm": 620.0,
+      "learning_rate": 1.2740740740740743e-06,
+      "loss": 1.9862,
+      "step": 215
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 456.0,
+      "learning_rate": 1.28e-06,
+      "loss": 2.6504,
+      "step": 216
+    },
+    {
+      "epoch": 0.01928888888888889,
+      "grad_norm": 524.0,
+      "learning_rate": 1.285925925925926e-06,
+      "loss": 1.8197,
+      "step": 217
+    },
+    {
+      "epoch": 0.01937777777777778,
+      "grad_norm": 157.0,
+      "learning_rate": 1.291851851851852e-06,
+      "loss": 1.6566,
+      "step": 218
+    },
+    {
+      "epoch": 0.019466666666666667,
+      "grad_norm": 450.0,
+      "learning_rate": 1.2977777777777779e-06,
+      "loss": 1.805,
+      "step": 219
+    },
+    {
+      "epoch": 0.019555555555555555,
+      "grad_norm": 520.0,
+      "learning_rate": 1.3037037037037036e-06,
+      "loss": 1.6847,
+      "step": 220
+    },
+    {
+      "epoch": 0.019644444444444444,
+      "grad_norm": 374.0,
+      "learning_rate": 1.3096296296296298e-06,
+      "loss": 1.7384,
+      "step": 221
+    },
+    {
+      "epoch": 0.019733333333333332,
+      "grad_norm": 187.0,
+      "learning_rate": 1.3155555555555558e-06,
+      "loss": 2.1447,
+      "step": 222
+    },
+    {
+      "epoch": 0.019822222222222224,
+      "grad_norm": 892.0,
+      "learning_rate": 1.3214814814814817e-06,
+      "loss": 2.1242,
+      "step": 223
+    },
+    {
+      "epoch": 0.019911111111111112,
+      "grad_norm": 161.0,
+      "learning_rate": 1.3274074074074075e-06,
+      "loss": 1.9482,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 338.0,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 2.2184,
+      "step": 225
+    },
+    {
+      "epoch": 0.02008888888888889,
+      "grad_norm": 92.5,
+      "learning_rate": 1.3392592592592594e-06,
+      "loss": 1.9917,
+      "step": 226
+    },
+    {
+      "epoch": 0.020177777777777777,
+      "grad_norm": 464.0,
+      "learning_rate": 1.3451851851851851e-06,
+      "loss": 2.0897,
+      "step": 227
+    },
+    {
+      "epoch": 0.020266666666666665,
+      "grad_norm": 376.0,
+      "learning_rate": 1.351111111111111e-06,
+      "loss": 1.5719,
+      "step": 228
+    },
+    {
+      "epoch": 0.020355555555555557,
+      "grad_norm": 624.0,
+      "learning_rate": 1.3570370370370372e-06,
+      "loss": 2.3121,
+      "step": 229
+    },
+    {
+      "epoch": 0.020444444444444446,
+      "grad_norm": 560.0,
+      "learning_rate": 1.3629629629629632e-06,
+      "loss": 1.8598,
+      "step": 230
+    },
+    {
+      "epoch": 0.020533333333333334,
+      "grad_norm": 428.0,
+      "learning_rate": 1.3688888888888891e-06,
+      "loss": 1.7444,
+      "step": 231
+    },
+    {
+      "epoch": 0.020622222222222222,
+      "grad_norm": 398.0,
+      "learning_rate": 1.3748148148148149e-06,
+      "loss": 1.7866,
+      "step": 232
+    },
+    {
+      "epoch": 0.02071111111111111,
+      "grad_norm": 414.0,
+      "learning_rate": 1.3807407407407408e-06,
+      "loss": 1.8359,
+      "step": 233
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 366.0,
+      "learning_rate": 1.3866666666666668e-06,
+      "loss": 2.1752,
+      "step": 234
+    },
+    {
+      "epoch": 0.020888888888888887,
+      "grad_norm": 200.0,
+      "learning_rate": 1.3925925925925925e-06,
+      "loss": 1.9274,
+      "step": 235
+    },
+    {
+      "epoch": 0.02097777777777778,
+      "grad_norm": 776.0,
+      "learning_rate": 1.3985185185185185e-06,
+      "loss": 1.6493,
+      "step": 236
+    },
+    {
+      "epoch": 0.021066666666666668,
+      "grad_norm": 520.0,
+      "learning_rate": 1.4044444444444447e-06,
+      "loss": 2.127,
+      "step": 237
+    },
+    {
+      "epoch": 0.021155555555555556,
+      "grad_norm": 131.0,
+      "learning_rate": 1.4103703703703706e-06,
+      "loss": 1.9301,
+      "step": 238
+    },
+    {
+      "epoch": 0.021244444444444444,
+      "grad_norm": 127.5,
+      "learning_rate": 1.4162962962962964e-06,
+      "loss": 1.9289,
+      "step": 239
+    },
+    {
+      "epoch": 0.021333333333333333,
+      "grad_norm": 180.0,
+      "learning_rate": 1.4222222222222223e-06,
+      "loss": 2.1142,
+      "step": 240
+    },
+    {
+      "epoch": 0.02142222222222222,
+      "grad_norm": 206.0,
+      "learning_rate": 1.4281481481481483e-06,
+      "loss": 1.8084,
+      "step": 241
+    },
+    {
+      "epoch": 0.021511111111111113,
+      "grad_norm": 124.5,
+      "learning_rate": 1.434074074074074e-06,
+      "loss": 2.0736,
+      "step": 242
+    },
+    {
+      "epoch": 0.0216,
+      "grad_norm": 199.0,
+      "learning_rate": 1.44e-06,
+      "loss": 2.1418,
+      "step": 243
+    },
+    {
+      "epoch": 0.02168888888888889,
+      "grad_norm": 127.0,
+      "learning_rate": 1.4459259259259261e-06,
+      "loss": 1.8402,
+      "step": 244
+    },
+    {
+      "epoch": 0.021777777777777778,
+      "grad_norm": 544.0,
+      "learning_rate": 1.451851851851852e-06,
+      "loss": 2.0848,
+      "step": 245
+    },
+    {
+      "epoch": 0.021866666666666666,
+      "grad_norm": 68.0,
+      "learning_rate": 1.457777777777778e-06,
+      "loss": 1.7449,
+      "step": 246
+    },
+    {
+      "epoch": 0.021955555555555555,
+      "grad_norm": 262.0,
+      "learning_rate": 1.4637037037037038e-06,
+      "loss": 1.7924,
+      "step": 247
+    },
+    {
+      "epoch": 0.022044444444444443,
+      "grad_norm": 408.0,
+      "learning_rate": 1.4696296296296298e-06,
+      "loss": 1.8314,
+      "step": 248
+    },
+    {
+      "epoch": 0.022133333333333335,
+      "grad_norm": 188.0,
+      "learning_rate": 1.4755555555555557e-06,
+      "loss": 2.1662,
+      "step": 249
+    },
+    {
+      "epoch": 0.022222222222222223,
+      "grad_norm": 73.0,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 2.4121,
+      "step": 250
+    },
+    {
+      "epoch": 0.02231111111111111,
+      "grad_norm": 1020.0,
+      "learning_rate": 1.4874074074074074e-06,
+      "loss": 1.8504,
+      "step": 251
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 596.0,
+      "learning_rate": 1.4933333333333336e-06,
+      "loss": 2.2239,
+      "step": 252
+    },
+    {
+      "epoch": 0.022488888888888888,
+      "grad_norm": 346.0,
+      "learning_rate": 1.4992592592592595e-06,
+      "loss": 1.935,
+      "step": 253
+    },
+    {
+      "epoch": 0.022577777777777776,
+      "grad_norm": 211.0,
+      "learning_rate": 1.5051851851851853e-06,
+      "loss": 1.5215,
+      "step": 254
+    },
+    {
+      "epoch": 0.02266666666666667,
+      "grad_norm": 264.0,
+      "learning_rate": 1.5111111111111112e-06,
+      "loss": 1.6145,
+      "step": 255
+    },
+    {
+      "epoch": 0.022755555555555557,
+      "grad_norm": 832.0,
+      "learning_rate": 1.5170370370370372e-06,
+      "loss": 1.7378,
+      "step": 256
+    },
+    {
+      "epoch": 0.022844444444444445,
+      "grad_norm": 788.0,
+      "learning_rate": 1.522962962962963e-06,
+      "loss": 1.6191,
+      "step": 257
+    },
+    {
+      "epoch": 0.022933333333333333,
+      "grad_norm": 136.0,
+      "learning_rate": 1.5288888888888889e-06,
+      "loss": 1.9586,
+      "step": 258
+    },
+    {
+      "epoch": 0.02302222222222222,
+      "grad_norm": 190.0,
+      "learning_rate": 1.5348148148148148e-06,
+      "loss": 1.5423,
+      "step": 259
+    },
+    {
+      "epoch": 0.02311111111111111,
+      "grad_norm": 596.0,
+      "learning_rate": 1.540740740740741e-06,
+      "loss": 1.9816,
+      "step": 260
+    },
+    {
+      "epoch": 0.0232,
+      "grad_norm": 672.0,
+      "learning_rate": 1.546666666666667e-06,
+      "loss": 2.5644,
+      "step": 261
+    },
+    {
+      "epoch": 0.02328888888888889,
+      "grad_norm": 472.0,
+      "learning_rate": 1.5525925925925927e-06,
+      "loss": 2.1356,
+      "step": 262
+    },
+    {
+      "epoch": 0.02337777777777778,
+      "grad_norm": 272.0,
+      "learning_rate": 1.5585185185185187e-06,
+      "loss": 2.1309,
+      "step": 263
+    },
+    {
+      "epoch": 0.023466666666666667,
+      "grad_norm": 808.0,
+      "learning_rate": 1.5644444444444446e-06,
+      "loss": 2.06,
+      "step": 264
+    },
+    {
+      "epoch": 0.023555555555555555,
+      "grad_norm": 418.0,
+      "learning_rate": 1.5703703703703704e-06,
+      "loss": 1.5306,
+      "step": 265
+    },
+    {
+      "epoch": 0.023644444444444444,
+      "grad_norm": 266.0,
+      "learning_rate": 1.5762962962962963e-06,
+      "loss": 1.5774,
+      "step": 266
+    },
+    {
+      "epoch": 0.023733333333333332,
+      "grad_norm": 784.0,
+      "learning_rate": 1.5822222222222223e-06,
+      "loss": 1.687,
+      "step": 267
+    },
+    {
+      "epoch": 0.023822222222222224,
+      "grad_norm": 115.5,
+      "learning_rate": 1.5881481481481484e-06,
+      "loss": 1.4123,
+      "step": 268
+    },
+    {
+      "epoch": 0.023911111111111112,
+      "grad_norm": 183.0,
+      "learning_rate": 1.5940740740740742e-06,
+      "loss": 1.5784,
+      "step": 269
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 140.0,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.6272,
+      "step": 270
+    },
+    {
+      "epoch": 0.02408888888888889,
+      "grad_norm": 241.0,
+      "learning_rate": 1.605925925925926e-06,
+      "loss": 1.8811,
+      "step": 271
+    },
+    {
+      "epoch": 0.024177777777777777,
+      "grad_norm": 87.0,
+      "learning_rate": 1.611851851851852e-06,
+      "loss": 1.839,
+      "step": 272
+    },
+    {
+      "epoch": 0.024266666666666666,
+      "grad_norm": 256.0,
+      "learning_rate": 1.6177777777777778e-06,
+      "loss": 1.8364,
+      "step": 273
+    },
+    {
+      "epoch": 0.024355555555555554,
+      "grad_norm": 314.0,
+      "learning_rate": 1.6237037037037037e-06,
+      "loss": 1.6312,
+      "step": 274
+    },
+    {
+      "epoch": 0.024444444444444446,
+      "grad_norm": 524.0,
+      "learning_rate": 1.62962962962963e-06,
+      "loss": 1.8397,
+      "step": 275
+    },
+    {
+      "epoch": 0.024533333333333334,
+      "grad_norm": 418.0,
+      "learning_rate": 1.6355555555555559e-06,
+      "loss": 1.8445,
+      "step": 276
+    },
+    {
+      "epoch": 0.024622222222222222,
+      "grad_norm": 278.0,
+      "learning_rate": 1.6414814814814816e-06,
+      "loss": 2.0695,
+      "step": 277
+    },
+    {
+      "epoch": 0.02471111111111111,
+      "grad_norm": 848.0,
+      "learning_rate": 1.6474074074074076e-06,
+      "loss": 1.7845,
+      "step": 278
+    },
+    {
+      "epoch": 0.0248,
+      "grad_norm": 290.0,
+      "learning_rate": 1.6533333333333335e-06,
+      "loss": 2.1102,
+      "step": 279
+    },
+    {
+      "epoch": 0.024888888888888887,
+      "grad_norm": 756.0,
+      "learning_rate": 1.6592592592592593e-06,
+      "loss": 1.7362,
+      "step": 280
+    },
+    {
+      "epoch": 0.02497777777777778,
+      "grad_norm": 348.0,
+      "learning_rate": 1.6651851851851852e-06,
+      "loss": 1.6714,
+      "step": 281
+    },
+    {
+      "epoch": 0.025066666666666668,
+      "grad_norm": 368.0,
+      "learning_rate": 1.6711111111111112e-06,
+      "loss": 2.1332,
+      "step": 282
+    },
+    {
+      "epoch": 0.025155555555555556,
+      "grad_norm": 139.0,
+      "learning_rate": 1.6770370370370373e-06,
+      "loss": 1.8834,
+      "step": 283
+    },
+    {
+      "epoch": 0.025244444444444444,
+      "grad_norm": 78.0,
+      "learning_rate": 1.682962962962963e-06,
+      "loss": 1.7177,
+      "step": 284
+    },
+    {
+      "epoch": 0.025333333333333333,
+      "grad_norm": 70.5,
+      "learning_rate": 1.688888888888889e-06,
+      "loss": 1.5344,
+      "step": 285
+    },
+    {
+      "epoch": 0.02542222222222222,
+      "grad_norm": 226.0,
+      "learning_rate": 1.694814814814815e-06,
+      "loss": 1.7134,
+      "step": 286
+    },
+    {
+      "epoch": 0.02551111111111111,
+      "grad_norm": 360.0,
+      "learning_rate": 1.700740740740741e-06,
+      "loss": 1.7622,
+      "step": 287
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 246.0,
+      "learning_rate": 1.7066666666666667e-06,
+      "loss": 1.9511,
+      "step": 288
+    },
+    {
+      "epoch": 0.02568888888888889,
+      "grad_norm": 388.0,
+      "learning_rate": 1.7125925925925927e-06,
+      "loss": 1.8314,
+      "step": 289
+    },
+    {
+      "epoch": 0.025777777777777778,
+      "grad_norm": 352.0,
+      "learning_rate": 1.7185185185185186e-06,
+      "loss": 1.6169,
+      "step": 290
+    },
+    {
+      "epoch": 0.025866666666666666,
+      "grad_norm": 209.0,
+      "learning_rate": 1.7244444444444448e-06,
+      "loss": 1.8993,
+      "step": 291
+    },
+    {
+      "epoch": 0.025955555555555555,
+      "grad_norm": 59.75,
+      "learning_rate": 1.7303703703703705e-06,
+      "loss": 1.6836,
+      "step": 292
+    },
+    {
+      "epoch": 0.026044444444444443,
+      "grad_norm": 344.0,
+      "learning_rate": 1.7362962962962965e-06,
+      "loss": 1.8091,
+      "step": 293
+    },
+    {
+      "epoch": 0.026133333333333335,
+      "grad_norm": 296.0,
+      "learning_rate": 1.7422222222222224e-06,
+      "loss": 1.8711,
+      "step": 294
+    },
+    {
+      "epoch": 0.026222222222222223,
+      "grad_norm": 207.0,
+      "learning_rate": 1.7481481481481482e-06,
+      "loss": 2.1451,
+      "step": 295
+    },
+    {
+      "epoch": 0.02631111111111111,
+      "grad_norm": 310.0,
+      "learning_rate": 1.7540740740740741e-06,
+      "loss": 1.7102,
+      "step": 296
+    },
+    {
+      "epoch": 0.0264,
+      "grad_norm": 86.0,
+      "learning_rate": 1.76e-06,
+      "loss": 1.6804,
+      "step": 297
+    },
+    {
+      "epoch": 0.026488888888888888,
+      "grad_norm": 98.0,
+      "learning_rate": 1.7659259259259258e-06,
+      "loss": 1.5353,
+      "step": 298
+    },
+    {
+      "epoch": 0.026577777777777777,
+      "grad_norm": 410.0,
+      "learning_rate": 1.771851851851852e-06,
+      "loss": 1.331,
+      "step": 299
+    },
+    {
+      "epoch": 0.02666666666666667,
+      "grad_norm": 358.0,
+      "learning_rate": 1.777777777777778e-06,
+      "loss": 1.9572,
+      "step": 300
+    },
+    {
+      "epoch": 0.026755555555555557,
+      "grad_norm": 568.0,
+      "learning_rate": 1.783703703703704e-06,
+      "loss": 1.4911,
+      "step": 301
+    },
+    {
+      "epoch": 0.026844444444444445,
+      "grad_norm": 45.75,
+      "learning_rate": 1.7896296296296299e-06,
+      "loss": 1.7458,
+      "step": 302
+    },
+    {
+      "epoch": 0.026933333333333333,
+      "grad_norm": 270.0,
+      "learning_rate": 1.7955555555555556e-06,
+      "loss": 1.6834,
+      "step": 303
+    },
+    {
+      "epoch": 0.027022222222222222,
+      "grad_norm": 206.0,
+      "learning_rate": 1.8014814814814816e-06,
+      "loss": 1.9324,
+      "step": 304
+    },
+    {
+      "epoch": 0.02711111111111111,
+      "grad_norm": 96.0,
+      "learning_rate": 1.8074074074074075e-06,
+      "loss": 1.8226,
+      "step": 305
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 84.5,
+      "learning_rate": 1.8133333333333337e-06,
+      "loss": 1.8229,
+      "step": 306
+    },
+    {
+      "epoch": 0.02728888888888889,
+      "grad_norm": 243.0,
+      "learning_rate": 1.8192592592592594e-06,
+      "loss": 1.947,
+      "step": 307
+    },
+    {
+      "epoch": 0.02737777777777778,
+      "grad_norm": 272.0,
+      "learning_rate": 1.8251851851851854e-06,
+      "loss": 1.7109,
+      "step": 308
+    },
+    {
+      "epoch": 0.027466666666666667,
+      "grad_norm": 242.0,
+      "learning_rate": 1.8311111111111113e-06,
+      "loss": 1.4618,
+      "step": 309
+    },
+    {
+      "epoch": 0.027555555555555555,
+      "grad_norm": 186.0,
+      "learning_rate": 1.837037037037037e-06,
+      "loss": 1.662,
+      "step": 310
+    },
+    {
+      "epoch": 0.027644444444444444,
+      "grad_norm": 253.0,
+      "learning_rate": 1.842962962962963e-06,
+      "loss": 1.5705,
+      "step": 311
+    },
+    {
+      "epoch": 0.027733333333333332,
+      "grad_norm": 146.0,
+      "learning_rate": 1.848888888888889e-06,
+      "loss": 1.4771,
+      "step": 312
+    },
+    {
+      "epoch": 0.027822222222222224,
+      "grad_norm": 548.0,
+      "learning_rate": 1.8548148148148147e-06,
+      "loss": 1.7168,
+      "step": 313
+    },
+    {
+      "epoch": 0.027911111111111112,
+      "grad_norm": 64.0,
+      "learning_rate": 1.860740740740741e-06,
+      "loss": 1.6319,
+      "step": 314
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 460.0,
+      "learning_rate": 1.8666666666666669e-06,
+      "loss": 1.5996,
+      "step": 315
+    },
+    {
+      "epoch": 0.02808888888888889,
+      "grad_norm": 102.5,
+      "learning_rate": 1.8725925925925928e-06,
+      "loss": 2.0821,
+      "step": 316
+    },
+    {
+      "epoch": 0.028177777777777777,
+      "grad_norm": 454.0,
+      "learning_rate": 1.8785185185185188e-06,
+      "loss": 1.4013,
+      "step": 317
+    },
+    {
+      "epoch": 0.028266666666666666,
+      "grad_norm": 490.0,
+      "learning_rate": 1.8844444444444445e-06,
+      "loss": 1.4409,
+      "step": 318
+    },
+    {
+      "epoch": 0.028355555555555554,
+      "grad_norm": 90.0,
+      "learning_rate": 1.8903703703703705e-06,
+      "loss": 1.497,
+      "step": 319
+    },
+    {
+      "epoch": 0.028444444444444446,
+      "grad_norm": 548.0,
+      "learning_rate": 1.8962962962962964e-06,
+      "loss": 1.7435,
+      "step": 320
+    },
+    {
+      "epoch": 0.028533333333333334,
+      "grad_norm": 502.0,
+      "learning_rate": 1.9022222222222222e-06,
+      "loss": 1.6075,
+      "step": 321
+    },
+    {
+      "epoch": 0.028622222222222223,
+      "grad_norm": 138.0,
+      "learning_rate": 1.9081481481481483e-06,
+      "loss": 1.3411,
+      "step": 322
+    },
+    {
+      "epoch": 0.02871111111111111,
+      "grad_norm": 233.0,
+      "learning_rate": 1.9140740740740745e-06,
+      "loss": 1.5461,
+      "step": 323
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 438.0,
+      "learning_rate": 1.9200000000000003e-06,
+      "loss": 1.8493,
+      "step": 324
+    },
+    {
+      "epoch": 0.028888888888888888,
+      "grad_norm": 430.0,
+      "learning_rate": 1.925925925925926e-06,
+      "loss": 1.3589,
+      "step": 325
+    },
+    {
+      "epoch": 0.02897777777777778,
+      "grad_norm": 163.0,
+      "learning_rate": 1.931851851851852e-06,
+      "loss": 1.5907,
+      "step": 326
+    },
+    {
+      "epoch": 0.029066666666666668,
+      "grad_norm": 72.5,
+      "learning_rate": 1.937777777777778e-06,
+      "loss": 1.4163,
+      "step": 327
+    },
+    {
+      "epoch": 0.029155555555555556,
+      "grad_norm": 185.0,
+      "learning_rate": 1.9437037037037037e-06,
+      "loss": 1.4259,
+      "step": 328
+    },
+    {
+      "epoch": 0.029244444444444444,
+      "grad_norm": 354.0,
+      "learning_rate": 1.94962962962963e-06,
+      "loss": 1.6694,
+      "step": 329
+    },
+    {
+      "epoch": 0.029333333333333333,
+      "grad_norm": 61.0,
+      "learning_rate": 1.955555555555556e-06,
+      "loss": 1.5262,
+      "step": 330
+    },
+    {
+      "epoch": 0.02942222222222222,
+      "grad_norm": 87.0,
+      "learning_rate": 1.9614814814814817e-06,
+      "loss": 1.5915,
+      "step": 331
+    },
+    {
+      "epoch": 0.02951111111111111,
+      "grad_norm": 304.0,
+      "learning_rate": 1.9674074074074075e-06,
+      "loss": 1.4768,
+      "step": 332
+    },
+    {
+      "epoch": 0.0296,
+      "grad_norm": 342.0,
+      "learning_rate": 1.9733333333333336e-06,
+      "loss": 1.5028,
+      "step": 333
+    },
+    {
+      "epoch": 0.02968888888888889,
+      "grad_norm": 410.0,
+      "learning_rate": 1.9792592592592594e-06,
+      "loss": 1.5221,
+      "step": 334
+    },
+    {
+      "epoch": 0.029777777777777778,
+      "grad_norm": 59.0,
+      "learning_rate": 1.985185185185185e-06,
+      "loss": 1.4783,
+      "step": 335
+    },
+    {
+      "epoch": 0.029866666666666666,
+      "grad_norm": 107.5,
+      "learning_rate": 1.9911111111111113e-06,
+      "loss": 1.4197,
+      "step": 336
+    },
+    {
+      "epoch": 0.029955555555555555,
+      "grad_norm": 600.0,
+      "learning_rate": 1.9970370370370375e-06,
+      "loss": 1.5368,
+      "step": 337
+    },
+    {
+      "epoch": 0.030044444444444443,
+      "grad_norm": 109.5,
+      "learning_rate": 2.002962962962963e-06,
+      "loss": 1.3078,
+      "step": 338
+    },
+    {
+      "epoch": 0.030133333333333335,
+      "grad_norm": 330.0,
+      "learning_rate": 2.008888888888889e-06,
+      "loss": 1.7949,
+      "step": 339
+    },
+    {
+      "epoch": 0.030222222222222223,
+      "grad_norm": 72.5,
+      "learning_rate": 2.014814814814815e-06,
+      "loss": 1.645,
+      "step": 340
+    },
+    {
+      "epoch": 0.03031111111111111,
+      "grad_norm": 130.0,
+      "learning_rate": 2.020740740740741e-06,
+      "loss": 1.7376,
+      "step": 341
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 314.0,
+      "learning_rate": 2.0266666666666666e-06,
+      "loss": 1.4562,
+      "step": 342
+    },
+    {
+      "epoch": 0.03048888888888889,
+      "grad_norm": 326.0,
+      "learning_rate": 2.0325925925925928e-06,
+      "loss": 1.4022,
+      "step": 343
+    },
+    {
+      "epoch": 0.030577777777777777,
+      "grad_norm": 306.0,
+      "learning_rate": 2.0385185185185185e-06,
+      "loss": 1.5097,
+      "step": 344
+    },
+    {
+      "epoch": 0.030666666666666665,
+      "grad_norm": 396.0,
+      "learning_rate": 2.0444444444444447e-06,
+      "loss": 1.494,
+      "step": 345
+    },
+    {
+      "epoch": 0.030755555555555557,
+      "grad_norm": 334.0,
+      "learning_rate": 2.0503703703703704e-06,
+      "loss": 1.5265,
+      "step": 346
+    },
+    {
+      "epoch": 0.030844444444444445,
+      "grad_norm": 528.0,
+      "learning_rate": 2.0562962962962966e-06,
+      "loss": 1.4911,
+      "step": 347
+    },
+    {
+      "epoch": 0.030933333333333334,
+      "grad_norm": 1024.0,
+      "learning_rate": 2.0622222222222223e-06,
+      "loss": 1.6338,
+      "step": 348
+    },
+    {
+      "epoch": 0.031022222222222222,
+      "grad_norm": 320.0,
+      "learning_rate": 2.0681481481481485e-06,
+      "loss": 1.6709,
+      "step": 349
+    },
+    {
+      "epoch": 0.03111111111111111,
+      "grad_norm": 302.0,
+      "learning_rate": 2.0740740740740742e-06,
+      "loss": 1.5537,
+      "step": 350
+    },
+    {
+      "epoch": 0.0312,
+      "grad_norm": 260.0,
+      "learning_rate": 2.08e-06,
+      "loss": 1.4419,
+      "step": 351
+    },
+    {
+      "epoch": 0.03128888888888889,
+      "grad_norm": 222.0,
+      "learning_rate": 2.085925925925926e-06,
+      "loss": 1.4615,
+      "step": 352
+    },
+    {
+      "epoch": 0.031377777777777775,
+      "grad_norm": 354.0,
+      "learning_rate": 2.0918518518518523e-06,
+      "loss": 1.3701,
+      "step": 353
+    },
+    {
+      "epoch": 0.031466666666666664,
+      "grad_norm": 250.0,
+      "learning_rate": 2.097777777777778e-06,
+      "loss": 1.4044,
+      "step": 354
+    },
+    {
+      "epoch": 0.03155555555555556,
+      "grad_norm": 96.5,
+      "learning_rate": 2.103703703703704e-06,
+      "loss": 1.7443,
+      "step": 355
+    },
+    {
+      "epoch": 0.03164444444444445,
+      "grad_norm": 312.0,
+      "learning_rate": 2.10962962962963e-06,
+      "loss": 1.288,
+      "step": 356
+    },
+    {
+      "epoch": 0.031733333333333336,
+      "grad_norm": 198.0,
+      "learning_rate": 2.1155555555555557e-06,
+      "loss": 1.7305,
+      "step": 357
+    },
+    {
+      "epoch": 0.031822222222222224,
+      "grad_norm": 107.0,
+      "learning_rate": 2.1214814814814815e-06,
+      "loss": 1.2298,
+      "step": 358
+    },
+    {
+      "epoch": 0.03191111111111111,
+      "grad_norm": 364.0,
+      "learning_rate": 2.1274074074074076e-06,
+      "loss": 1.5715,
+      "step": 359
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 248.0,
+      "learning_rate": 2.133333333333334e-06,
+      "loss": 1.3665,
+      "step": 360
+    },
+    {
+      "epoch": 0.03208888888888889,
+      "grad_norm": 268.0,
+      "learning_rate": 2.1392592592592595e-06,
+      "loss": 1.6009,
+      "step": 361
+    },
+    {
+      "epoch": 0.03217777777777778,
+      "grad_norm": 290.0,
+      "learning_rate": 2.1451851851851853e-06,
+      "loss": 1.5291,
+      "step": 362
+    },
+    {
+      "epoch": 0.032266666666666666,
+      "grad_norm": 264.0,
+      "learning_rate": 2.1511111111111115e-06,
+      "loss": 1.463,
+      "step": 363
+    },
+    {
+      "epoch": 0.032355555555555554,
+      "grad_norm": 240.0,
+      "learning_rate": 2.157037037037037e-06,
+      "loss": 1.2396,
+      "step": 364
+    },
+    {
+      "epoch": 0.03244444444444444,
+      "grad_norm": 528.0,
+      "learning_rate": 2.162962962962963e-06,
+      "loss": 1.2609,
+      "step": 365
+    },
+    {
+      "epoch": 0.03253333333333333,
+      "grad_norm": 374.0,
+      "learning_rate": 2.168888888888889e-06,
+      "loss": 1.5059,
+      "step": 366
+    },
+    {
+      "epoch": 0.03262222222222222,
+      "grad_norm": 118.0,
+      "learning_rate": 2.174814814814815e-06,
+      "loss": 1.538,
+      "step": 367
+    },
+    {
+      "epoch": 0.032711111111111114,
+      "grad_norm": 138.0,
+      "learning_rate": 2.180740740740741e-06,
+      "loss": 1.361,
+      "step": 368
+    },
+    {
+      "epoch": 0.0328,
+      "grad_norm": 46.0,
+      "learning_rate": 2.1866666666666668e-06,
+      "loss": 1.4983,
+      "step": 369
+    },
+    {
+      "epoch": 0.03288888888888889,
+      "grad_norm": 254.0,
+      "learning_rate": 2.192592592592593e-06,
+      "loss": 1.2636,
+      "step": 370
+    },
+    {
+      "epoch": 0.03297777777777778,
+      "grad_norm": 235.0,
+      "learning_rate": 2.1985185185185187e-06,
+      "loss": 1.7093,
+      "step": 371
+    },
+    {
+      "epoch": 0.03306666666666667,
+      "grad_norm": 272.0,
+      "learning_rate": 2.2044444444444444e-06,
+      "loss": 1.2772,
+      "step": 372
+    },
+    {
+      "epoch": 0.033155555555555556,
+      "grad_norm": 106.0,
+      "learning_rate": 2.2103703703703706e-06,
+      "loss": 1.5283,
+      "step": 373
+    },
+    {
+      "epoch": 0.033244444444444445,
+      "grad_norm": 243.0,
+      "learning_rate": 2.2162962962962963e-06,
+      "loss": 1.4666,
+      "step": 374
+    },
+    {
+      "epoch": 0.03333333333333333,
+      "grad_norm": 278.0,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.464,
+      "step": 375
+    },
+    {
+      "epoch": 0.03342222222222222,
+      "grad_norm": 508.0,
+      "learning_rate": 2.2281481481481482e-06,
+      "loss": 1.5069,
+      "step": 376
+    },
+    {
+      "epoch": 0.03351111111111111,
+      "grad_norm": 230.0,
+      "learning_rate": 2.2340740740740744e-06,
+      "loss": 1.4688,
+      "step": 377
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 316.0,
+      "learning_rate": 2.24e-06,
+      "loss": 1.3593,
+      "step": 378
+    },
+    {
+      "epoch": 0.033688888888888886,
+      "grad_norm": 516.0,
+      "learning_rate": 2.2459259259259263e-06,
+      "loss": 1.5535,
+      "step": 379
+    },
+    {
+      "epoch": 0.033777777777777775,
+      "grad_norm": 260.0,
+      "learning_rate": 2.251851851851852e-06,
+      "loss": 1.5377,
+      "step": 380
+    },
+    {
+      "epoch": 0.03386666666666667,
+      "grad_norm": 258.0,
+      "learning_rate": 2.257777777777778e-06,
+      "loss": 1.3784,
+      "step": 381
+    },
+    {
+      "epoch": 0.03395555555555556,
+      "grad_norm": 394.0,
+      "learning_rate": 2.263703703703704e-06,
+      "loss": 1.3672,
+      "step": 382
+    },
+    {
+      "epoch": 0.03404444444444445,
+      "grad_norm": 136.0,
+      "learning_rate": 2.2696296296296297e-06,
+      "loss": 1.1547,
+      "step": 383
+    },
+    {
+      "epoch": 0.034133333333333335,
+      "grad_norm": 120.0,
+      "learning_rate": 2.275555555555556e-06,
+      "loss": 1.5968,
+      "step": 384
+    },
+    {
+      "epoch": 0.03422222222222222,
+      "grad_norm": 218.0,
+      "learning_rate": 2.2814814814814816e-06,
+      "loss": 1.4749,
+      "step": 385
+    },
+    {
+      "epoch": 0.03431111111111111,
+      "grad_norm": 178.0,
+      "learning_rate": 2.287407407407408e-06,
+      "loss": 1.6154,
+      "step": 386
+    },
+    {
+      "epoch": 0.0344,
+      "grad_norm": 117.5,
+      "learning_rate": 2.2933333333333335e-06,
+      "loss": 1.2563,
+      "step": 387
+    },
+    {
+      "epoch": 0.03448888888888889,
+      "grad_norm": 276.0,
+      "learning_rate": 2.2992592592592593e-06,
+      "loss": 1.4849,
+      "step": 388
+    },
+    {
+      "epoch": 0.03457777777777778,
+      "grad_norm": 50.5,
+      "learning_rate": 2.3051851851851855e-06,
+      "loss": 1.5206,
+      "step": 389
+    },
+    {
+      "epoch": 0.034666666666666665,
+      "grad_norm": 90.0,
+      "learning_rate": 2.311111111111111e-06,
+      "loss": 1.4921,
+      "step": 390
+    },
+    {
+      "epoch": 0.03475555555555555,
+      "grad_norm": 133.0,
+      "learning_rate": 2.3170370370370374e-06,
+      "loss": 1.2981,
+      "step": 391
+    },
+    {
+      "epoch": 0.03484444444444444,
+      "grad_norm": 264.0,
+      "learning_rate": 2.322962962962963e-06,
+      "loss": 1.2118,
+      "step": 392
+    },
+    {
+      "epoch": 0.03493333333333333,
+      "grad_norm": 43.75,
+      "learning_rate": 2.3288888888888893e-06,
+      "loss": 1.4324,
+      "step": 393
+    },
+    {
+      "epoch": 0.035022222222222225,
+      "grad_norm": 330.0,
+      "learning_rate": 2.334814814814815e-06,
+      "loss": 1.2464,
+      "step": 394
+    },
+    {
+      "epoch": 0.035111111111111114,
+      "grad_norm": 200.0,
+      "learning_rate": 2.3407407407407408e-06,
+      "loss": 1.4062,
+      "step": 395
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 364.0,
+      "learning_rate": 2.346666666666667e-06,
+      "loss": 1.5116,
+      "step": 396
+    },
+    {
+      "epoch": 0.03528888888888889,
+      "grad_norm": 660.0,
+      "learning_rate": 2.3525925925925927e-06,
+      "loss": 1.3662,
+      "step": 397
+    },
+    {
+      "epoch": 0.03537777777777778,
+      "grad_norm": 53.5,
+      "learning_rate": 2.3585185185185184e-06,
+      "loss": 1.1816,
+      "step": 398
+    },
+    {
+      "epoch": 0.03546666666666667,
+      "grad_norm": 198.0,
+      "learning_rate": 2.3644444444444446e-06,
+      "loss": 1.4046,
+      "step": 399
+    },
+    {
+      "epoch": 0.035555555555555556,
+      "grad_norm": 800.0,
+      "learning_rate": 2.3703703703703707e-06,
+      "loss": 1.1903,
+      "step": 400
+    },
+    {
+      "epoch": 0.035644444444444444,
+      "grad_norm": 183.0,
+      "learning_rate": 2.3762962962962965e-06,
+      "loss": 1.2934,
+      "step": 401
+    },
+    {
+      "epoch": 0.03573333333333333,
+      "grad_norm": 232.0,
+      "learning_rate": 2.3822222222222222e-06,
+      "loss": 1.3737,
+      "step": 402
+    },
+    {
+      "epoch": 0.03582222222222222,
+      "grad_norm": 57.0,
+      "learning_rate": 2.3881481481481484e-06,
+      "loss": 1.157,
+      "step": 403
+    },
+    {
+      "epoch": 0.03591111111111111,
+      "grad_norm": 368.0,
+      "learning_rate": 2.394074074074074e-06,
+      "loss": 1.6452,
+      "step": 404
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 498.0,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.304,
+      "step": 405
+    },
+    {
+      "epoch": 0.036088888888888886,
+      "grad_norm": 888.0,
+      "learning_rate": 2.405925925925926e-06,
+      "loss": 1.1431,
+      "step": 406
+    },
+    {
+      "epoch": 0.03617777777777778,
+      "grad_norm": 219.0,
+      "learning_rate": 2.4118518518518522e-06,
+      "loss": 1.4469,
+      "step": 407
+    },
+    {
+      "epoch": 0.03626666666666667,
+      "grad_norm": 169.0,
+      "learning_rate": 2.417777777777778e-06,
+      "loss": 1.224,
+      "step": 408
+    },
+    {
+      "epoch": 0.03635555555555556,
+      "grad_norm": 113.0,
+      "learning_rate": 2.423703703703704e-06,
+      "loss": 1.2378,
+      "step": 409
+    },
+    {
+      "epoch": 0.036444444444444446,
+      "grad_norm": 168.0,
+      "learning_rate": 2.42962962962963e-06,
+      "loss": 1.2484,
+      "step": 410
+    },
+    {
+      "epoch": 0.036533333333333334,
+      "grad_norm": 103.0,
+      "learning_rate": 2.4355555555555556e-06,
+      "loss": 1.3836,
+      "step": 411
+    },
+    {
+      "epoch": 0.03662222222222222,
+      "grad_norm": 95.5,
+      "learning_rate": 2.441481481481482e-06,
+      "loss": 1.3539,
+      "step": 412
+    },
+    {
+      "epoch": 0.03671111111111111,
+      "grad_norm": 79.5,
+      "learning_rate": 2.4474074074074075e-06,
+      "loss": 1.3549,
+      "step": 413
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 95.5,
+      "learning_rate": 2.4533333333333333e-06,
+      "loss": 1.2751,
+      "step": 414
+    },
+    {
+      "epoch": 0.03688888888888889,
+      "grad_norm": 200.0,
+      "learning_rate": 2.4592592592592594e-06,
+      "loss": 1.5609,
+      "step": 415
+    },
+    {
+      "epoch": 0.036977777777777776,
+      "grad_norm": 876.0,
+      "learning_rate": 2.4651851851851856e-06,
+      "loss": 1.255,
+      "step": 416
+    },
+    {
+      "epoch": 0.037066666666666664,
+      "grad_norm": 454.0,
+      "learning_rate": 2.4711111111111114e-06,
+      "loss": 1.4239,
+      "step": 417
+    },
+    {
+      "epoch": 0.03715555555555555,
+      "grad_norm": 217.0,
+      "learning_rate": 2.477037037037037e-06,
+      "loss": 1.3941,
+      "step": 418
+    },
+    {
+      "epoch": 0.03724444444444444,
+      "grad_norm": 334.0,
+      "learning_rate": 2.4829629629629633e-06,
+      "loss": 1.3322,
+      "step": 419
+    },
+    {
+      "epoch": 0.037333333333333336,
+      "grad_norm": 688.0,
+      "learning_rate": 2.488888888888889e-06,
+      "loss": 1.3129,
+      "step": 420
+    },
+    {
+      "epoch": 0.037422222222222225,
+      "grad_norm": 149.0,
+      "learning_rate": 2.4948148148148148e-06,
+      "loss": 1.5394,
+      "step": 421
+    },
+    {
+      "epoch": 0.03751111111111111,
+      "grad_norm": 442.0,
+      "learning_rate": 2.500740740740741e-06,
+      "loss": 1.3924,
+      "step": 422
+    },
+    {
+      "epoch": 0.0376,
+      "grad_norm": 226.0,
+      "learning_rate": 2.5066666666666667e-06,
+      "loss": 1.2487,
+      "step": 423
+    },
+    {
+      "epoch": 0.03768888888888889,
+      "grad_norm": 120.0,
+      "learning_rate": 2.5125925925925924e-06,
+      "loss": 1.2121,
+      "step": 424
+    },
+    {
+      "epoch": 0.03777777777777778,
+      "grad_norm": 276.0,
+      "learning_rate": 2.5185185185185186e-06,
+      "loss": 1.4465,
+      "step": 425
+    },
+    {
+      "epoch": 0.037866666666666667,
+      "grad_norm": 213.0,
+      "learning_rate": 2.5244444444444447e-06,
+      "loss": 1.4812,
+      "step": 426
+    },
+    {
+      "epoch": 0.037955555555555555,
+      "grad_norm": 90.5,
+      "learning_rate": 2.530370370370371e-06,
+      "loss": 1.2627,
+      "step": 427
+    },
+    {
+      "epoch": 0.03804444444444444,
+      "grad_norm": 120.5,
+      "learning_rate": 2.5362962962962967e-06,
+      "loss": 1.2314,
+      "step": 428
+    },
+    {
+      "epoch": 0.03813333333333333,
+      "grad_norm": 292.0,
+      "learning_rate": 2.5422222222222224e-06,
+      "loss": 1.2771,
+      "step": 429
+    },
+    {
+      "epoch": 0.03822222222222222,
+      "grad_norm": 181.0,
+      "learning_rate": 2.5481481481481486e-06,
+      "loss": 1.1305,
+      "step": 430
+    },
+    {
+      "epoch": 0.03831111111111111,
+      "grad_norm": 119.0,
+      "learning_rate": 2.5540740740740743e-06,
+      "loss": 1.4545,
+      "step": 431
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 61.75,
+      "learning_rate": 2.56e-06,
+      "loss": 1.3775,
+      "step": 432
+    },
+    {
+      "epoch": 0.03848888888888889,
+      "grad_norm": 45.0,
+      "learning_rate": 2.5659259259259262e-06,
+      "loss": 1.4056,
+      "step": 433
+    },
+    {
+      "epoch": 0.03857777777777778,
+      "grad_norm": 85.0,
+      "learning_rate": 2.571851851851852e-06,
+      "loss": 1.2984,
+      "step": 434
+    },
+    {
+      "epoch": 0.03866666666666667,
+      "grad_norm": 28.125,
+      "learning_rate": 2.577777777777778e-06,
+      "loss": 1.4112,
+      "step": 435
+    },
+    {
+      "epoch": 0.03875555555555556,
+      "grad_norm": 47.75,
+      "learning_rate": 2.583703703703704e-06,
+      "loss": 1.474,
+      "step": 436
+    },
+    {
+      "epoch": 0.038844444444444445,
+      "grad_norm": 728.0,
+      "learning_rate": 2.5896296296296296e-06,
+      "loss": 1.3893,
+      "step": 437
+    },
+    {
+      "epoch": 0.038933333333333334,
+      "grad_norm": 300.0,
+      "learning_rate": 2.5955555555555558e-06,
+      "loss": 1.2765,
+      "step": 438
+    },
+    {
+      "epoch": 0.03902222222222222,
+      "grad_norm": 296.0,
+      "learning_rate": 2.6014814814814815e-06,
+      "loss": 1.2866,
+      "step": 439
+    },
+    {
+      "epoch": 0.03911111111111111,
+      "grad_norm": 179.0,
+      "learning_rate": 2.6074074074074073e-06,
+      "loss": 1.3804,
+      "step": 440
+    },
+    {
+      "epoch": 0.0392,
+      "grad_norm": 233.0,
+      "learning_rate": 2.6133333333333334e-06,
+      "loss": 1.2363,
+      "step": 441
+    },
+    {
+      "epoch": 0.03928888888888889,
+      "grad_norm": 198.0,
+      "learning_rate": 2.6192592592592596e-06,
+      "loss": 1.3471,
+      "step": 442
+    },
+    {
+      "epoch": 0.039377777777777775,
+      "grad_norm": 352.0,
+      "learning_rate": 2.6251851851851858e-06,
+      "loss": 1.4551,
+      "step": 443
+    },
+    {
+      "epoch": 0.039466666666666664,
+      "grad_norm": 616.0,
+      "learning_rate": 2.6311111111111115e-06,
+      "loss": 1.1124,
+      "step": 444
+    },
+    {
+      "epoch": 0.03955555555555555,
+      "grad_norm": 248.0,
+      "learning_rate": 2.6370370370370373e-06,
+      "loss": 1.1019,
+      "step": 445
+    },
+    {
+      "epoch": 0.03964444444444445,
+      "grad_norm": 432.0,
+      "learning_rate": 2.6429629629629634e-06,
+      "loss": 1.3506,
+      "step": 446
+    },
+    {
+      "epoch": 0.039733333333333336,
+      "grad_norm": 139.0,
+      "learning_rate": 2.648888888888889e-06,
+      "loss": 1.2422,
+      "step": 447
+    },
+    {
+      "epoch": 0.039822222222222224,
+      "grad_norm": 316.0,
+      "learning_rate": 2.654814814814815e-06,
+      "loss": 1.19,
+      "step": 448
+    },
+    {
+      "epoch": 0.03991111111111111,
+      "grad_norm": 219.0,
+      "learning_rate": 2.660740740740741e-06,
+      "loss": 1.2549,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 157.0,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.144,
+      "step": 450
+    },
+    {
+      "epoch": 0.04008888888888889,
+      "grad_norm": 50.0,
+      "learning_rate": 2.6725925925925926e-06,
+      "loss": 1.3555,
+      "step": 451
+    },
+    {
+      "epoch": 0.04017777777777778,
+      "grad_norm": 234.0,
+      "learning_rate": 2.6785185185185187e-06,
+      "loss": 1.3689,
+      "step": 452
+    },
+    {
+      "epoch": 0.040266666666666666,
+      "grad_norm": 95.0,
+      "learning_rate": 2.6844444444444445e-06,
+      "loss": 1.4077,
+      "step": 453
+    },
+    {
+      "epoch": 0.040355555555555554,
+      "grad_norm": 153.0,
+      "learning_rate": 2.6903703703703702e-06,
+      "loss": 1.4151,
+      "step": 454
+    },
+    {
+      "epoch": 0.04044444444444444,
+      "grad_norm": 388.0,
+      "learning_rate": 2.6962962962962964e-06,
+      "loss": 1.1729,
+      "step": 455
+    },
+    {
+      "epoch": 0.04053333333333333,
+      "grad_norm": 290.0,
+      "learning_rate": 2.702222222222222e-06,
+      "loss": 1.3076,
+      "step": 456
+    },
+    {
+      "epoch": 0.04062222222222222,
+      "grad_norm": 223.0,
+      "learning_rate": 2.7081481481481487e-06,
+      "loss": 1.1007,
+      "step": 457
+    },
+    {
+      "epoch": 0.040711111111111115,
+      "grad_norm": 628.0,
+      "learning_rate": 2.7140740740740745e-06,
+      "loss": 1.1923,
+      "step": 458
+    },
+    {
+      "epoch": 0.0408,
+      "grad_norm": 80.0,
+      "learning_rate": 2.7200000000000002e-06,
+      "loss": 1.2104,
+      "step": 459
+    },
+    {
+      "epoch": 0.04088888888888889,
+      "grad_norm": 318.0,
+      "learning_rate": 2.7259259259259264e-06,
+      "loss": 1.3317,
+      "step": 460
+    },
+    {
+      "epoch": 0.04097777777777778,
+      "grad_norm": 234.0,
+      "learning_rate": 2.731851851851852e-06,
+      "loss": 1.3094,
+      "step": 461
+    },
+    {
+      "epoch": 0.04106666666666667,
+      "grad_norm": 109.5,
+      "learning_rate": 2.7377777777777783e-06,
+      "loss": 1.2681,
+      "step": 462
+    },
+    {
+      "epoch": 0.041155555555555556,
+      "grad_norm": 93.5,
+      "learning_rate": 2.743703703703704e-06,
+      "loss": 1.3734,
+      "step": 463
+    },
+    {
+      "epoch": 0.041244444444444445,
+      "grad_norm": 55.5,
+      "learning_rate": 2.7496296296296298e-06,
+      "loss": 1.42,
+      "step": 464
+    },
+    {
+      "epoch": 0.04133333333333333,
+      "grad_norm": 212.0,
+      "learning_rate": 2.755555555555556e-06,
+      "loss": 1.2326,
+      "step": 465
+    },
+    {
+      "epoch": 0.04142222222222222,
+      "grad_norm": 350.0,
+      "learning_rate": 2.7614814814814817e-06,
+      "loss": 1.1649,
+      "step": 466
+    },
+    {
+      "epoch": 0.04151111111111111,
+      "grad_norm": 153.0,
+      "learning_rate": 2.7674074074074074e-06,
+      "loss": 1.3701,
+      "step": 467
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 208.0,
+      "learning_rate": 2.7733333333333336e-06,
+      "loss": 1.306,
+      "step": 468
+    },
+    {
+      "epoch": 0.041688888888888886,
+      "grad_norm": 220.0,
+      "learning_rate": 2.7792592592592593e-06,
+      "loss": 1.3752,
+      "step": 469
+    },
+    {
+      "epoch": 0.041777777777777775,
+      "grad_norm": 200.0,
+      "learning_rate": 2.785185185185185e-06,
+      "loss": 1.2066,
+      "step": 470
+    },
+    {
+      "epoch": 0.04186666666666667,
+      "grad_norm": 189.0,
+      "learning_rate": 2.7911111111111113e-06,
+      "loss": 1.3922,
+      "step": 471
+    },
+    {
+      "epoch": 0.04195555555555556,
+      "grad_norm": 552.0,
+      "learning_rate": 2.797037037037037e-06,
+      "loss": 1.3839,
+      "step": 472
+    },
+    {
+      "epoch": 0.04204444444444445,
+      "grad_norm": 462.0,
+      "learning_rate": 2.8029629629629636e-06,
+      "loss": 1.3641,
+      "step": 473
+    },
+    {
+      "epoch": 0.042133333333333335,
+      "grad_norm": 85.0,
+      "learning_rate": 2.8088888888888893e-06,
+      "loss": 1.2774,
+      "step": 474
+    },
+    {
+      "epoch": 0.042222222222222223,
+      "grad_norm": 47.75,
+      "learning_rate": 2.814814814814815e-06,
+      "loss": 1.4247,
+      "step": 475
+    },
+    {
+      "epoch": 0.04231111111111111,
+      "grad_norm": 708.0,
+      "learning_rate": 2.8207407407407412e-06,
+      "loss": 1.179,
+      "step": 476
+    },
+    {
+      "epoch": 0.0424,
+      "grad_norm": 192.0,
+      "learning_rate": 2.826666666666667e-06,
+      "loss": 1.4095,
+      "step": 477
+    },
+    {
+      "epoch": 0.04248888888888889,
+      "grad_norm": 84.0,
+      "learning_rate": 2.8325925925925927e-06,
+      "loss": 1.289,
+      "step": 478
+    },
+    {
+      "epoch": 0.04257777777777778,
+      "grad_norm": 112.5,
+      "learning_rate": 2.838518518518519e-06,
+      "loss": 1.4485,
+      "step": 479
+    },
+    {
+      "epoch": 0.042666666666666665,
+      "grad_norm": 253.0,
+      "learning_rate": 2.8444444444444446e-06,
+      "loss": 1.1784,
+      "step": 480
+    },
+    {
+      "epoch": 0.042755555555555554,
+      "grad_norm": 155.0,
+      "learning_rate": 2.8503703703703704e-06,
+      "loss": 1.3439,
+      "step": 481
+    },
+    {
+      "epoch": 0.04284444444444444,
+      "grad_norm": 438.0,
+      "learning_rate": 2.8562962962962966e-06,
+      "loss": 1.3068,
+      "step": 482
+    },
+    {
+      "epoch": 0.04293333333333333,
+      "grad_norm": 54.5,
+      "learning_rate": 2.8622222222222223e-06,
+      "loss": 1.1445,
+      "step": 483
+    },
+    {
+      "epoch": 0.043022222222222226,
+      "grad_norm": 157.0,
+      "learning_rate": 2.868148148148148e-06,
+      "loss": 1.2683,
+      "step": 484
+    },
+    {
+      "epoch": 0.043111111111111114,
+      "grad_norm": 328.0,
+      "learning_rate": 2.874074074074074e-06,
+      "loss": 1.4379,
+      "step": 485
+    },
+    {
+      "epoch": 0.0432,
+      "grad_norm": 178.0,
+      "learning_rate": 2.88e-06,
+      "loss": 1.3707,
+      "step": 486
+    },
+    {
+      "epoch": 0.04328888888888889,
+      "grad_norm": 394.0,
+      "learning_rate": 2.8859259259259257e-06,
+      "loss": 1.151,
+      "step": 487
+    },
+    {
+      "epoch": 0.04337777777777778,
+      "grad_norm": 160.0,
+      "learning_rate": 2.8918518518518523e-06,
+      "loss": 1.3168,
+      "step": 488
+    },
+    {
+      "epoch": 0.04346666666666667,
+      "grad_norm": 113.5,
+      "learning_rate": 2.8977777777777785e-06,
+      "loss": 1.0492,
+      "step": 489
+    },
+    {
+      "epoch": 0.043555555555555556,
+      "grad_norm": 80.5,
+      "learning_rate": 2.903703703703704e-06,
+      "loss": 1.1937,
+      "step": 490
+    },
+    {
+      "epoch": 0.043644444444444444,
+      "grad_norm": 192.0,
+      "learning_rate": 2.90962962962963e-06,
+      "loss": 1.3354,
+      "step": 491
+    },
+    {
+      "epoch": 0.04373333333333333,
+      "grad_norm": 252.0,
+      "learning_rate": 2.915555555555556e-06,
+      "loss": 1.2029,
+      "step": 492
+    },
+    {
+      "epoch": 0.04382222222222222,
+      "grad_norm": 134.0,
+      "learning_rate": 2.921481481481482e-06,
+      "loss": 1.1746,
+      "step": 493
+    },
+    {
+      "epoch": 0.04391111111111111,
+      "grad_norm": 125.0,
+      "learning_rate": 2.9274074074074076e-06,
+      "loss": 1.0901,
+      "step": 494
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 110.5,
+      "learning_rate": 2.9333333333333338e-06,
+      "loss": 1.2751,
+      "step": 495
+    },
+    {
+      "epoch": 0.044088888888888886,
+      "grad_norm": 248.0,
+      "learning_rate": 2.9392592592592595e-06,
+      "loss": 1.3616,
+      "step": 496
+    },
+    {
+      "epoch": 0.04417777777777778,
+      "grad_norm": 520.0,
+      "learning_rate": 2.9451851851851853e-06,
+      "loss": 1.3063,
+      "step": 497
+    },
+    {
+      "epoch": 0.04426666666666667,
+      "grad_norm": 136.0,
+      "learning_rate": 2.9511111111111114e-06,
+      "loss": 1.3194,
+      "step": 498
+    },
+    {
+      "epoch": 0.04435555555555556,
+      "grad_norm": 132.0,
+      "learning_rate": 2.957037037037037e-06,
+      "loss": 1.1492,
+      "step": 499
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 238.0,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 1.3519,
+      "step": 500
+    },
+    {
+      "epoch": 0.044533333333333334,
+      "grad_norm": 520.0,
+      "learning_rate": 2.968888888888889e-06,
+      "loss": 1.1294,
+      "step": 501
+    },
+    {
+      "epoch": 0.04462222222222222,
+      "grad_norm": 568.0,
+      "learning_rate": 2.974814814814815e-06,
+      "loss": 1.1979,
+      "step": 502
+    },
+    {
+      "epoch": 0.04471111111111111,
+      "grad_norm": 310.0,
+      "learning_rate": 2.9807407407407406e-06,
+      "loss": 1.1988,
+      "step": 503
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 344.0,
+      "learning_rate": 2.986666666666667e-06,
+      "loss": 1.0695,
+      "step": 504
+    },
+    {
+      "epoch": 0.04488888888888889,
+      "grad_norm": 326.0,
+      "learning_rate": 2.992592592592593e-06,
+      "loss": 1.1706,
+      "step": 505
+    },
+    {
+      "epoch": 0.044977777777777776,
+      "grad_norm": 242.0,
+      "learning_rate": 2.998518518518519e-06,
+      "loss": 1.2985,
+      "step": 506
+    },
+    {
+      "epoch": 0.045066666666666665,
+      "grad_norm": 170.0,
+      "learning_rate": 3.004444444444445e-06,
+      "loss": 1.1275,
+      "step": 507
+    },
+    {
+      "epoch": 0.04515555555555555,
+      "grad_norm": 106.5,
+      "learning_rate": 3.0103703703703706e-06,
+      "loss": 1.2757,
+      "step": 508
+    },
+    {
+      "epoch": 0.04524444444444444,
+      "grad_norm": 96.0,
+      "learning_rate": 3.0162962962962967e-06,
+      "loss": 1.2064,
+      "step": 509
+    },
+    {
+      "epoch": 0.04533333333333334,
+      "grad_norm": 88.0,
+      "learning_rate": 3.0222222222222225e-06,
+      "loss": 1.3689,
+      "step": 510
+    },
+    {
+      "epoch": 0.045422222222222225,
+      "grad_norm": 25.375,
+      "learning_rate": 3.028148148148148e-06,
+      "loss": 1.2352,
+      "step": 511
+    },
+    {
+      "epoch": 0.04551111111111111,
+      "grad_norm": 47.75,
+      "learning_rate": 3.0340740740740744e-06,
+      "loss": 1.2787,
+      "step": 512
+    },
+    {
+      "epoch": 0.0456,
+      "grad_norm": 432.0,
+      "learning_rate": 3.04e-06,
+      "loss": 1.2989,
+      "step": 513
+    },
+    {
+      "epoch": 0.04568888888888889,
+      "grad_norm": 368.0,
+      "learning_rate": 3.045925925925926e-06,
+      "loss": 1.2123,
+      "step": 514
+    },
+    {
+      "epoch": 0.04577777777777778,
+      "grad_norm": 114.0,
+      "learning_rate": 3.051851851851852e-06,
+      "loss": 1.2673,
+      "step": 515
+    },
+    {
+      "epoch": 0.04586666666666667,
+      "grad_norm": 41.5,
+      "learning_rate": 3.0577777777777778e-06,
+      "loss": 1.4405,
+      "step": 516
+    },
+    {
+      "epoch": 0.045955555555555555,
+      "grad_norm": 89.0,
+      "learning_rate": 3.063703703703704e-06,
+      "loss": 1.1164,
+      "step": 517
+    },
+    {
+      "epoch": 0.04604444444444444,
+      "grad_norm": 168.0,
+      "learning_rate": 3.0696296296296297e-06,
+      "loss": 1.2673,
+      "step": 518
+    },
+    {
+      "epoch": 0.04613333333333333,
+      "grad_norm": 91.0,
+      "learning_rate": 3.0755555555555563e-06,
+      "loss": 1.2423,
+      "step": 519
+    },
+    {
+      "epoch": 0.04622222222222222,
+      "grad_norm": 96.0,
+      "learning_rate": 3.081481481481482e-06,
+      "loss": 1.3676,
+      "step": 520
+    },
+    {
+      "epoch": 0.04631111111111111,
+      "grad_norm": 65.0,
+      "learning_rate": 3.0874074074074078e-06,
+      "loss": 1.012,
+      "step": 521
+    },
+    {
+      "epoch": 0.0464,
+      "grad_norm": 51.25,
+      "learning_rate": 3.093333333333334e-06,
+      "loss": 1.207,
+      "step": 522
+    },
+    {
+      "epoch": 0.04648888888888889,
+      "grad_norm": 55.5,
+      "learning_rate": 3.0992592592592597e-06,
+      "loss": 1.101,
+      "step": 523
+    },
+    {
+      "epoch": 0.04657777777777778,
+      "grad_norm": 61.25,
+      "learning_rate": 3.1051851851851854e-06,
+      "loss": 1.1937,
+      "step": 524
+    },
+    {
+      "epoch": 0.04666666666666667,
+      "grad_norm": 49.0,
+      "learning_rate": 3.1111111111111116e-06,
+      "loss": 1.3712,
+      "step": 525
+    },
+    {
+      "epoch": 0.04675555555555556,
+      "grad_norm": 86.0,
+      "learning_rate": 3.1170370370370373e-06,
+      "loss": 1.0443,
+      "step": 526
+    },
+    {
+      "epoch": 0.046844444444444445,
+      "grad_norm": 55.0,
+      "learning_rate": 3.122962962962963e-06,
+      "loss": 1.2971,
+      "step": 527
+    },
+    {
+      "epoch": 0.046933333333333334,
+      "grad_norm": 68.5,
+      "learning_rate": 3.1288888888888892e-06,
+      "loss": 1.4113,
+      "step": 528
+    },
+    {
+      "epoch": 0.04702222222222222,
+      "grad_norm": 71.5,
+      "learning_rate": 3.134814814814815e-06,
+      "loss": 1.0775,
+      "step": 529
+    },
+    {
+      "epoch": 0.04711111111111111,
+      "grad_norm": 115.5,
+      "learning_rate": 3.1407407407407407e-06,
+      "loss": 1.1993,
+      "step": 530
+    },
+    {
+      "epoch": 0.0472,
+      "grad_norm": 236.0,
+      "learning_rate": 3.146666666666667e-06,
+      "loss": 1.2128,
+      "step": 531
+    },
+    {
+      "epoch": 0.04728888888888889,
+      "grad_norm": 264.0,
+      "learning_rate": 3.1525925925925926e-06,
+      "loss": 1.2806,
+      "step": 532
+    },
+    {
+      "epoch": 0.047377777777777776,
+      "grad_norm": 241.0,
+      "learning_rate": 3.1585185185185184e-06,
+      "loss": 1.168,
+      "step": 533
+    },
+    {
+      "epoch": 0.047466666666666664,
+      "grad_norm": 89.5,
+      "learning_rate": 3.1644444444444445e-06,
+      "loss": 1.2409,
+      "step": 534
+    },
+    {
+      "epoch": 0.04755555555555555,
+      "grad_norm": 126.0,
+      "learning_rate": 3.1703703703703707e-06,
+      "loss": 1.1316,
+      "step": 535
+    },
+    {
+      "epoch": 0.04764444444444445,
+      "grad_norm": 71.0,
+      "learning_rate": 3.176296296296297e-06,
+      "loss": 1.3036,
+      "step": 536
+    },
+    {
+      "epoch": 0.047733333333333336,
+      "grad_norm": 48.25,
+      "learning_rate": 3.1822222222222226e-06,
+      "loss": 1.2673,
+      "step": 537
+    },
+    {
+      "epoch": 0.047822222222222224,
+      "grad_norm": 680.0,
+      "learning_rate": 3.1881481481481484e-06,
+      "loss": 1.1692,
+      "step": 538
+    },
+    {
+      "epoch": 0.04791111111111111,
+      "grad_norm": 524.0,
+      "learning_rate": 3.1940740740740745e-06,
+      "loss": 1.2614,
+      "step": 539
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 600.0,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.2062,
+      "step": 540
+    },
+    {
+      "epoch": 0.04808888888888889,
+      "grad_norm": 624.0,
+      "learning_rate": 3.205925925925926e-06,
+      "loss": 1.1416,
+      "step": 541
+    },
+    {
+      "epoch": 0.04817777777777778,
+      "grad_norm": 492.0,
+      "learning_rate": 3.211851851851852e-06,
+      "loss": 1.3563,
+      "step": 542
+    },
+    {
+      "epoch": 0.048266666666666666,
+      "grad_norm": 298.0,
+      "learning_rate": 3.217777777777778e-06,
+      "loss": 1.3695,
+      "step": 543
+    },
+    {
+      "epoch": 0.048355555555555554,
+      "grad_norm": 61.5,
+      "learning_rate": 3.223703703703704e-06,
+      "loss": 1.0497,
+      "step": 544
+    },
+    {
+      "epoch": 0.04844444444444444,
+      "grad_norm": 79.0,
+      "learning_rate": 3.22962962962963e-06,
+      "loss": 1.1243,
+      "step": 545
+    },
+    {
+      "epoch": 0.04853333333333333,
+      "grad_norm": 26.0,
+      "learning_rate": 3.2355555555555556e-06,
+      "loss": 1.3504,
+      "step": 546
+    },
+    {
+      "epoch": 0.04862222222222222,
+      "grad_norm": 140.0,
+      "learning_rate": 3.2414814814814818e-06,
+      "loss": 1.0935,
+      "step": 547
+    },
+    {
+      "epoch": 0.04871111111111111,
+      "grad_norm": 59.75,
+      "learning_rate": 3.2474074074074075e-06,
+      "loss": 1.2413,
+      "step": 548
+    },
+    {
+      "epoch": 0.0488,
+      "grad_norm": 70.5,
+      "learning_rate": 3.2533333333333332e-06,
+      "loss": 1.2238,
+      "step": 549
+    },
+    {
+      "epoch": 0.04888888888888889,
+      "grad_norm": 93.5,
+      "learning_rate": 3.25925925925926e-06,
+      "loss": 1.1277,
+      "step": 550
+    },
+    {
+      "epoch": 0.04897777777777778,
+      "grad_norm": 44.75,
+      "learning_rate": 3.2651851851851856e-06,
+      "loss": 1.2097,
+      "step": 551
+    },
+    {
+      "epoch": 0.04906666666666667,
+      "grad_norm": 28.75,
+      "learning_rate": 3.2711111111111117e-06,
+      "loss": 1.2791,
+      "step": 552
+    },
+    {
+      "epoch": 0.049155555555555557,
+      "grad_norm": 21.875,
+      "learning_rate": 3.2770370370370375e-06,
+      "loss": 1.3612,
+      "step": 553
+    },
+    {
+      "epoch": 0.049244444444444445,
+      "grad_norm": 27.125,
+      "learning_rate": 3.2829629629629632e-06,
+      "loss": 1.3539,
+      "step": 554
+    },
+    {
+      "epoch": 0.04933333333333333,
+      "grad_norm": 31.125,
+      "learning_rate": 3.2888888888888894e-06,
+      "loss": 1.2273,
+      "step": 555
+    },
+    {
+      "epoch": 0.04942222222222222,
+      "grad_norm": 89.5,
+      "learning_rate": 3.294814814814815e-06,
+      "loss": 1.3107,
+      "step": 556
+    },
+    {
+      "epoch": 0.04951111111111111,
+      "grad_norm": 48.25,
+      "learning_rate": 3.300740740740741e-06,
+      "loss": 1.0585,
+      "step": 557
+    },
+    {
+      "epoch": 0.0496,
+      "grad_norm": 28.375,
+      "learning_rate": 3.306666666666667e-06,
+      "loss": 1.1148,
+      "step": 558
+    },
+    {
+      "epoch": 0.04968888888888889,
+      "grad_norm": 196.0,
+      "learning_rate": 3.312592592592593e-06,
+      "loss": 1.2712,
+      "step": 559
+    },
+    {
+      "epoch": 0.049777777777777775,
+      "grad_norm": 280.0,
+      "learning_rate": 3.3185185185185185e-06,
+      "loss": 1.1633,
+      "step": 560
+    },
+    {
+      "epoch": 0.04986666666666666,
+      "grad_norm": 228.0,
+      "learning_rate": 3.3244444444444447e-06,
+      "loss": 1.2046,
+      "step": 561
+    },
+    {
+      "epoch": 0.04995555555555556,
+      "grad_norm": 266.0,
+      "learning_rate": 3.3303703703703705e-06,
+      "loss": 1.3186,
+      "step": 562
+    },
+    {
+      "epoch": 0.05004444444444445,
+      "grad_norm": 165.0,
+      "learning_rate": 3.336296296296296e-06,
+      "loss": 1.2676,
+      "step": 563
+    },
+    {
+      "epoch": 0.050133333333333335,
+      "grad_norm": 256.0,
+      "learning_rate": 3.3422222222222224e-06,
+      "loss": 1.2269,
+      "step": 564
+    },
+    {
+      "epoch": 0.050222222222222224,
+      "grad_norm": 396.0,
+      "learning_rate": 3.348148148148148e-06,
+      "loss": 1.1724,
+      "step": 565
+    },
+    {
+      "epoch": 0.05031111111111111,
+      "grad_norm": 592.0,
+      "learning_rate": 3.3540740740740747e-06,
+      "loss": 1.2076,
+      "step": 566
+    },
+    {
+      "epoch": 0.0504,
+      "grad_norm": 253.0,
+      "learning_rate": 3.3600000000000004e-06,
+      "loss": 1.2768,
+      "step": 567
+    },
+    {
+      "epoch": 0.05048888888888889,
+      "grad_norm": 596.0,
+      "learning_rate": 3.365925925925926e-06,
+      "loss": 1.3073,
+      "step": 568
+    },
+    {
+      "epoch": 0.05057777777777778,
+      "grad_norm": 87.0,
+      "learning_rate": 3.3718518518518524e-06,
+      "loss": 1.1307,
+      "step": 569
+    },
+    {
+      "epoch": 0.050666666666666665,
+      "grad_norm": 62.0,
+      "learning_rate": 3.377777777777778e-06,
+      "loss": 1.0852,
+      "step": 570
+    },
+    {
+      "epoch": 0.050755555555555554,
+      "grad_norm": 74.0,
+      "learning_rate": 3.383703703703704e-06,
+      "loss": 1.2641,
+      "step": 571
+    },
+    {
+      "epoch": 0.05084444444444444,
+      "grad_norm": 25.5,
+      "learning_rate": 3.38962962962963e-06,
+      "loss": 1.1371,
+      "step": 572
+    },
+    {
+      "epoch": 0.05093333333333333,
+      "grad_norm": 29.25,
+      "learning_rate": 3.3955555555555558e-06,
+      "loss": 0.9764,
+      "step": 573
+    },
+    {
+      "epoch": 0.05102222222222222,
+      "grad_norm": 380.0,
+      "learning_rate": 3.401481481481482e-06,
+      "loss": 1.1951,
+      "step": 574
+    },
+    {
+      "epoch": 0.051111111111111114,
+      "grad_norm": 246.0,
+      "learning_rate": 3.4074074074074077e-06,
+      "loss": 1.0953,
+      "step": 575
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 340.0,
+      "learning_rate": 3.4133333333333334e-06,
+      "loss": 1.1671,
+      "step": 576
+    },
+    {
+      "epoch": 0.05128888888888889,
+      "grad_norm": 536.0,
+      "learning_rate": 3.4192592592592596e-06,
+      "loss": 1.0578,
+      "step": 577
+    },
+    {
+      "epoch": 0.05137777777777778,
+      "grad_norm": 96.5,
+      "learning_rate": 3.4251851851851853e-06,
+      "loss": 1.0278,
+      "step": 578
+    },
+    {
+      "epoch": 0.05146666666666667,
+      "grad_norm": 48.25,
+      "learning_rate": 3.431111111111111e-06,
+      "loss": 1.228,
+      "step": 579
+    },
+    {
+      "epoch": 0.051555555555555556,
+      "grad_norm": 20.5,
+      "learning_rate": 3.4370370370370372e-06,
+      "loss": 1.1282,
+      "step": 580
+    },
+    {
+      "epoch": 0.051644444444444444,
+      "grad_norm": 33.5,
+      "learning_rate": 3.4429629629629634e-06,
+      "loss": 1.2421,
+      "step": 581
+    },
+    {
+      "epoch": 0.05173333333333333,
+      "grad_norm": 51.75,
+      "learning_rate": 3.4488888888888896e-06,
+      "loss": 1.259,
+      "step": 582
+    },
+    {
+      "epoch": 0.05182222222222222,
+      "grad_norm": 95.0,
+      "learning_rate": 3.4548148148148153e-06,
+      "loss": 1.279,
+      "step": 583
+    },
+    {
+      "epoch": 0.05191111111111111,
+      "grad_norm": 89.5,
+      "learning_rate": 3.460740740740741e-06,
+      "loss": 1.1804,
+      "step": 584
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 225.0,
+      "learning_rate": 3.4666666666666672e-06,
+      "loss": 1.0458,
+      "step": 585
+    },
+    {
+      "epoch": 0.052088888888888886,
+      "grad_norm": 276.0,
+      "learning_rate": 3.472592592592593e-06,
+      "loss": 1.1224,
+      "step": 586
+    },
+    {
+      "epoch": 0.052177777777777774,
+      "grad_norm": 520.0,
+      "learning_rate": 3.4785185185185187e-06,
+      "loss": 1.0985,
+      "step": 587
+    },
+    {
+      "epoch": 0.05226666666666667,
+      "grad_norm": 58.5,
+      "learning_rate": 3.484444444444445e-06,
+      "loss": 1.149,
+      "step": 588
+    },
+    {
+      "epoch": 0.05235555555555556,
+      "grad_norm": 73.0,
+      "learning_rate": 3.4903703703703706e-06,
+      "loss": 1.1701,
+      "step": 589
+    },
+    {
+      "epoch": 0.052444444444444446,
+      "grad_norm": 30.125,
+      "learning_rate": 3.4962962962962964e-06,
+      "loss": 1.2726,
+      "step": 590
+    },
+    {
+      "epoch": 0.052533333333333335,
+      "grad_norm": 43.5,
+      "learning_rate": 3.5022222222222225e-06,
+      "loss": 1.1579,
+      "step": 591
+    },
+    {
+      "epoch": 0.05262222222222222,
+      "grad_norm": 56.75,
+      "learning_rate": 3.5081481481481483e-06,
+      "loss": 1.1924,
+      "step": 592
+    },
+    {
+      "epoch": 0.05271111111111111,
+      "grad_norm": 138.0,
+      "learning_rate": 3.514074074074074e-06,
+      "loss": 1.022,
+      "step": 593
+    },
+    {
+      "epoch": 0.0528,
+      "grad_norm": 25.125,
+      "learning_rate": 3.52e-06,
+      "loss": 1.0847,
+      "step": 594
+    },
+    {
+      "epoch": 0.05288888888888889,
+      "grad_norm": 58.25,
+      "learning_rate": 3.525925925925926e-06,
+      "loss": 1.2376,
+      "step": 595
+    },
+    {
+      "epoch": 0.052977777777777776,
+      "grad_norm": 46.75,
+      "learning_rate": 3.5318518518518517e-06,
+      "loss": 1.1919,
+      "step": 596
+    },
+    {
+      "epoch": 0.053066666666666665,
+      "grad_norm": 120.0,
+      "learning_rate": 3.5377777777777783e-06,
+      "loss": 1.1256,
+      "step": 597
+    },
+    {
+      "epoch": 0.05315555555555555,
+      "grad_norm": 15.3125,
+      "learning_rate": 3.543703703703704e-06,
+      "loss": 1.2851,
+      "step": 598
+    },
+    {
+      "epoch": 0.05324444444444444,
+      "grad_norm": 29.875,
+      "learning_rate": 3.54962962962963e-06,
+      "loss": 1.0702,
+      "step": 599
+    },
+    {
+      "epoch": 0.05333333333333334,
+      "grad_norm": 14.625,
+      "learning_rate": 3.555555555555556e-06,
+      "loss": 1.0969,
+      "step": 600
+    },
+    {
+      "epoch": 0.053422222222222225,
+      "grad_norm": 29.75,
+      "learning_rate": 3.561481481481482e-06,
+      "loss": 1.0773,
+      "step": 601
+    },
+    {
+      "epoch": 0.05351111111111111,
+      "grad_norm": 29.5,
+      "learning_rate": 3.567407407407408e-06,
+      "loss": 1.287,
+      "step": 602
+    },
+    {
+      "epoch": 0.0536,
+      "grad_norm": 61.0,
+      "learning_rate": 3.5733333333333336e-06,
+      "loss": 0.8934,
+      "step": 603
+    },
+    {
+      "epoch": 0.05368888888888889,
+      "grad_norm": 40.0,
+      "learning_rate": 3.5792592592592597e-06,
+      "loss": 1.2661,
+      "step": 604
+    },
+    {
+      "epoch": 0.05377777777777778,
+      "grad_norm": 35.5,
+      "learning_rate": 3.5851851851851855e-06,
+      "loss": 1.1686,
+      "step": 605
+    },
+    {
+      "epoch": 0.05386666666666667,
+      "grad_norm": 36.5,
+      "learning_rate": 3.5911111111111112e-06,
+      "loss": 1.0214,
+      "step": 606
+    },
+    {
+      "epoch": 0.053955555555555555,
+      "grad_norm": 22.5,
+      "learning_rate": 3.5970370370370374e-06,
+      "loss": 1.2646,
+      "step": 607
+    },
+    {
+      "epoch": 0.054044444444444444,
+      "grad_norm": 29.125,
+      "learning_rate": 3.602962962962963e-06,
+      "loss": 1.0063,
+      "step": 608
+    },
+    {
+      "epoch": 0.05413333333333333,
+      "grad_norm": 29.625,
+      "learning_rate": 3.608888888888889e-06,
+      "loss": 1.1175,
+      "step": 609
+    },
+    {
+      "epoch": 0.05422222222222222,
+      "grad_norm": 26.125,
+      "learning_rate": 3.614814814814815e-06,
+      "loss": 1.0434,
+      "step": 610
+    },
+    {
+      "epoch": 0.05431111111111111,
+      "grad_norm": 31.125,
+      "learning_rate": 3.6207407407407408e-06,
+      "loss": 1.1931,
+      "step": 611
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 23.125,
+      "learning_rate": 3.6266666666666674e-06,
+      "loss": 1.1518,
+      "step": 612
+    },
+    {
+      "epoch": 0.05448888888888889,
+      "grad_norm": 56.5,
+      "learning_rate": 3.632592592592593e-06,
+      "loss": 1.1137,
+      "step": 613
+    },
+    {
+      "epoch": 0.05457777777777778,
+      "grad_norm": 18.125,
+      "learning_rate": 3.638518518518519e-06,
+      "loss": 1.2909,
+      "step": 614
+    },
+    {
+      "epoch": 0.05466666666666667,
+      "grad_norm": 26.0,
+      "learning_rate": 3.644444444444445e-06,
+      "loss": 0.9847,
+      "step": 615
+    },
+    {
+      "epoch": 0.05475555555555556,
+      "grad_norm": 31.625,
+      "learning_rate": 3.6503703703703708e-06,
+      "loss": 1.1759,
+      "step": 616
+    },
+    {
+      "epoch": 0.054844444444444446,
+      "grad_norm": 27.25,
+      "learning_rate": 3.6562962962962965e-06,
+      "loss": 0.9791,
+      "step": 617
+    },
+    {
+      "epoch": 0.054933333333333334,
+      "grad_norm": 29.625,
+      "learning_rate": 3.6622222222222227e-06,
+      "loss": 1.201,
+      "step": 618
+    },
+    {
+      "epoch": 0.05502222222222222,
+      "grad_norm": 36.5,
+      "learning_rate": 3.6681481481481484e-06,
+      "loss": 1.2501,
+      "step": 619
+    },
+    {
+      "epoch": 0.05511111111111111,
+      "grad_norm": 22.875,
+      "learning_rate": 3.674074074074074e-06,
+      "loss": 1.1249,
+      "step": 620
+    },
+    {
+      "epoch": 0.0552,
+      "grad_norm": 11.6875,
+      "learning_rate": 3.6800000000000003e-06,
+      "loss": 1.1149,
+      "step": 621
+    },
+    {
+      "epoch": 0.05528888888888889,
+      "grad_norm": 33.25,
+      "learning_rate": 3.685925925925926e-06,
+      "loss": 1.2135,
+      "step": 622
+    },
+    {
+      "epoch": 0.055377777777777776,
+      "grad_norm": 77.5,
+      "learning_rate": 3.691851851851852e-06,
+      "loss": 1.2393,
+      "step": 623
+    },
+    {
+      "epoch": 0.055466666666666664,
+      "grad_norm": 54.0,
+      "learning_rate": 3.697777777777778e-06,
+      "loss": 1.2062,
+      "step": 624
+    },
+    {
+      "epoch": 0.05555555555555555,
+      "grad_norm": 54.0,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 1.0201,
+      "step": 625
+    },
+    {
+      "epoch": 0.05564444444444445,
+      "grad_norm": 56.75,
+      "learning_rate": 3.7096296296296295e-06,
+      "loss": 1.2156,
+      "step": 626
+    },
+    {
+      "epoch": 0.055733333333333336,
+      "grad_norm": 30.25,
+      "learning_rate": 3.7155555555555557e-06,
+      "loss": 1.2136,
+      "step": 627
+    },
+    {
+      "epoch": 0.055822222222222224,
+      "grad_norm": 21.0,
+      "learning_rate": 3.721481481481482e-06,
+      "loss": 1.203,
+      "step": 628
+    },
+    {
+      "epoch": 0.05591111111111111,
+      "grad_norm": 19.75,
+      "learning_rate": 3.727407407407408e-06,
+      "loss": 1.2054,
+      "step": 629
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 12.3125,
+      "learning_rate": 3.7333333333333337e-06,
+      "loss": 1.1215,
+      "step": 630
+    },
+    {
+      "epoch": 0.05608888888888889,
+      "grad_norm": 32.25,
+      "learning_rate": 3.73925925925926e-06,
+      "loss": 1.0029,
+      "step": 631
+    },
+    {
+      "epoch": 0.05617777777777778,
+      "grad_norm": 20.875,
+      "learning_rate": 3.7451851851851856e-06,
+      "loss": 1.157,
+      "step": 632
+    },
+    {
+      "epoch": 0.056266666666666666,
+      "grad_norm": 28.625,
+      "learning_rate": 3.7511111111111114e-06,
+      "loss": 0.946,
+      "step": 633
+    },
+    {
+      "epoch": 0.056355555555555555,
+      "grad_norm": 19.5,
+      "learning_rate": 3.7570370370370376e-06,
+      "loss": 0.9869,
+      "step": 634
+    },
+    {
+      "epoch": 0.05644444444444444,
+      "grad_norm": 17.375,
+      "learning_rate": 3.7629629629629633e-06,
+      "loss": 1.2238,
+      "step": 635
+    },
+    {
+      "epoch": 0.05653333333333333,
+      "grad_norm": 17.0,
+      "learning_rate": 3.768888888888889e-06,
+      "loss": 1.0982,
+      "step": 636
+    },
+    {
+      "epoch": 0.05662222222222222,
+      "grad_norm": 28.875,
+      "learning_rate": 3.774814814814815e-06,
+      "loss": 1.1333,
+      "step": 637
+    },
+    {
+      "epoch": 0.05671111111111111,
+      "grad_norm": 100.0,
+      "learning_rate": 3.780740740740741e-06,
+      "loss": 1.0195,
+      "step": 638
+    },
+    {
+      "epoch": 0.0568,
+      "grad_norm": 57.5,
+      "learning_rate": 3.7866666666666667e-06,
+      "loss": 1.0143,
+      "step": 639
+    },
+    {
+      "epoch": 0.05688888888888889,
+      "grad_norm": 188.0,
+      "learning_rate": 3.792592592592593e-06,
+      "loss": 1.1285,
+      "step": 640
+    },
+    {
+      "epoch": 0.05697777777777778,
+      "grad_norm": 268.0,
+      "learning_rate": 3.7985185185185186e-06,
+      "loss": 1.0341,
+      "step": 641
+    },
+    {
+      "epoch": 0.05706666666666667,
+      "grad_norm": 79.0,
+      "learning_rate": 3.8044444444444443e-06,
+      "loss": 1.0445,
+      "step": 642
+    },
+    {
+      "epoch": 0.05715555555555556,
+      "grad_norm": 120.5,
+      "learning_rate": 3.810370370370371e-06,
+      "loss": 1.0415,
+      "step": 643
+    },
+    {
+      "epoch": 0.057244444444444445,
+      "grad_norm": 31.75,
+      "learning_rate": 3.816296296296297e-06,
+      "loss": 1.0505,
+      "step": 644
+    },
+    {
+      "epoch": 0.05733333333333333,
+      "grad_norm": 25.375,
+      "learning_rate": 3.8222222222222224e-06,
+      "loss": 1.0958,
+      "step": 645
+    },
+    {
+      "epoch": 0.05742222222222222,
+      "grad_norm": 37.75,
+      "learning_rate": 3.828148148148149e-06,
+      "loss": 1.0616,
+      "step": 646
+    },
+    {
+      "epoch": 0.05751111111111111,
+      "grad_norm": 28.125,
+      "learning_rate": 3.834074074074075e-06,
+      "loss": 1.1199,
+      "step": 647
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 30.125,
+      "learning_rate": 3.8400000000000005e-06,
+      "loss": 1.0492,
+      "step": 648
+    },
+    {
+      "epoch": 0.05768888888888889,
+      "grad_norm": 53.25,
+      "learning_rate": 3.845925925925926e-06,
+      "loss": 1.1051,
+      "step": 649
+    },
+    {
+      "epoch": 0.057777777777777775,
+      "grad_norm": 10.8125,
+      "learning_rate": 3.851851851851852e-06,
+      "loss": 1.0603,
+      "step": 650
+    },
+    {
+      "epoch": 0.057866666666666663,
+      "grad_norm": 9.375,
+      "learning_rate": 3.857777777777778e-06,
+      "loss": 1.0928,
+      "step": 651
+    },
+    {
+      "epoch": 0.05795555555555556,
+      "grad_norm": 13.3125,
+      "learning_rate": 3.863703703703704e-06,
+      "loss": 1.0614,
+      "step": 652
+    },
+    {
+      "epoch": 0.05804444444444445,
+      "grad_norm": 48.75,
+      "learning_rate": 3.86962962962963e-06,
+      "loss": 1.3082,
+      "step": 653
+    },
+    {
+      "epoch": 0.058133333333333335,
+      "grad_norm": 24.625,
+      "learning_rate": 3.875555555555556e-06,
+      "loss": 0.9697,
+      "step": 654
+    },
+    {
+      "epoch": 0.058222222222222224,
+      "grad_norm": 10.3125,
+      "learning_rate": 3.8814814814814816e-06,
+      "loss": 1.2311,
+      "step": 655
+    },
+    {
+      "epoch": 0.05831111111111111,
+      "grad_norm": 11.4375,
+      "learning_rate": 3.887407407407407e-06,
+      "loss": 1.0543,
+      "step": 656
+    },
+    {
+      "epoch": 0.0584,
+      "grad_norm": 13.1875,
+      "learning_rate": 3.893333333333333e-06,
+      "loss": 1.3093,
+      "step": 657
+    },
+    {
+      "epoch": 0.05848888888888889,
+      "grad_norm": 18.375,
+      "learning_rate": 3.89925925925926e-06,
+      "loss": 1.0728,
+      "step": 658
+    },
+    {
+      "epoch": 0.05857777777777778,
+      "grad_norm": 20.625,
+      "learning_rate": 3.905185185185185e-06,
+      "loss": 1.2136,
+      "step": 659
+    },
+    {
+      "epoch": 0.058666666666666666,
+      "grad_norm": 63.25,
+      "learning_rate": 3.911111111111112e-06,
+      "loss": 1.156,
+      "step": 660
+    },
+    {
+      "epoch": 0.058755555555555554,
+      "grad_norm": 63.5,
+      "learning_rate": 3.917037037037038e-06,
+      "loss": 1.1131,
+      "step": 661
+    },
+    {
+      "epoch": 0.05884444444444444,
+      "grad_norm": 91.5,
+      "learning_rate": 3.9229629629629635e-06,
+      "loss": 1.1508,
+      "step": 662
+    },
+    {
+      "epoch": 0.05893333333333333,
+      "grad_norm": 53.25,
+      "learning_rate": 3.928888888888889e-06,
+      "loss": 1.1532,
+      "step": 663
+    },
+    {
+      "epoch": 0.05902222222222222,
+      "grad_norm": 15.5,
+      "learning_rate": 3.934814814814815e-06,
+      "loss": 1.2799,
+      "step": 664
+    },
+    {
+      "epoch": 0.059111111111111114,
+      "grad_norm": 43.25,
+      "learning_rate": 3.940740740740741e-06,
+      "loss": 1.1179,
+      "step": 665
+    },
+    {
+      "epoch": 0.0592,
+      "grad_norm": 28.0,
+      "learning_rate": 3.946666666666667e-06,
+      "loss": 1.2575,
+      "step": 666
+    },
+    {
+      "epoch": 0.05928888888888889,
+      "grad_norm": 143.0,
+      "learning_rate": 3.952592592592593e-06,
+      "loss": 1.2174,
+      "step": 667
+    },
+    {
+      "epoch": 0.05937777777777778,
+      "grad_norm": 159.0,
+      "learning_rate": 3.958518518518519e-06,
+      "loss": 1.1929,
+      "step": 668
+    },
+    {
+      "epoch": 0.05946666666666667,
+      "grad_norm": 43.25,
+      "learning_rate": 3.9644444444444445e-06,
+      "loss": 1.0859,
+      "step": 669
+    },
+    {
+      "epoch": 0.059555555555555556,
+      "grad_norm": 21.25,
+      "learning_rate": 3.97037037037037e-06,
+      "loss": 1.0207,
+      "step": 670
+    },
+    {
+      "epoch": 0.059644444444444444,
+      "grad_norm": 46.5,
+      "learning_rate": 3.976296296296297e-06,
+      "loss": 1.174,
+      "step": 671
+    },
+    {
+      "epoch": 0.05973333333333333,
+      "grad_norm": 47.25,
+      "learning_rate": 3.982222222222223e-06,
+      "loss": 1.1613,
+      "step": 672
+    },
+    {
+      "epoch": 0.05982222222222222,
+      "grad_norm": 33.0,
+      "learning_rate": 3.988148148148148e-06,
+      "loss": 1.1291,
+      "step": 673
+    },
+    {
+      "epoch": 0.05991111111111111,
+      "grad_norm": 33.75,
+      "learning_rate": 3.994074074074075e-06,
+      "loss": 1.0431,
+      "step": 674
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 116.5,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.1213,
+      "step": 675
+    },
+    {
+      "epoch": 0.060088888888888886,
+      "grad_norm": 130.0,
+      "learning_rate": 4.005925925925926e-06,
+      "loss": 0.9577,
+      "step": 676
+    },
+    {
+      "epoch": 0.060177777777777774,
+      "grad_norm": 148.0,
+      "learning_rate": 4.011851851851852e-06,
+      "loss": 1.0642,
+      "step": 677
+    },
+    {
+      "epoch": 0.06026666666666667,
+      "grad_norm": 131.0,
+      "learning_rate": 4.017777777777778e-06,
+      "loss": 1.2898,
+      "step": 678
+    },
+    {
+      "epoch": 0.06035555555555556,
+      "grad_norm": 121.5,
+      "learning_rate": 4.0237037037037045e-06,
+      "loss": 1.1322,
+      "step": 679
+    },
+    {
+      "epoch": 0.060444444444444446,
+      "grad_norm": 11.0,
+      "learning_rate": 4.02962962962963e-06,
+      "loss": 1.1269,
+      "step": 680
+    },
+    {
+      "epoch": 0.060533333333333335,
+      "grad_norm": 17.75,
+      "learning_rate": 4.035555555555556e-06,
+      "loss": 1.1911,
+      "step": 681
+    },
+    {
+      "epoch": 0.06062222222222222,
+      "grad_norm": 59.5,
+      "learning_rate": 4.041481481481482e-06,
+      "loss": 0.9455,
+      "step": 682
+    },
+    {
+      "epoch": 0.06071111111111111,
+      "grad_norm": 73.0,
+      "learning_rate": 4.0474074074074075e-06,
+      "loss": 1.106,
+      "step": 683
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 51.75,
+      "learning_rate": 4.053333333333333e-06,
+      "loss": 1.1375,
+      "step": 684
+    },
+    {
+      "epoch": 0.06088888888888889,
+      "grad_norm": 96.5,
+      "learning_rate": 4.05925925925926e-06,
+      "loss": 1.0839,
+      "step": 685
+    },
+    {
+      "epoch": 0.06097777777777778,
+      "grad_norm": 34.0,
+      "learning_rate": 4.0651851851851855e-06,
+      "loss": 0.9496,
+      "step": 686
+    },
+    {
+      "epoch": 0.061066666666666665,
+      "grad_norm": 50.25,
+      "learning_rate": 4.071111111111111e-06,
+      "loss": 1.1221,
+      "step": 687
+    },
+    {
+      "epoch": 0.06115555555555555,
+      "grad_norm": 30.875,
+      "learning_rate": 4.077037037037037e-06,
+      "loss": 1.1167,
+      "step": 688
+    },
+    {
+      "epoch": 0.06124444444444444,
+      "grad_norm": 50.25,
+      "learning_rate": 4.082962962962964e-06,
+      "loss": 1.2134,
+      "step": 689
+    },
+    {
+      "epoch": 0.06133333333333333,
+      "grad_norm": 26.5,
+      "learning_rate": 4.088888888888889e-06,
+      "loss": 1.1893,
+      "step": 690
+    },
+    {
+      "epoch": 0.061422222222222225,
+      "grad_norm": 16.25,
+      "learning_rate": 4.094814814814815e-06,
+      "loss": 0.8992,
+      "step": 691
+    },
+    {
+      "epoch": 0.061511111111111114,
+      "grad_norm": 30.125,
+      "learning_rate": 4.100740740740741e-06,
+      "loss": 0.9613,
+      "step": 692
+    },
+    {
+      "epoch": 0.0616,
+      "grad_norm": 20.0,
+      "learning_rate": 4.1066666666666674e-06,
+      "loss": 1.0279,
+      "step": 693
+    },
+    {
+      "epoch": 0.06168888888888889,
+      "grad_norm": 13.25,
+      "learning_rate": 4.112592592592593e-06,
+      "loss": 1.0105,
+      "step": 694
+    },
+    {
+      "epoch": 0.06177777777777778,
+      "grad_norm": 27.125,
+      "learning_rate": 4.118518518518519e-06,
+      "loss": 1.1867,
+      "step": 695
+    },
+    {
+      "epoch": 0.06186666666666667,
+      "grad_norm": 78.0,
+      "learning_rate": 4.124444444444445e-06,
+      "loss": 1.0181,
+      "step": 696
+    },
+    {
+      "epoch": 0.061955555555555555,
+      "grad_norm": 47.75,
+      "learning_rate": 4.13037037037037e-06,
+      "loss": 0.9885,
+      "step": 697
+    },
+    {
+      "epoch": 0.062044444444444444,
+      "grad_norm": 56.5,
+      "learning_rate": 4.136296296296297e-06,
+      "loss": 1.0773,
+      "step": 698
+    },
+    {
+      "epoch": 0.06213333333333333,
+      "grad_norm": 28.375,
+      "learning_rate": 4.142222222222223e-06,
+      "loss": 1.2266,
+      "step": 699
+    },
+    {
+      "epoch": 0.06222222222222222,
+      "grad_norm": 30.125,
+      "learning_rate": 4.1481481481481485e-06,
+      "loss": 1.0714,
+      "step": 700
+    },
+    {
+      "epoch": 0.06231111111111111,
+      "grad_norm": 23.625,
+      "learning_rate": 4.154074074074074e-06,
+      "loss": 1.0256,
+      "step": 701
+    },
+    {
+      "epoch": 0.0624,
+      "grad_norm": 27.375,
+      "learning_rate": 4.16e-06,
+      "loss": 0.9773,
+      "step": 702
+    },
+    {
+      "epoch": 0.062488888888888885,
+      "grad_norm": 18.125,
+      "learning_rate": 4.165925925925926e-06,
+      "loss": 1.0056,
+      "step": 703
+    },
+    {
+      "epoch": 0.06257777777777777,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.171851851851852e-06,
+      "loss": 0.9904,
+      "step": 704
+    },
+    {
+      "epoch": 0.06266666666666666,
+      "grad_norm": 13.5,
+      "learning_rate": 4.177777777777778e-06,
+      "loss": 1.1006,
+      "step": 705
+    },
+    {
+      "epoch": 0.06275555555555555,
+      "grad_norm": 24.25,
+      "learning_rate": 4.183703703703705e-06,
+      "loss": 1.0261,
+      "step": 706
+    },
+    {
+      "epoch": 0.06284444444444444,
+      "grad_norm": 29.0,
+      "learning_rate": 4.18962962962963e-06,
+      "loss": 1.0851,
+      "step": 707
+    },
+    {
+      "epoch": 0.06293333333333333,
+      "grad_norm": 13.0,
+      "learning_rate": 4.195555555555556e-06,
+      "loss": 1.0507,
+      "step": 708
+    },
+    {
+      "epoch": 0.06302222222222222,
+      "grad_norm": 9.9375,
+      "learning_rate": 4.201481481481482e-06,
+      "loss": 1.145,
+      "step": 709
+    },
+    {
+      "epoch": 0.06311111111111112,
+      "grad_norm": 21.0,
+      "learning_rate": 4.207407407407408e-06,
+      "loss": 1.1426,
+      "step": 710
+    },
+    {
+      "epoch": 0.0632,
+      "grad_norm": 78.5,
+      "learning_rate": 4.213333333333333e-06,
+      "loss": 1.1366,
+      "step": 711
+    },
+    {
+      "epoch": 0.0632888888888889,
+      "grad_norm": 56.0,
+      "learning_rate": 4.21925925925926e-06,
+      "loss": 1.1032,
+      "step": 712
+    },
+    {
+      "epoch": 0.06337777777777778,
+      "grad_norm": 15.0,
+      "learning_rate": 4.225185185185186e-06,
+      "loss": 1.1046,
+      "step": 713
+    },
+    {
+      "epoch": 0.06346666666666667,
+      "grad_norm": 28.5,
+      "learning_rate": 4.2311111111111114e-06,
+      "loss": 1.0281,
+      "step": 714
+    },
+    {
+      "epoch": 0.06355555555555556,
+      "grad_norm": 32.0,
+      "learning_rate": 4.237037037037037e-06,
+      "loss": 1.0088,
+      "step": 715
+    },
+    {
+      "epoch": 0.06364444444444445,
+      "grad_norm": 55.5,
+      "learning_rate": 4.242962962962963e-06,
+      "loss": 0.9971,
+      "step": 716
+    },
+    {
+      "epoch": 0.06373333333333334,
+      "grad_norm": 28.75,
+      "learning_rate": 4.248888888888889e-06,
+      "loss": 1.1258,
+      "step": 717
+    },
+    {
+      "epoch": 0.06382222222222222,
+      "grad_norm": 10.5625,
+      "learning_rate": 4.254814814814815e-06,
+      "loss": 0.8606,
+      "step": 718
+    },
+    {
+      "epoch": 0.06391111111111111,
+      "grad_norm": 15.875,
+      "learning_rate": 4.260740740740741e-06,
+      "loss": 1.098,
+      "step": 719
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 32.5,
+      "learning_rate": 4.266666666666668e-06,
+      "loss": 0.9896,
+      "step": 720
+    },
+    {
+      "epoch": 0.06408888888888889,
+      "grad_norm": 33.25,
+      "learning_rate": 4.272592592592593e-06,
+      "loss": 1.0333,
+      "step": 721
+    },
+    {
+      "epoch": 0.06417777777777778,
+      "grad_norm": 14.0,
+      "learning_rate": 4.278518518518519e-06,
+      "loss": 1.01,
+      "step": 722
+    },
+    {
+      "epoch": 0.06426666666666667,
+      "grad_norm": 33.75,
+      "learning_rate": 4.284444444444445e-06,
+      "loss": 0.9981,
+      "step": 723
+    },
+    {
+      "epoch": 0.06435555555555555,
+      "grad_norm": 140.0,
+      "learning_rate": 4.290370370370371e-06,
+      "loss": 0.9732,
+      "step": 724
+    },
+    {
+      "epoch": 0.06444444444444444,
+      "grad_norm": 54.0,
+      "learning_rate": 4.296296296296296e-06,
+      "loss": 1.1874,
+      "step": 725
+    },
+    {
+      "epoch": 0.06453333333333333,
+      "grad_norm": 48.0,
+      "learning_rate": 4.302222222222223e-06,
+      "loss": 1.1188,
+      "step": 726
+    },
+    {
+      "epoch": 0.06462222222222222,
+      "grad_norm": 46.75,
+      "learning_rate": 4.308148148148149e-06,
+      "loss": 0.961,
+      "step": 727
+    },
+    {
+      "epoch": 0.06471111111111111,
+      "grad_norm": 70.0,
+      "learning_rate": 4.314074074074074e-06,
+      "loss": 1.0773,
+      "step": 728
+    },
+    {
+      "epoch": 0.0648,
+      "grad_norm": 49.25,
+      "learning_rate": 4.32e-06,
+      "loss": 1.2543,
+      "step": 729
+    },
+    {
+      "epoch": 0.06488888888888888,
+      "grad_norm": 43.0,
+      "learning_rate": 4.325925925925926e-06,
+      "loss": 1.177,
+      "step": 730
+    },
+    {
+      "epoch": 0.06497777777777777,
+      "grad_norm": 15.375,
+      "learning_rate": 4.3318518518518525e-06,
+      "loss": 0.982,
+      "step": 731
+    },
+    {
+      "epoch": 0.06506666666666666,
+      "grad_norm": 10.5625,
+      "learning_rate": 4.337777777777778e-06,
+      "loss": 1.1697,
+      "step": 732
+    },
+    {
+      "epoch": 0.06515555555555555,
+      "grad_norm": 58.0,
+      "learning_rate": 4.343703703703704e-06,
+      "loss": 1.003,
+      "step": 733
+    },
+    {
+      "epoch": 0.06524444444444444,
+      "grad_norm": 66.0,
+      "learning_rate": 4.34962962962963e-06,
+      "loss": 1.1894,
+      "step": 734
+    },
+    {
+      "epoch": 0.06533333333333333,
+      "grad_norm": 185.0,
+      "learning_rate": 4.3555555555555555e-06,
+      "loss": 1.0279,
+      "step": 735
+    },
+    {
+      "epoch": 0.06542222222222223,
+      "grad_norm": 89.5,
+      "learning_rate": 4.361481481481482e-06,
+      "loss": 1.0144,
+      "step": 736
+    },
+    {
+      "epoch": 0.06551111111111112,
+      "grad_norm": 55.5,
+      "learning_rate": 4.367407407407408e-06,
+      "loss": 1.0821,
+      "step": 737
+    },
+    {
+      "epoch": 0.0656,
+      "grad_norm": 83.0,
+      "learning_rate": 4.3733333333333335e-06,
+      "loss": 1.2513,
+      "step": 738
+    },
+    {
+      "epoch": 0.0656888888888889,
+      "grad_norm": 60.25,
+      "learning_rate": 4.37925925925926e-06,
+      "loss": 1.1043,
+      "step": 739
+    },
+    {
+      "epoch": 0.06577777777777778,
+      "grad_norm": 160.0,
+      "learning_rate": 4.385185185185186e-06,
+      "loss": 1.0672,
+      "step": 740
+    },
+    {
+      "epoch": 0.06586666666666667,
+      "grad_norm": 235.0,
+      "learning_rate": 4.391111111111112e-06,
+      "loss": 1.045,
+      "step": 741
+    },
+    {
+      "epoch": 0.06595555555555556,
+      "grad_norm": 150.0,
+      "learning_rate": 4.397037037037037e-06,
+      "loss": 1.2013,
+      "step": 742
+    },
+    {
+      "epoch": 0.06604444444444445,
+      "grad_norm": 124.5,
+      "learning_rate": 4.402962962962963e-06,
+      "loss": 1.0656,
+      "step": 743
+    },
+    {
+      "epoch": 0.06613333333333334,
+      "grad_norm": 38.75,
+      "learning_rate": 4.408888888888889e-06,
+      "loss": 1.0831,
+      "step": 744
+    },
+    {
+      "epoch": 0.06622222222222222,
+      "grad_norm": 16.0,
+      "learning_rate": 4.4148148148148154e-06,
+      "loss": 1.1137,
+      "step": 745
+    },
+    {
+      "epoch": 0.06631111111111111,
+      "grad_norm": 17.375,
+      "learning_rate": 4.420740740740741e-06,
+      "loss": 0.9669,
+      "step": 746
+    },
+    {
+      "epoch": 0.0664,
+      "grad_norm": 14.1875,
+      "learning_rate": 4.426666666666667e-06,
+      "loss": 1.1022,
+      "step": 747
+    },
+    {
+      "epoch": 0.06648888888888889,
+      "grad_norm": 20.875,
+      "learning_rate": 4.432592592592593e-06,
+      "loss": 1.0302,
+      "step": 748
+    },
+    {
+      "epoch": 0.06657777777777778,
+      "grad_norm": 47.5,
+      "learning_rate": 4.438518518518518e-06,
+      "loss": 1.0697,
+      "step": 749
+    },
+    {
+      "epoch": 0.06666666666666667,
+      "grad_norm": 17.5,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.1617,
+      "step": 750
+    },
+    {
+      "epoch": 0.06675555555555555,
+      "grad_norm": 75.5,
+      "learning_rate": 4.450370370370371e-06,
+      "loss": 0.8373,
+      "step": 751
+    },
+    {
+      "epoch": 0.06684444444444444,
+      "grad_norm": 11.5625,
+      "learning_rate": 4.4562962962962965e-06,
+      "loss": 0.9627,
+      "step": 752
+    },
+    {
+      "epoch": 0.06693333333333333,
+      "grad_norm": 30.125,
+      "learning_rate": 4.462222222222223e-06,
+      "loss": 1.03,
+      "step": 753
+    },
+    {
+      "epoch": 0.06702222222222222,
+      "grad_norm": 25.125,
+      "learning_rate": 4.468148148148149e-06,
+      "loss": 1.0237,
+      "step": 754
+    },
+    {
+      "epoch": 0.06711111111111111,
+      "grad_norm": 34.5,
+      "learning_rate": 4.4740740740740746e-06,
+      "loss": 0.9877,
+      "step": 755
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 22.625,
+      "learning_rate": 4.48e-06,
+      "loss": 1.0924,
+      "step": 756
+    },
+    {
+      "epoch": 0.06728888888888888,
+      "grad_norm": 28.625,
+      "learning_rate": 4.485925925925926e-06,
+      "loss": 0.9866,
+      "step": 757
+    },
+    {
+      "epoch": 0.06737777777777777,
+      "grad_norm": 19.125,
+      "learning_rate": 4.491851851851853e-06,
+      "loss": 1.0774,
+      "step": 758
+    },
+    {
+      "epoch": 0.06746666666666666,
+      "grad_norm": 51.0,
+      "learning_rate": 4.497777777777778e-06,
+      "loss": 0.9677,
+      "step": 759
+    },
+    {
+      "epoch": 0.06755555555555555,
+      "grad_norm": 16.25,
+      "learning_rate": 4.503703703703704e-06,
+      "loss": 1.0342,
+      "step": 760
+    },
+    {
+      "epoch": 0.06764444444444444,
+      "grad_norm": 60.5,
+      "learning_rate": 4.50962962962963e-06,
+      "loss": 1.2428,
+      "step": 761
+    },
+    {
+      "epoch": 0.06773333333333334,
+      "grad_norm": 57.75,
+      "learning_rate": 4.515555555555556e-06,
+      "loss": 0.8983,
+      "step": 762
+    },
+    {
+      "epoch": 0.06782222222222223,
+      "grad_norm": 71.5,
+      "learning_rate": 4.521481481481481e-06,
+      "loss": 1.2271,
+      "step": 763
+    },
+    {
+      "epoch": 0.06791111111111112,
+      "grad_norm": 25.75,
+      "learning_rate": 4.527407407407408e-06,
+      "loss": 1.063,
+      "step": 764
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 74.5,
+      "learning_rate": 4.533333333333334e-06,
+      "loss": 1.0649,
+      "step": 765
+    },
+    {
+      "epoch": 0.0680888888888889,
+      "grad_norm": 64.0,
+      "learning_rate": 4.5392592592592594e-06,
+      "loss": 1.1402,
+      "step": 766
+    },
+    {
+      "epoch": 0.06817777777777778,
+      "grad_norm": 63.25,
+      "learning_rate": 4.545185185185186e-06,
+      "loss": 1.0373,
+      "step": 767
+    },
+    {
+      "epoch": 0.06826666666666667,
+      "grad_norm": 69.0,
+      "learning_rate": 4.551111111111112e-06,
+      "loss": 1.0294,
+      "step": 768
+    },
+    {
+      "epoch": 0.06835555555555556,
+      "grad_norm": 42.75,
+      "learning_rate": 4.5570370370370375e-06,
+      "loss": 1.1386,
+      "step": 769
+    },
+    {
+      "epoch": 0.06844444444444445,
+      "grad_norm": 22.875,
+      "learning_rate": 4.562962962962963e-06,
+      "loss": 0.9291,
+      "step": 770
+    },
+    {
+      "epoch": 0.06853333333333333,
+      "grad_norm": 12.9375,
+      "learning_rate": 4.568888888888889e-06,
+      "loss": 1.1283,
+      "step": 771
+    },
+    {
+      "epoch": 0.06862222222222222,
+      "grad_norm": 29.75,
+      "learning_rate": 4.574814814814816e-06,
+      "loss": 0.981,
+      "step": 772
+    },
+    {
+      "epoch": 0.06871111111111111,
+      "grad_norm": 77.5,
+      "learning_rate": 4.580740740740741e-06,
+      "loss": 1.043,
+      "step": 773
+    },
+    {
+      "epoch": 0.0688,
+      "grad_norm": 45.75,
+      "learning_rate": 4.586666666666667e-06,
+      "loss": 1.0731,
+      "step": 774
+    },
+    {
+      "epoch": 0.06888888888888889,
+      "grad_norm": 27.0,
+      "learning_rate": 4.592592592592593e-06,
+      "loss": 0.9298,
+      "step": 775
+    },
+    {
+      "epoch": 0.06897777777777778,
+      "grad_norm": 15.0,
+      "learning_rate": 4.5985185185185186e-06,
+      "loss": 0.7747,
+      "step": 776
+    },
+    {
+      "epoch": 0.06906666666666667,
+      "grad_norm": 13.0625,
+      "learning_rate": 4.604444444444444e-06,
+      "loss": 1.0377,
+      "step": 777
+    },
+    {
+      "epoch": 0.06915555555555555,
+      "grad_norm": 29.375,
+      "learning_rate": 4.610370370370371e-06,
+      "loss": 1.1766,
+      "step": 778
+    },
+    {
+      "epoch": 0.06924444444444444,
+      "grad_norm": 23.0,
+      "learning_rate": 4.616296296296297e-06,
+      "loss": 1.0957,
+      "step": 779
+    },
+    {
+      "epoch": 0.06933333333333333,
+      "grad_norm": 11.0,
+      "learning_rate": 4.622222222222222e-06,
+      "loss": 1.0044,
+      "step": 780
+    },
+    {
+      "epoch": 0.06942222222222222,
+      "grad_norm": 23.5,
+      "learning_rate": 4.628148148148148e-06,
+      "loss": 1.0502,
+      "step": 781
+    },
+    {
+      "epoch": 0.0695111111111111,
+      "grad_norm": 32.5,
+      "learning_rate": 4.634074074074075e-06,
+      "loss": 1.2083,
+      "step": 782
+    },
+    {
+      "epoch": 0.0696,
+      "grad_norm": 76.5,
+      "learning_rate": 4.6400000000000005e-06,
+      "loss": 1.0362,
+      "step": 783
+    },
+    {
+      "epoch": 0.06968888888888888,
+      "grad_norm": 37.75,
+      "learning_rate": 4.645925925925926e-06,
+      "loss": 1.0807,
+      "step": 784
+    },
+    {
+      "epoch": 0.06977777777777777,
+      "grad_norm": 35.25,
+      "learning_rate": 4.651851851851853e-06,
+      "loss": 0.9496,
+      "step": 785
+    },
+    {
+      "epoch": 0.06986666666666666,
+      "grad_norm": 88.0,
+      "learning_rate": 4.6577777777777785e-06,
+      "loss": 1.0296,
+      "step": 786
+    },
+    {
+      "epoch": 0.06995555555555556,
+      "grad_norm": 74.5,
+      "learning_rate": 4.663703703703704e-06,
+      "loss": 1.0848,
+      "step": 787
+    },
+    {
+      "epoch": 0.07004444444444445,
+      "grad_norm": 20.25,
+      "learning_rate": 4.66962962962963e-06,
+      "loss": 1.0708,
+      "step": 788
+    },
+    {
+      "epoch": 0.07013333333333334,
+      "grad_norm": 56.0,
+      "learning_rate": 4.675555555555556e-06,
+      "loss": 0.963,
+      "step": 789
+    },
+    {
+      "epoch": 0.07022222222222223,
+      "grad_norm": 59.0,
+      "learning_rate": 4.6814814814814815e-06,
+      "loss": 1.0138,
+      "step": 790
+    },
+    {
+      "epoch": 0.07031111111111112,
+      "grad_norm": 59.25,
+      "learning_rate": 4.687407407407408e-06,
+      "loss": 1.0761,
+      "step": 791
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 34.0,
+      "learning_rate": 4.693333333333334e-06,
+      "loss": 1.0452,
+      "step": 792
+    },
+    {
+      "epoch": 0.07048888888888889,
+      "grad_norm": 25.875,
+      "learning_rate": 4.69925925925926e-06,
+      "loss": 0.9051,
+      "step": 793
+    },
+    {
+      "epoch": 0.07057777777777778,
+      "grad_norm": 46.0,
+      "learning_rate": 4.705185185185185e-06,
+      "loss": 1.0745,
+      "step": 794
+    },
+    {
+      "epoch": 0.07066666666666667,
+      "grad_norm": 68.5,
+      "learning_rate": 4.711111111111111e-06,
+      "loss": 1.0009,
+      "step": 795
+    },
+    {
+      "epoch": 0.07075555555555556,
+      "grad_norm": 31.875,
+      "learning_rate": 4.717037037037037e-06,
+      "loss": 1.2865,
+      "step": 796
+    },
+    {
+      "epoch": 0.07084444444444445,
+      "grad_norm": 18.875,
+      "learning_rate": 4.722962962962963e-06,
+      "loss": 1.0875,
+      "step": 797
+    },
+    {
+      "epoch": 0.07093333333333333,
+      "grad_norm": 24.75,
+      "learning_rate": 4.728888888888889e-06,
+      "loss": 0.8993,
+      "step": 798
+    },
+    {
+      "epoch": 0.07102222222222222,
+      "grad_norm": 17.5,
+      "learning_rate": 4.734814814814816e-06,
+      "loss": 1.2044,
+      "step": 799
+    },
+    {
+      "epoch": 0.07111111111111111,
+      "grad_norm": 45.5,
+      "learning_rate": 4.7407407407407415e-06,
+      "loss": 1.0602,
+      "step": 800
+    },
+    {
+      "epoch": 0.0712,
+      "grad_norm": 30.375,
+      "learning_rate": 4.746666666666667e-06,
+      "loss": 0.8644,
+      "step": 801
+    },
+    {
+      "epoch": 0.07128888888888889,
+      "grad_norm": 20.375,
+      "learning_rate": 4.752592592592593e-06,
+      "loss": 1.0675,
+      "step": 802
+    },
+    {
+      "epoch": 0.07137777777777778,
+      "grad_norm": 14.0625,
+      "learning_rate": 4.758518518518519e-06,
+      "loss": 1.0944,
+      "step": 803
+    },
+    {
+      "epoch": 0.07146666666666666,
+      "grad_norm": 17.0,
+      "learning_rate": 4.7644444444444445e-06,
+      "loss": 1.067,
+      "step": 804
+    },
+    {
+      "epoch": 0.07155555555555555,
+      "grad_norm": 11.125,
+      "learning_rate": 4.770370370370371e-06,
+      "loss": 1.0066,
+      "step": 805
+    },
+    {
+      "epoch": 0.07164444444444444,
+      "grad_norm": 37.5,
+      "learning_rate": 4.776296296296297e-06,
+      "loss": 1.0111,
+      "step": 806
+    },
+    {
+      "epoch": 0.07173333333333333,
+      "grad_norm": 42.75,
+      "learning_rate": 4.7822222222222226e-06,
+      "loss": 1.0235,
+      "step": 807
+    },
+    {
+      "epoch": 0.07182222222222222,
+      "grad_norm": 18.375,
+      "learning_rate": 4.788148148148148e-06,
+      "loss": 1.2106,
+      "step": 808
+    },
+    {
+      "epoch": 0.0719111111111111,
+      "grad_norm": 31.5,
+      "learning_rate": 4.794074074074074e-06,
+      "loss": 1.2418,
+      "step": 809
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 17.625,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.1508,
+      "step": 810
+    },
+    {
+      "epoch": 0.07208888888888888,
+      "grad_norm": 19.75,
+      "learning_rate": 4.805925925925926e-06,
+      "loss": 1.0852,
+      "step": 811
+    },
+    {
+      "epoch": 0.07217777777777777,
+      "grad_norm": 53.0,
+      "learning_rate": 4.811851851851852e-06,
+      "loss": 0.8923,
+      "step": 812
+    },
+    {
+      "epoch": 0.07226666666666667,
+      "grad_norm": 24.875,
+      "learning_rate": 4.817777777777779e-06,
+      "loss": 1.0401,
+      "step": 813
+    },
+    {
+      "epoch": 0.07235555555555556,
+      "grad_norm": 52.25,
+      "learning_rate": 4.8237037037037045e-06,
+      "loss": 1.051,
+      "step": 814
+    },
+    {
+      "epoch": 0.07244444444444445,
+      "grad_norm": 26.125,
+      "learning_rate": 4.82962962962963e-06,
+      "loss": 1.1046,
+      "step": 815
+    },
+    {
+      "epoch": 0.07253333333333334,
+      "grad_norm": 22.5,
+      "learning_rate": 4.835555555555556e-06,
+      "loss": 1.052,
+      "step": 816
+    },
+    {
+      "epoch": 0.07262222222222223,
+      "grad_norm": 29.375,
+      "learning_rate": 4.841481481481482e-06,
+      "loss": 1.0253,
+      "step": 817
+    },
+    {
+      "epoch": 0.07271111111111112,
+      "grad_norm": 18.5,
+      "learning_rate": 4.847407407407408e-06,
+      "loss": 1.1075,
+      "step": 818
+    },
+    {
+      "epoch": 0.0728,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.853333333333334e-06,
+      "loss": 0.914,
+      "step": 819
+    },
+    {
+      "epoch": 0.07288888888888889,
+      "grad_norm": 9.3125,
+      "learning_rate": 4.85925925925926e-06,
+      "loss": 1.1162,
+      "step": 820
+    },
+    {
+      "epoch": 0.07297777777777778,
+      "grad_norm": 15.3125,
+      "learning_rate": 4.8651851851851855e-06,
+      "loss": 0.9541,
+      "step": 821
+    },
+    {
+      "epoch": 0.07306666666666667,
+      "grad_norm": 19.875,
+      "learning_rate": 4.871111111111111e-06,
+      "loss": 1.0685,
+      "step": 822
+    },
+    {
+      "epoch": 0.07315555555555556,
+      "grad_norm": 26.0,
+      "learning_rate": 4.877037037037037e-06,
+      "loss": 1.0583,
+      "step": 823
+    },
+    {
+      "epoch": 0.07324444444444445,
+      "grad_norm": 19.5,
+      "learning_rate": 4.882962962962964e-06,
+      "loss": 1.0506,
+      "step": 824
+    },
+    {
+      "epoch": 0.07333333333333333,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.888888888888889e-06,
+      "loss": 0.9984,
+      "step": 825
+    },
+    {
+      "epoch": 0.07342222222222222,
+      "grad_norm": 10.5,
+      "learning_rate": 4.894814814814815e-06,
+      "loss": 1.0167,
+      "step": 826
+    },
+    {
+      "epoch": 0.07351111111111111,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.900740740740741e-06,
+      "loss": 1.2049,
+      "step": 827
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 28.25,
+      "learning_rate": 4.9066666666666666e-06,
+      "loss": 0.9359,
+      "step": 828
+    },
+    {
+      "epoch": 0.07368888888888889,
+      "grad_norm": 23.125,
+      "learning_rate": 4.912592592592593e-06,
+      "loss": 1.0808,
+      "step": 829
+    },
+    {
+      "epoch": 0.07377777777777778,
+      "grad_norm": 22.375,
+      "learning_rate": 4.918518518518519e-06,
+      "loss": 0.8995,
+      "step": 830
+    },
+    {
+      "epoch": 0.07386666666666666,
+      "grad_norm": 22.75,
+      "learning_rate": 4.924444444444445e-06,
+      "loss": 0.9842,
+      "step": 831
+    },
+    {
+      "epoch": 0.07395555555555555,
+      "grad_norm": 17.875,
+      "learning_rate": 4.930370370370371e-06,
+      "loss": 0.9832,
+      "step": 832
+    },
+    {
+      "epoch": 0.07404444444444444,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.936296296296297e-06,
+      "loss": 0.9475,
+      "step": 833
+    },
+    {
+      "epoch": 0.07413333333333333,
+      "grad_norm": 19.875,
+      "learning_rate": 4.942222222222223e-06,
+      "loss": 1.0851,
+      "step": 834
+    },
+    {
+      "epoch": 0.07422222222222222,
+      "grad_norm": 18.625,
+      "learning_rate": 4.9481481481481485e-06,
+      "loss": 1.0909,
+      "step": 835
+    },
+    {
+      "epoch": 0.0743111111111111,
+      "grad_norm": 15.375,
+      "learning_rate": 4.954074074074074e-06,
+      "loss": 1.0086,
+      "step": 836
+    },
+    {
+      "epoch": 0.0744,
+      "grad_norm": 20.375,
+      "learning_rate": 4.960000000000001e-06,
+      "loss": 1.1095,
+      "step": 837
+    },
+    {
+      "epoch": 0.07448888888888888,
+      "grad_norm": 10.0,
+      "learning_rate": 4.9659259259259265e-06,
+      "loss": 0.9838,
+      "step": 838
+    },
+    {
+      "epoch": 0.07457777777777778,
+      "grad_norm": 21.625,
+      "learning_rate": 4.971851851851852e-06,
+      "loss": 0.9988,
+      "step": 839
+    },
+    {
+      "epoch": 0.07466666666666667,
+      "grad_norm": 25.0,
+      "learning_rate": 4.977777777777778e-06,
+      "loss": 1.1166,
+      "step": 840
+    },
+    {
+      "epoch": 0.07475555555555556,
+      "grad_norm": 74.0,
+      "learning_rate": 4.983703703703704e-06,
+      "loss": 0.9179,
+      "step": 841
+    },
+    {
+      "epoch": 0.07484444444444445,
+      "grad_norm": 66.0,
+      "learning_rate": 4.9896296296296295e-06,
+      "loss": 1.1782,
+      "step": 842
+    },
+    {
+      "epoch": 0.07493333333333334,
+      "grad_norm": 45.75,
+      "learning_rate": 4.995555555555556e-06,
+      "loss": 1.0508,
+      "step": 843
+    },
+    {
+      "epoch": 0.07502222222222223,
+      "grad_norm": 16.0,
+      "learning_rate": 5.001481481481482e-06,
+      "loss": 0.9969,
+      "step": 844
+    },
+    {
+      "epoch": 0.07511111111111111,
+      "grad_norm": 23.625,
+      "learning_rate": 5.007407407407408e-06,
+      "loss": 0.856,
+      "step": 845
+    },
+    {
+      "epoch": 0.0752,
+      "grad_norm": 10.0,
+      "learning_rate": 5.013333333333333e-06,
+      "loss": 1.0301,
+      "step": 846
+    },
+    {
+      "epoch": 0.07528888888888889,
+      "grad_norm": 19.0,
+      "learning_rate": 5.019259259259259e-06,
+      "loss": 1.0441,
+      "step": 847
+    },
+    {
+      "epoch": 0.07537777777777778,
+      "grad_norm": 7.5,
+      "learning_rate": 5.025185185185185e-06,
+      "loss": 1.1653,
+      "step": 848
+    },
+    {
+      "epoch": 0.07546666666666667,
+      "grad_norm": 15.375,
+      "learning_rate": 5.031111111111111e-06,
+      "loss": 0.9441,
+      "step": 849
+    },
+    {
+      "epoch": 0.07555555555555556,
+      "grad_norm": 19.875,
+      "learning_rate": 5.037037037037037e-06,
+      "loss": 0.9735,
+      "step": 850
+    },
+    {
+      "epoch": 0.07564444444444444,
+      "grad_norm": 22.875,
+      "learning_rate": 5.042962962962963e-06,
+      "loss": 0.9408,
+      "step": 851
+    },
+    {
+      "epoch": 0.07573333333333333,
+      "grad_norm": 15.3125,
+      "learning_rate": 5.0488888888888895e-06,
+      "loss": 1.1003,
+      "step": 852
+    },
+    {
+      "epoch": 0.07582222222222222,
+      "grad_norm": 18.0,
+      "learning_rate": 5.054814814814816e-06,
+      "loss": 1.0347,
+      "step": 853
+    },
+    {
+      "epoch": 0.07591111111111111,
+      "grad_norm": 20.875,
+      "learning_rate": 5.060740740740742e-06,
+      "loss": 1.0857,
+      "step": 854
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 23.5,
+      "learning_rate": 5.0666666666666676e-06,
+      "loss": 1.1775,
+      "step": 855
+    },
+    {
+      "epoch": 0.07608888888888889,
+      "grad_norm": 10.75,
+      "learning_rate": 5.072592592592593e-06,
+      "loss": 1.1004,
+      "step": 856
+    },
+    {
+      "epoch": 0.07617777777777777,
+      "grad_norm": 30.875,
+      "learning_rate": 5.078518518518519e-06,
+      "loss": 0.9809,
+      "step": 857
+    },
+    {
+      "epoch": 0.07626666666666666,
+      "grad_norm": 30.375,
+      "learning_rate": 5.084444444444445e-06,
+      "loss": 1.0408,
+      "step": 858
+    },
+    {
+      "epoch": 0.07635555555555555,
+      "grad_norm": 68.5,
+      "learning_rate": 5.090370370370371e-06,
+      "loss": 1.0698,
+      "step": 859
+    },
+    {
+      "epoch": 0.07644444444444444,
+      "grad_norm": 74.0,
+      "learning_rate": 5.096296296296297e-06,
+      "loss": 1.0136,
+      "step": 860
+    },
+    {
+      "epoch": 0.07653333333333333,
+      "grad_norm": 74.0,
+      "learning_rate": 5.102222222222223e-06,
+      "loss": 0.9514,
+      "step": 861
+    },
+    {
+      "epoch": 0.07662222222222222,
+      "grad_norm": 46.5,
+      "learning_rate": 5.108148148148149e-06,
+      "loss": 1.1253,
+      "step": 862
+    },
+    {
+      "epoch": 0.0767111111111111,
+      "grad_norm": 16.375,
+      "learning_rate": 5.114074074074074e-06,
+      "loss": 0.988,
+      "step": 863
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 7.78125,
+      "learning_rate": 5.12e-06,
+      "loss": 0.9667,
+      "step": 864
+    },
+    {
+      "epoch": 0.0768888888888889,
+      "grad_norm": 26.5,
+      "learning_rate": 5.125925925925927e-06,
+      "loss": 1.0608,
+      "step": 865
+    },
+    {
+      "epoch": 0.07697777777777778,
+      "grad_norm": 177.0,
+      "learning_rate": 5.1318518518518524e-06,
+      "loss": 0.8872,
+      "step": 866
+    },
+    {
+      "epoch": 0.07706666666666667,
+      "grad_norm": 40.5,
+      "learning_rate": 5.137777777777778e-06,
+      "loss": 0.9849,
+      "step": 867
+    },
+    {
+      "epoch": 0.07715555555555556,
+      "grad_norm": 77.0,
+      "learning_rate": 5.143703703703704e-06,
+      "loss": 1.1822,
+      "step": 868
+    },
+    {
+      "epoch": 0.07724444444444445,
+      "grad_norm": 21.375,
+      "learning_rate": 5.14962962962963e-06,
+      "loss": 0.8604,
+      "step": 869
+    },
+    {
+      "epoch": 0.07733333333333334,
+      "grad_norm": 13.8125,
+      "learning_rate": 5.155555555555556e-06,
+      "loss": 1.0739,
+      "step": 870
+    },
+    {
+      "epoch": 0.07742222222222223,
+      "grad_norm": 24.0,
+      "learning_rate": 5.161481481481482e-06,
+      "loss": 1.0772,
+      "step": 871
+    },
+    {
+      "epoch": 0.07751111111111111,
+      "grad_norm": 18.875,
+      "learning_rate": 5.167407407407408e-06,
+      "loss": 0.9472,
+      "step": 872
+    },
+    {
+      "epoch": 0.0776,
+      "grad_norm": 17.0,
+      "learning_rate": 5.1733333333333335e-06,
+      "loss": 0.9283,
+      "step": 873
+    },
+    {
+      "epoch": 0.07768888888888889,
+      "grad_norm": 15.5625,
+      "learning_rate": 5.179259259259259e-06,
+      "loss": 1.0509,
+      "step": 874
+    },
+    {
+      "epoch": 0.07777777777777778,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 1.0652,
+      "step": 875
+    },
+    {
+      "epoch": 0.07786666666666667,
+      "grad_norm": 13.0625,
+      "learning_rate": 5.1911111111111116e-06,
+      "loss": 1.0946,
+      "step": 876
+    },
+    {
+      "epoch": 0.07795555555555556,
+      "grad_norm": 11.0625,
+      "learning_rate": 5.197037037037037e-06,
+      "loss": 0.8748,
+      "step": 877
+    },
+    {
+      "epoch": 0.07804444444444444,
+      "grad_norm": 18.75,
+      "learning_rate": 5.202962962962963e-06,
+      "loss": 1.0584,
+      "step": 878
+    },
+    {
+      "epoch": 0.07813333333333333,
+      "grad_norm": 12.125,
+      "learning_rate": 5.208888888888889e-06,
+      "loss": 1.0502,
+      "step": 879
+    },
+    {
+      "epoch": 0.07822222222222222,
+      "grad_norm": 20.125,
+      "learning_rate": 5.2148148148148145e-06,
+      "loss": 1.055,
+      "step": 880
+    },
+    {
+      "epoch": 0.07831111111111111,
+      "grad_norm": 17.0,
+      "learning_rate": 5.22074074074074e-06,
+      "loss": 1.0016,
+      "step": 881
+    },
+    {
+      "epoch": 0.0784,
+      "grad_norm": 23.75,
+      "learning_rate": 5.226666666666667e-06,
+      "loss": 0.9828,
+      "step": 882
+    },
+    {
+      "epoch": 0.07848888888888889,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.2325925925925935e-06,
+      "loss": 1.0704,
+      "step": 883
+    },
+    {
+      "epoch": 0.07857777777777777,
+      "grad_norm": 15.25,
+      "learning_rate": 5.238518518518519e-06,
+      "loss": 0.9186,
+      "step": 884
+    },
+    {
+      "epoch": 0.07866666666666666,
+      "grad_norm": 9.5,
+      "learning_rate": 5.244444444444445e-06,
+      "loss": 1.0712,
+      "step": 885
+    },
+    {
+      "epoch": 0.07875555555555555,
+      "grad_norm": 14.875,
+      "learning_rate": 5.2503703703703716e-06,
+      "loss": 1.1987,
+      "step": 886
+    },
+    {
+      "epoch": 0.07884444444444444,
+      "grad_norm": 21.25,
+      "learning_rate": 5.256296296296297e-06,
+      "loss": 1.1212,
+      "step": 887
+    },
+    {
+      "epoch": 0.07893333333333333,
+      "grad_norm": 13.0,
+      "learning_rate": 5.262222222222223e-06,
+      "loss": 0.9041,
+      "step": 888
+    },
+    {
+      "epoch": 0.07902222222222222,
+      "grad_norm": 25.0,
+      "learning_rate": 5.268148148148149e-06,
+      "loss": 0.9844,
+      "step": 889
+    },
+    {
+      "epoch": 0.0791111111111111,
+      "grad_norm": 15.0,
+      "learning_rate": 5.2740740740740745e-06,
+      "loss": 1.0423,
+      "step": 890
+    },
+    {
+      "epoch": 0.0792,
+      "grad_norm": 14.0625,
+      "learning_rate": 5.28e-06,
+      "loss": 1.0796,
+      "step": 891
+    },
+    {
+      "epoch": 0.0792888888888889,
+      "grad_norm": 8.625,
+      "learning_rate": 5.285925925925927e-06,
+      "loss": 1.0597,
+      "step": 892
+    },
+    {
+      "epoch": 0.07937777777777778,
+      "grad_norm": 12.0,
+      "learning_rate": 5.291851851851853e-06,
+      "loss": 0.9701,
+      "step": 893
+    },
+    {
+      "epoch": 0.07946666666666667,
+      "grad_norm": 9.75,
+      "learning_rate": 5.297777777777778e-06,
+      "loss": 0.9341,
+      "step": 894
+    },
+    {
+      "epoch": 0.07955555555555556,
+      "grad_norm": 5.21875,
+      "learning_rate": 5.303703703703704e-06,
+      "loss": 1.038,
+      "step": 895
+    },
+    {
+      "epoch": 0.07964444444444445,
+      "grad_norm": 15.4375,
+      "learning_rate": 5.30962962962963e-06,
+      "loss": 1.1918,
+      "step": 896
+    },
+    {
+      "epoch": 0.07973333333333334,
+      "grad_norm": 30.25,
+      "learning_rate": 5.3155555555555564e-06,
+      "loss": 0.9809,
+      "step": 897
+    },
+    {
+      "epoch": 0.07982222222222222,
+      "grad_norm": 15.6875,
+      "learning_rate": 5.321481481481482e-06,
+      "loss": 0.9291,
+      "step": 898
+    },
+    {
+      "epoch": 0.07991111111111111,
+      "grad_norm": 13.0625,
+      "learning_rate": 5.327407407407408e-06,
+      "loss": 0.9623,
+      "step": 899
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.5,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.0123,
+      "step": 900
+    },
+    {
+      "epoch": 0.08008888888888889,
+      "grad_norm": 15.625,
+      "learning_rate": 5.339259259259259e-06,
+      "loss": 0.9463,
+      "step": 901
+    },
+    {
+      "epoch": 0.08017777777777778,
+      "grad_norm": 22.25,
+      "learning_rate": 5.345185185185185e-06,
+      "loss": 0.9269,
+      "step": 902
+    },
+    {
+      "epoch": 0.08026666666666667,
+      "grad_norm": 23.375,
+      "learning_rate": 5.351111111111112e-06,
+      "loss": 1.0053,
+      "step": 903
+    },
+    {
+      "epoch": 0.08035555555555556,
+      "grad_norm": 17.5,
+      "learning_rate": 5.3570370370370375e-06,
+      "loss": 0.9401,
+      "step": 904
+    },
+    {
+      "epoch": 0.08044444444444444,
+      "grad_norm": 16.875,
+      "learning_rate": 5.362962962962963e-06,
+      "loss": 1.0765,
+      "step": 905
+    },
+    {
+      "epoch": 0.08053333333333333,
+      "grad_norm": 25.125,
+      "learning_rate": 5.368888888888889e-06,
+      "loss": 0.9657,
+      "step": 906
+    },
+    {
+      "epoch": 0.08062222222222222,
+      "grad_norm": 17.25,
+      "learning_rate": 5.374814814814815e-06,
+      "loss": 1.0011,
+      "step": 907
+    },
+    {
+      "epoch": 0.08071111111111111,
+      "grad_norm": 19.0,
+      "learning_rate": 5.3807407407407405e-06,
+      "loss": 0.9674,
+      "step": 908
+    },
+    {
+      "epoch": 0.0808,
+      "grad_norm": 13.5625,
+      "learning_rate": 5.386666666666667e-06,
+      "loss": 1.1011,
+      "step": 909
+    },
+    {
+      "epoch": 0.08088888888888889,
+      "grad_norm": 48.0,
+      "learning_rate": 5.392592592592593e-06,
+      "loss": 0.7799,
+      "step": 910
+    },
+    {
+      "epoch": 0.08097777777777777,
+      "grad_norm": 40.0,
+      "learning_rate": 5.3985185185185185e-06,
+      "loss": 0.8798,
+      "step": 911
+    },
+    {
+      "epoch": 0.08106666666666666,
+      "grad_norm": 14.75,
+      "learning_rate": 5.404444444444444e-06,
+      "loss": 0.9889,
+      "step": 912
+    },
+    {
+      "epoch": 0.08115555555555555,
+      "grad_norm": 8.625,
+      "learning_rate": 5.41037037037037e-06,
+      "loss": 0.9277,
+      "step": 913
+    },
+    {
+      "epoch": 0.08124444444444444,
+      "grad_norm": 28.625,
+      "learning_rate": 5.4162962962962975e-06,
+      "loss": 1.12,
+      "step": 914
+    },
+    {
+      "epoch": 0.08133333333333333,
+      "grad_norm": 51.5,
+      "learning_rate": 5.422222222222223e-06,
+      "loss": 0.9853,
+      "step": 915
+    },
+    {
+      "epoch": 0.08142222222222223,
+      "grad_norm": 37.25,
+      "learning_rate": 5.428148148148149e-06,
+      "loss": 0.9175,
+      "step": 916
+    },
+    {
+      "epoch": 0.08151111111111112,
+      "grad_norm": 18.5,
+      "learning_rate": 5.434074074074075e-06,
+      "loss": 1.0471,
+      "step": 917
+    },
+    {
+      "epoch": 0.0816,
+      "grad_norm": 13.25,
+      "learning_rate": 5.4400000000000004e-06,
+      "loss": 0.9901,
+      "step": 918
+    },
+    {
+      "epoch": 0.0816888888888889,
+      "grad_norm": 6.53125,
+      "learning_rate": 5.445925925925927e-06,
+      "loss": 1.0058,
+      "step": 919
+    },
+    {
+      "epoch": 0.08177777777777778,
+      "grad_norm": 19.25,
+      "learning_rate": 5.451851851851853e-06,
+      "loss": 0.8776,
+      "step": 920
+    },
+    {
+      "epoch": 0.08186666666666667,
+      "grad_norm": 14.8125,
+      "learning_rate": 5.4577777777777785e-06,
+      "loss": 0.8998,
+      "step": 921
+    },
+    {
+      "epoch": 0.08195555555555556,
+      "grad_norm": 12.0,
+      "learning_rate": 5.463703703703704e-06,
+      "loss": 0.922,
+      "step": 922
+    },
+    {
+      "epoch": 0.08204444444444445,
+      "grad_norm": 7.96875,
+      "learning_rate": 5.46962962962963e-06,
+      "loss": 1.0777,
+      "step": 923
+    },
+    {
+      "epoch": 0.08213333333333334,
+      "grad_norm": 24.125,
+      "learning_rate": 5.475555555555557e-06,
+      "loss": 0.8317,
+      "step": 924
+    },
+    {
+      "epoch": 0.08222222222222222,
+      "grad_norm": 19.125,
+      "learning_rate": 5.481481481481482e-06,
+      "loss": 1.029,
+      "step": 925
+    },
+    {
+      "epoch": 0.08231111111111111,
+      "grad_norm": 50.0,
+      "learning_rate": 5.487407407407408e-06,
+      "loss": 1.1379,
+      "step": 926
+    },
+    {
+      "epoch": 0.0824,
+      "grad_norm": 60.75,
+      "learning_rate": 5.493333333333334e-06,
+      "loss": 0.9609,
+      "step": 927
+    },
+    {
+      "epoch": 0.08248888888888889,
+      "grad_norm": 49.75,
+      "learning_rate": 5.4992592592592596e-06,
+      "loss": 0.8845,
+      "step": 928
+    },
+    {
+      "epoch": 0.08257777777777778,
+      "grad_norm": 28.5,
+      "learning_rate": 5.505185185185185e-06,
+      "loss": 0.9022,
+      "step": 929
+    },
+    {
+      "epoch": 0.08266666666666667,
+      "grad_norm": 10.875,
+      "learning_rate": 5.511111111111112e-06,
+      "loss": 1.1057,
+      "step": 930
+    },
+    {
+      "epoch": 0.08275555555555555,
+      "grad_norm": 23.0,
+      "learning_rate": 5.517037037037038e-06,
+      "loss": 0.7815,
+      "step": 931
+    },
+    {
+      "epoch": 0.08284444444444444,
+      "grad_norm": 5.40625,
+      "learning_rate": 5.522962962962963e-06,
+      "loss": 0.9139,
+      "step": 932
+    },
+    {
+      "epoch": 0.08293333333333333,
+      "grad_norm": 9.1875,
+      "learning_rate": 5.528888888888889e-06,
+      "loss": 1.0064,
+      "step": 933
+    },
+    {
+      "epoch": 0.08302222222222222,
+      "grad_norm": 18.75,
+      "learning_rate": 5.534814814814815e-06,
+      "loss": 1.0324,
+      "step": 934
+    },
+    {
+      "epoch": 0.08311111111111111,
+      "grad_norm": 13.375,
+      "learning_rate": 5.540740740740741e-06,
+      "loss": 1.0154,
+      "step": 935
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 13.5,
+      "learning_rate": 5.546666666666667e-06,
+      "loss": 1.0638,
+      "step": 936
+    },
+    {
+      "epoch": 0.08328888888888888,
+      "grad_norm": 26.25,
+      "learning_rate": 5.552592592592593e-06,
+      "loss": 0.9699,
+      "step": 937
+    },
+    {
+      "epoch": 0.08337777777777777,
+      "grad_norm": 37.0,
+      "learning_rate": 5.558518518518519e-06,
+      "loss": 1.026,
+      "step": 938
+    },
+    {
+      "epoch": 0.08346666666666666,
+      "grad_norm": 16.0,
+      "learning_rate": 5.5644444444444444e-06,
+      "loss": 0.9056,
+      "step": 939
+    },
+    {
+      "epoch": 0.08355555555555555,
+      "grad_norm": 22.875,
+      "learning_rate": 5.57037037037037e-06,
+      "loss": 0.8375,
+      "step": 940
+    },
+    {
+      "epoch": 0.08364444444444444,
+      "grad_norm": 25.625,
+      "learning_rate": 5.576296296296296e-06,
+      "loss": 0.9627,
+      "step": 941
+    },
+    {
+      "epoch": 0.08373333333333334,
+      "grad_norm": 12.125,
+      "learning_rate": 5.5822222222222225e-06,
+      "loss": 0.838,
+      "step": 942
+    },
+    {
+      "epoch": 0.08382222222222223,
+      "grad_norm": 29.375,
+      "learning_rate": 5.588148148148148e-06,
+      "loss": 1.169,
+      "step": 943
+    },
+    {
+      "epoch": 0.08391111111111112,
+      "grad_norm": 32.25,
+      "learning_rate": 5.594074074074074e-06,
+      "loss": 1.1005,
+      "step": 944
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 27.25,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.9303,
+      "step": 945
+    },
+    {
+      "epoch": 0.0840888888888889,
+      "grad_norm": 12.6875,
+      "learning_rate": 5.605925925925927e-06,
+      "loss": 0.9645,
+      "step": 946
+    },
+    {
+      "epoch": 0.08417777777777778,
+      "grad_norm": 17.875,
+      "learning_rate": 5.611851851851853e-06,
+      "loss": 0.9513,
+      "step": 947
+    },
+    {
+      "epoch": 0.08426666666666667,
+      "grad_norm": 35.5,
+      "learning_rate": 5.617777777777779e-06,
+      "loss": 0.8811,
+      "step": 948
+    },
+    {
+      "epoch": 0.08435555555555556,
+      "grad_norm": 18.625,
+      "learning_rate": 5.623703703703704e-06,
+      "loss": 0.9656,
+      "step": 949
+    },
+    {
+      "epoch": 0.08444444444444445,
+      "grad_norm": 9.4375,
+      "learning_rate": 5.62962962962963e-06,
+      "loss": 0.9613,
+      "step": 950
+    },
+    {
+      "epoch": 0.08453333333333334,
+      "grad_norm": 14.0625,
+      "learning_rate": 5.635555555555557e-06,
+      "loss": 0.9388,
+      "step": 951
+    },
+    {
+      "epoch": 0.08462222222222222,
+      "grad_norm": 21.375,
+      "learning_rate": 5.6414814814814825e-06,
+      "loss": 1.0081,
+      "step": 952
+    },
+    {
+      "epoch": 0.08471111111111111,
+      "grad_norm": 6.5625,
+      "learning_rate": 5.647407407407408e-06,
+      "loss": 0.8823,
+      "step": 953
+    },
+    {
+      "epoch": 0.0848,
+      "grad_norm": 8.125,
+      "learning_rate": 5.653333333333334e-06,
+      "loss": 0.9675,
+      "step": 954
+    },
+    {
+      "epoch": 0.08488888888888889,
+      "grad_norm": 9.6875,
+      "learning_rate": 5.65925925925926e-06,
+      "loss": 0.908,
+      "step": 955
+    },
+    {
+      "epoch": 0.08497777777777778,
+      "grad_norm": 41.0,
+      "learning_rate": 5.6651851851851855e-06,
+      "loss": 1.007,
+      "step": 956
+    },
+    {
+      "epoch": 0.08506666666666667,
+      "grad_norm": 7.46875,
+      "learning_rate": 5.671111111111112e-06,
+      "loss": 1.0873,
+      "step": 957
+    },
+    {
+      "epoch": 0.08515555555555555,
+      "grad_norm": 10.9375,
+      "learning_rate": 5.677037037037038e-06,
+      "loss": 1.07,
+      "step": 958
+    },
+    {
+      "epoch": 0.08524444444444444,
+      "grad_norm": 36.5,
+      "learning_rate": 5.6829629629629635e-06,
+      "loss": 0.9143,
+      "step": 959
+    },
+    {
+      "epoch": 0.08533333333333333,
+      "grad_norm": 16.0,
+      "learning_rate": 5.688888888888889e-06,
+      "loss": 0.9228,
+      "step": 960
+    },
+    {
+      "epoch": 0.08542222222222222,
+      "grad_norm": 12.75,
+      "learning_rate": 5.694814814814815e-06,
+      "loss": 0.9985,
+      "step": 961
+    },
+    {
+      "epoch": 0.08551111111111111,
+      "grad_norm": 6.3125,
+      "learning_rate": 5.700740740740741e-06,
+      "loss": 0.9101,
+      "step": 962
+    },
+    {
+      "epoch": 0.0856,
+      "grad_norm": 48.75,
+      "learning_rate": 5.706666666666667e-06,
+      "loss": 0.9106,
+      "step": 963
+    },
+    {
+      "epoch": 0.08568888888888888,
+      "grad_norm": 37.5,
+      "learning_rate": 5.712592592592593e-06,
+      "loss": 0.9535,
+      "step": 964
+    },
+    {
+      "epoch": 0.08577777777777777,
+      "grad_norm": 21.75,
+      "learning_rate": 5.718518518518519e-06,
+      "loss": 0.8622,
+      "step": 965
+    },
+    {
+      "epoch": 0.08586666666666666,
+      "grad_norm": 12.375,
+      "learning_rate": 5.724444444444445e-06,
+      "loss": 0.9044,
+      "step": 966
+    },
+    {
+      "epoch": 0.08595555555555555,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.73037037037037e-06,
+      "loss": 0.857,
+      "step": 967
+    },
+    {
+      "epoch": 0.08604444444444445,
+      "grad_norm": 9.0625,
+      "learning_rate": 5.736296296296296e-06,
+      "loss": 1.2415,
+      "step": 968
+    },
+    {
+      "epoch": 0.08613333333333334,
+      "grad_norm": 25.125,
+      "learning_rate": 5.742222222222223e-06,
+      "loss": 1.0327,
+      "step": 969
+    },
+    {
+      "epoch": 0.08622222222222223,
+      "grad_norm": 17.375,
+      "learning_rate": 5.748148148148148e-06,
+      "loss": 0.9202,
+      "step": 970
+    },
+    {
+      "epoch": 0.08631111111111112,
+      "grad_norm": 21.0,
+      "learning_rate": 5.754074074074074e-06,
+      "loss": 0.9149,
+      "step": 971
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 6.0,
+      "learning_rate": 5.76e-06,
+      "loss": 0.8889,
+      "step": 972
+    },
+    {
+      "epoch": 0.08648888888888889,
+      "grad_norm": 8.6875,
+      "learning_rate": 5.765925925925926e-06,
+      "loss": 1.0631,
+      "step": 973
+    },
+    {
+      "epoch": 0.08657777777777778,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.771851851851851e-06,
+      "loss": 1.0226,
+      "step": 974
+    },
+    {
+      "epoch": 0.08666666666666667,
+      "grad_norm": 11.75,
+      "learning_rate": 5.777777777777778e-06,
+      "loss": 0.8716,
+      "step": 975
+    },
+    {
+      "epoch": 0.08675555555555556,
+      "grad_norm": 14.5,
+      "learning_rate": 5.783703703703705e-06,
+      "loss": 0.834,
+      "step": 976
+    },
+    {
+      "epoch": 0.08684444444444445,
+      "grad_norm": 17.0,
+      "learning_rate": 5.78962962962963e-06,
+      "loss": 0.9354,
+      "step": 977
+    },
+    {
+      "epoch": 0.08693333333333333,
+      "grad_norm": 17.375,
+      "learning_rate": 5.795555555555557e-06,
+      "loss": 0.8918,
+      "step": 978
+    },
+    {
+      "epoch": 0.08702222222222222,
+      "grad_norm": 13.625,
+      "learning_rate": 5.801481481481483e-06,
+      "loss": 1.0007,
+      "step": 979
+    },
+    {
+      "epoch": 0.08711111111111111,
+      "grad_norm": 11.1875,
+      "learning_rate": 5.807407407407408e-06,
+      "loss": 1.039,
+      "step": 980
+    },
+    {
+      "epoch": 0.0872,
+      "grad_norm": 21.5,
+      "learning_rate": 5.813333333333334e-06,
+      "loss": 0.8528,
+      "step": 981
+    },
+    {
+      "epoch": 0.08728888888888889,
+      "grad_norm": 22.625,
+      "learning_rate": 5.81925925925926e-06,
+      "loss": 0.9071,
+      "step": 982
+    },
+    {
+      "epoch": 0.08737777777777778,
+      "grad_norm": 12.5,
+      "learning_rate": 5.825185185185186e-06,
+      "loss": 0.9539,
+      "step": 983
+    },
+    {
+      "epoch": 0.08746666666666666,
+      "grad_norm": 28.5,
+      "learning_rate": 5.831111111111112e-06,
+      "loss": 0.9258,
+      "step": 984
+    },
+    {
+      "epoch": 0.08755555555555555,
+      "grad_norm": 9.25,
+      "learning_rate": 5.837037037037038e-06,
+      "loss": 0.9701,
+      "step": 985
+    },
+    {
+      "epoch": 0.08764444444444444,
+      "grad_norm": 11.6875,
+      "learning_rate": 5.842962962962964e-06,
+      "loss": 0.9626,
+      "step": 986
+    },
+    {
+      "epoch": 0.08773333333333333,
+      "grad_norm": 27.375,
+      "learning_rate": 5.8488888888888895e-06,
+      "loss": 0.9498,
+      "step": 987
+    },
+    {
+      "epoch": 0.08782222222222222,
+      "grad_norm": 15.5,
+      "learning_rate": 5.854814814814815e-06,
+      "loss": 1.1438,
+      "step": 988
+    },
+    {
+      "epoch": 0.0879111111111111,
+      "grad_norm": 19.5,
+      "learning_rate": 5.860740740740741e-06,
+      "loss": 0.9251,
+      "step": 989
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 9.9375,
+      "learning_rate": 5.8666666666666675e-06,
+      "loss": 0.9546,
+      "step": 990
+    },
+    {
+      "epoch": 0.08808888888888888,
+      "grad_norm": 18.25,
+      "learning_rate": 5.872592592592593e-06,
+      "loss": 1.0136,
+      "step": 991
+    },
+    {
+      "epoch": 0.08817777777777777,
+      "grad_norm": 20.25,
+      "learning_rate": 5.878518518518519e-06,
+      "loss": 1.1301,
+      "step": 992
+    },
+    {
+      "epoch": 0.08826666666666666,
+      "grad_norm": 19.25,
+      "learning_rate": 5.884444444444445e-06,
+      "loss": 1.1384,
+      "step": 993
+    },
+    {
+      "epoch": 0.08835555555555556,
+      "grad_norm": 12.8125,
+      "learning_rate": 5.8903703703703705e-06,
+      "loss": 0.9239,
+      "step": 994
+    },
+    {
+      "epoch": 0.08844444444444445,
+      "grad_norm": 33.75,
+      "learning_rate": 5.896296296296296e-06,
+      "loss": 1.0624,
+      "step": 995
+    },
+    {
+      "epoch": 0.08853333333333334,
+      "grad_norm": 35.25,
+      "learning_rate": 5.902222222222223e-06,
+      "loss": 0.7896,
+      "step": 996
+    },
+    {
+      "epoch": 0.08862222222222223,
+      "grad_norm": 22.75,
+      "learning_rate": 5.908148148148149e-06,
+      "loss": 0.93,
+      "step": 997
+    },
+    {
+      "epoch": 0.08871111111111112,
+      "grad_norm": 9.0625,
+      "learning_rate": 5.914074074074074e-06,
+      "loss": 0.93,
+      "step": 998
+    },
+    {
+      "epoch": 0.0888,
+      "grad_norm": 13.0,
+      "learning_rate": 5.92e-06,
+      "loss": 1.0259,
+      "step": 999
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 25.375,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.8832,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08897777777777778,
+      "grad_norm": 25.625,
+      "learning_rate": 5.9318518518518516e-06,
+      "loss": 0.814,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08906666666666667,
+      "grad_norm": 14.375,
+      "learning_rate": 5.937777777777778e-06,
+      "loss": 0.9204,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08915555555555556,
+      "grad_norm": 9.8125,
+      "learning_rate": 5.943703703703704e-06,
+      "loss": 1.1244,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08924444444444445,
+      "grad_norm": 14.625,
+      "learning_rate": 5.94962962962963e-06,
+      "loss": 0.8929,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08933333333333333,
+      "grad_norm": 20.5,
+      "learning_rate": 5.955555555555555e-06,
+      "loss": 0.9888,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08942222222222222,
+      "grad_norm": 21.625,
+      "learning_rate": 5.961481481481481e-06,
+      "loss": 0.8717,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08951111111111111,
+      "grad_norm": 32.5,
+      "learning_rate": 5.9674074074074086e-06,
+      "loss": 0.91,
+      "step": 1007
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 21.875,
+      "learning_rate": 5.973333333333334e-06,
+      "loss": 0.8979,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08968888888888889,
+      "grad_norm": 21.25,
+      "learning_rate": 5.97925925925926e-06,
+      "loss": 0.9577,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08977777777777778,
+      "grad_norm": 10.25,
+      "learning_rate": 5.985185185185186e-06,
+      "loss": 0.9821,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08986666666666666,
+      "grad_norm": 33.5,
+      "learning_rate": 5.991111111111112e-06,
+      "loss": 0.9456,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08995555555555555,
+      "grad_norm": 23.625,
+      "learning_rate": 5.997037037037038e-06,
+      "loss": 0.9413,
+      "step": 1012
+    },
+    {
+      "epoch": 0.09004444444444444,
+      "grad_norm": 17.375,
+      "learning_rate": 6.002962962962964e-06,
+      "loss": 0.9847,
+      "step": 1013
+    },
+    {
+      "epoch": 0.09013333333333333,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.00888888888889e-06,
+      "loss": 0.9588,
+      "step": 1014
+    },
+    {
+      "epoch": 0.09022222222222222,
+      "grad_norm": 15.875,
+      "learning_rate": 6.014814814814815e-06,
+      "loss": 0.8533,
+      "step": 1015
+    },
+    {
+      "epoch": 0.0903111111111111,
+      "grad_norm": 15.0,
+      "learning_rate": 6.020740740740741e-06,
+      "loss": 0.8496,
+      "step": 1016
+    },
+    {
+      "epoch": 0.0904,
+      "grad_norm": 10.4375,
+      "learning_rate": 6.026666666666668e-06,
+      "loss": 1.0267,
+      "step": 1017
+    },
+    {
+      "epoch": 0.09048888888888888,
+      "grad_norm": 13.8125,
+      "learning_rate": 6.0325925925925934e-06,
+      "loss": 0.8934,
+      "step": 1018
+    },
+    {
+      "epoch": 0.09057777777777777,
+      "grad_norm": 15.5,
+      "learning_rate": 6.038518518518519e-06,
+      "loss": 1.0484,
+      "step": 1019
+    },
+    {
+      "epoch": 0.09066666666666667,
+      "grad_norm": 24.5,
+      "learning_rate": 6.044444444444445e-06,
+      "loss": 0.8808,
+      "step": 1020
+    },
+    {
+      "epoch": 0.09075555555555556,
+      "grad_norm": 6.6875,
+      "learning_rate": 6.050370370370371e-06,
+      "loss": 0.9334,
+      "step": 1021
+    },
+    {
+      "epoch": 0.09084444444444445,
+      "grad_norm": 10.8125,
+      "learning_rate": 6.056296296296296e-06,
+      "loss": 0.8819,
+      "step": 1022
+    },
+    {
+      "epoch": 0.09093333333333334,
+      "grad_norm": 9.125,
+      "learning_rate": 6.062222222222223e-06,
+      "loss": 1.0432,
+      "step": 1023
+    },
+    {
+      "epoch": 0.09102222222222223,
+      "grad_norm": 8.5,
+      "learning_rate": 6.068148148148149e-06,
+      "loss": 0.8768,
+      "step": 1024
+    },
+    {
+      "epoch": 0.09111111111111111,
+      "grad_norm": 19.625,
+      "learning_rate": 6.0740740740740745e-06,
+      "loss": 0.9464,
+      "step": 1025
+    },
+    {
+      "epoch": 0.0912,
+      "grad_norm": 16.875,
+      "learning_rate": 6.08e-06,
+      "loss": 0.9773,
+      "step": 1026
+    },
+    {
+      "epoch": 0.09128888888888889,
+      "grad_norm": 14.75,
+      "learning_rate": 6.085925925925926e-06,
+      "loss": 0.8902,
+      "step": 1027
+    },
+    {
+      "epoch": 0.09137777777777778,
+      "grad_norm": 14.375,
+      "learning_rate": 6.091851851851852e-06,
+      "loss": 0.8664,
+      "step": 1028
+    },
+    {
+      "epoch": 0.09146666666666667,
+      "grad_norm": 10.0,
+      "learning_rate": 6.097777777777778e-06,
+      "loss": 1.064,
+      "step": 1029
+    },
+    {
+      "epoch": 0.09155555555555556,
+      "grad_norm": 14.4375,
+      "learning_rate": 6.103703703703704e-06,
+      "loss": 1.1152,
+      "step": 1030
+    },
+    {
+      "epoch": 0.09164444444444445,
+      "grad_norm": 12.875,
+      "learning_rate": 6.10962962962963e-06,
+      "loss": 0.787,
+      "step": 1031
+    },
+    {
+      "epoch": 0.09173333333333333,
+      "grad_norm": 8.0,
+      "learning_rate": 6.1155555555555555e-06,
+      "loss": 0.9247,
+      "step": 1032
+    },
+    {
+      "epoch": 0.09182222222222222,
+      "grad_norm": 13.625,
+      "learning_rate": 6.121481481481481e-06,
+      "loss": 0.8224,
+      "step": 1033
+    },
+    {
+      "epoch": 0.09191111111111111,
+      "grad_norm": 25.875,
+      "learning_rate": 6.127407407407408e-06,
+      "loss": 0.972,
+      "step": 1034
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 19.625,
+      "learning_rate": 6.133333333333334e-06,
+      "loss": 1.0319,
+      "step": 1035
+    },
+    {
+      "epoch": 0.09208888888888889,
+      "grad_norm": 5.96875,
+      "learning_rate": 6.139259259259259e-06,
+      "loss": 1.0125,
+      "step": 1036
+    },
+    {
+      "epoch": 0.09217777777777778,
+      "grad_norm": 11.375,
+      "learning_rate": 6.145185185185185e-06,
+      "loss": 0.8842,
+      "step": 1037
+    },
+    {
+      "epoch": 0.09226666666666666,
+      "grad_norm": 10.625,
+      "learning_rate": 6.1511111111111125e-06,
+      "loss": 0.84,
+      "step": 1038
+    },
+    {
+      "epoch": 0.09235555555555555,
+      "grad_norm": 36.75,
+      "learning_rate": 6.157037037037038e-06,
+      "loss": 0.9803,
+      "step": 1039
+    },
+    {
+      "epoch": 0.09244444444444444,
+      "grad_norm": 39.75,
+      "learning_rate": 6.162962962962964e-06,
+      "loss": 0.7993,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09253333333333333,
+      "grad_norm": 10.3125,
+      "learning_rate": 6.16888888888889e-06,
+      "loss": 0.7938,
+      "step": 1041
+    },
+    {
+      "epoch": 0.09262222222222222,
+      "grad_norm": 12.25,
+      "learning_rate": 6.1748148148148155e-06,
+      "loss": 1.0007,
+      "step": 1042
+    },
+    {
+      "epoch": 0.0927111111111111,
+      "grad_norm": 9.1875,
+      "learning_rate": 6.180740740740741e-06,
+      "loss": 0.8314,
+      "step": 1043
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.186666666666668e-06,
+      "loss": 1.0201,
+      "step": 1044
+    },
+    {
+      "epoch": 0.09288888888888888,
+      "grad_norm": 22.75,
+      "learning_rate": 6.192592592592594e-06,
+      "loss": 0.8844,
+      "step": 1045
+    },
+    {
+      "epoch": 0.09297777777777778,
+      "grad_norm": 35.75,
+      "learning_rate": 6.198518518518519e-06,
+      "loss": 1.0508,
+      "step": 1046
+    },
+    {
+      "epoch": 0.09306666666666667,
+      "grad_norm": 14.125,
+      "learning_rate": 6.204444444444445e-06,
+      "loss": 0.9483,
+      "step": 1047
+    },
+    {
+      "epoch": 0.09315555555555556,
+      "grad_norm": 21.625,
+      "learning_rate": 6.210370370370371e-06,
+      "loss": 0.8275,
+      "step": 1048
+    },
+    {
+      "epoch": 0.09324444444444445,
+      "grad_norm": 12.25,
+      "learning_rate": 6.2162962962962966e-06,
+      "loss": 0.7298,
+      "step": 1049
+    },
+    {
+      "epoch": 0.09333333333333334,
+      "grad_norm": 14.0,
+      "learning_rate": 6.222222222222223e-06,
+      "loss": 0.8171,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09342222222222223,
+      "grad_norm": 10.6875,
+      "learning_rate": 6.228148148148149e-06,
+      "loss": 0.9722,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09351111111111111,
+      "grad_norm": 7.375,
+      "learning_rate": 6.234074074074075e-06,
+      "loss": 0.7929,
+      "step": 1052
+    },
+    {
+      "epoch": 0.0936,
+      "grad_norm": 13.5625,
+      "learning_rate": 6.24e-06,
+      "loss": 0.8834,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09368888888888889,
+      "grad_norm": 49.0,
+      "learning_rate": 6.245925925925926e-06,
+      "loss": 0.7753,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09377777777777778,
+      "grad_norm": 10.4375,
+      "learning_rate": 6.251851851851852e-06,
+      "loss": 0.9281,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09386666666666667,
+      "grad_norm": 9.8125,
+      "learning_rate": 6.2577777777777785e-06,
+      "loss": 0.9865,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09395555555555556,
+      "grad_norm": 37.75,
+      "learning_rate": 6.263703703703704e-06,
+      "loss": 1.0223,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09404444444444444,
+      "grad_norm": 47.75,
+      "learning_rate": 6.26962962962963e-06,
+      "loss": 0.9997,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09413333333333333,
+      "grad_norm": 44.75,
+      "learning_rate": 6.275555555555556e-06,
+      "loss": 0.8167,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09422222222222222,
+      "grad_norm": 37.75,
+      "learning_rate": 6.2814814814814814e-06,
+      "loss": 0.9412,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09431111111111111,
+      "grad_norm": 19.875,
+      "learning_rate": 6.287407407407408e-06,
+      "loss": 1.0371,
+      "step": 1061
+    },
+    {
+      "epoch": 0.0944,
+      "grad_norm": 36.25,
+      "learning_rate": 6.293333333333334e-06,
+      "loss": 0.7965,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09448888888888889,
+      "grad_norm": 11.125,
+      "learning_rate": 6.2992592592592595e-06,
+      "loss": 1.0447,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09457777777777777,
+      "grad_norm": 15.3125,
+      "learning_rate": 6.305185185185185e-06,
+      "loss": 1.0944,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09466666666666666,
+      "grad_norm": 11.3125,
+      "learning_rate": 6.311111111111111e-06,
+      "loss": 0.9053,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09475555555555555,
+      "grad_norm": 13.0625,
+      "learning_rate": 6.317037037037037e-06,
+      "loss": 1.0023,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09484444444444444,
+      "grad_norm": 10.125,
+      "learning_rate": 6.322962962962963e-06,
+      "loss": 1.0163,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09493333333333333,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.328888888888889e-06,
+      "loss": 1.0686,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09502222222222222,
+      "grad_norm": 9.3125,
+      "learning_rate": 6.334814814814816e-06,
+      "loss": 0.919,
+      "step": 1069
+    },
+    {
+      "epoch": 0.0951111111111111,
+      "grad_norm": 16.5,
+      "learning_rate": 6.3407407407407414e-06,
+      "loss": 0.8819,
+      "step": 1070
+    },
+    {
+      "epoch": 0.0952,
+      "grad_norm": 24.375,
+      "learning_rate": 6.346666666666668e-06,
+      "loss": 0.9231,
+      "step": 1071
+    },
+    {
+      "epoch": 0.0952888888888889,
+      "grad_norm": 8.25,
+      "learning_rate": 6.352592592592594e-06,
+      "loss": 1.1247,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09537777777777778,
+      "grad_norm": 9.0,
+      "learning_rate": 6.3585185185185195e-06,
+      "loss": 0.8918,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09546666666666667,
+      "grad_norm": 11.0625,
+      "learning_rate": 6.364444444444445e-06,
+      "loss": 0.8689,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09555555555555556,
+      "grad_norm": 9.25,
+      "learning_rate": 6.370370370370371e-06,
+      "loss": 0.7735,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09564444444444445,
+      "grad_norm": 21.25,
+      "learning_rate": 6.376296296296297e-06,
+      "loss": 0.868,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09573333333333334,
+      "grad_norm": 17.0,
+      "learning_rate": 6.382222222222223e-06,
+      "loss": 0.8701,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09582222222222223,
+      "grad_norm": 27.25,
+      "learning_rate": 6.388148148148149e-06,
+      "loss": 0.878,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09591111111111111,
+      "grad_norm": 31.75,
+      "learning_rate": 6.394074074074075e-06,
+      "loss": 0.8711,
+      "step": 1079
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 9.75,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.9271,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09608888888888889,
+      "grad_norm": 13.625,
+      "learning_rate": 6.405925925925926e-06,
+      "loss": 0.9755,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09617777777777778,
+      "grad_norm": 12.625,
+      "learning_rate": 6.411851851851852e-06,
+      "loss": 0.8105,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09626666666666667,
+      "grad_norm": 16.625,
+      "learning_rate": 6.417777777777779e-06,
+      "loss": 0.8298,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09635555555555556,
+      "grad_norm": 13.0625,
+      "learning_rate": 6.423703703703704e-06,
+      "loss": 0.7389,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09644444444444444,
+      "grad_norm": 15.25,
+      "learning_rate": 6.42962962962963e-06,
+      "loss": 0.7648,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09653333333333333,
+      "grad_norm": 17.0,
+      "learning_rate": 6.435555555555556e-06,
+      "loss": 0.9176,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09662222222222222,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.441481481481482e-06,
+      "loss": 0.876,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09671111111111111,
+      "grad_norm": 13.9375,
+      "learning_rate": 6.447407407407408e-06,
+      "loss": 0.9149,
+      "step": 1088
+    },
+    {
+      "epoch": 0.0968,
+      "grad_norm": 18.75,
+      "learning_rate": 6.453333333333334e-06,
+      "loss": 0.9759,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09688888888888889,
+      "grad_norm": 12.0,
+      "learning_rate": 6.45925925925926e-06,
+      "loss": 0.8678,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09697777777777777,
+      "grad_norm": 14.125,
+      "learning_rate": 6.4651851851851854e-06,
+      "loss": 0.9106,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09706666666666666,
+      "grad_norm": 19.375,
+      "learning_rate": 6.471111111111111e-06,
+      "loss": 0.814,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09715555555555555,
+      "grad_norm": 20.875,
+      "learning_rate": 6.477037037037037e-06,
+      "loss": 0.9231,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09724444444444444,
+      "grad_norm": 10.3125,
+      "learning_rate": 6.4829629629629635e-06,
+      "loss": 0.9281,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09733333333333333,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.488888888888889e-06,
+      "loss": 1.0715,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09742222222222222,
+      "grad_norm": 6.5625,
+      "learning_rate": 6.494814814814815e-06,
+      "loss": 0.8752,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09751111111111112,
+      "grad_norm": 9.0625,
+      "learning_rate": 6.500740740740741e-06,
+      "loss": 0.8744,
+      "step": 1097
+    },
+    {
+      "epoch": 0.0976,
+      "grad_norm": 16.375,
+      "learning_rate": 6.5066666666666665e-06,
+      "loss": 0.7764,
+      "step": 1098
+    },
+    {
+      "epoch": 0.0976888888888889,
+      "grad_norm": 48.5,
+      "learning_rate": 6.512592592592592e-06,
+      "loss": 0.8565,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09777777777777778,
+      "grad_norm": 43.75,
+      "learning_rate": 6.51851851851852e-06,
+      "loss": 1.022,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09786666666666667,
+      "grad_norm": 54.25,
+      "learning_rate": 6.524444444444445e-06,
+      "loss": 0.9185,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09795555555555556,
+      "grad_norm": 7.03125,
+      "learning_rate": 6.530370370370371e-06,
+      "loss": 0.7907,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09804444444444445,
+      "grad_norm": 15.25,
+      "learning_rate": 6.536296296296297e-06,
+      "loss": 0.7507,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09813333333333334,
+      "grad_norm": 8.375,
+      "learning_rate": 6.5422222222222235e-06,
+      "loss": 0.7937,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09822222222222222,
+      "grad_norm": 14.9375,
+      "learning_rate": 6.548148148148149e-06,
+      "loss": 0.7838,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09831111111111111,
+      "grad_norm": 10.25,
+      "learning_rate": 6.554074074074075e-06,
+      "loss": 1.0583,
+      "step": 1106
+    },
+    {
+      "epoch": 0.0984,
+      "grad_norm": 13.9375,
+      "learning_rate": 6.560000000000001e-06,
+      "loss": 0.8606,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09848888888888889,
+      "grad_norm": 18.375,
+      "learning_rate": 6.5659259259259265e-06,
+      "loss": 1.1111,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09857777777777778,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.571851851851852e-06,
+      "loss": 0.7741,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09866666666666667,
+      "grad_norm": 12.875,
+      "learning_rate": 6.577777777777779e-06,
+      "loss": 1.0664,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09875555555555555,
+      "grad_norm": 7.0625,
+      "learning_rate": 6.5837037037037045e-06,
+      "loss": 0.8058,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09884444444444444,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.58962962962963e-06,
+      "loss": 0.8916,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09893333333333333,
+      "grad_norm": 9.875,
+      "learning_rate": 6.595555555555556e-06,
+      "loss": 0.8462,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09902222222222222,
+      "grad_norm": 36.0,
+      "learning_rate": 6.601481481481482e-06,
+      "loss": 0.9232,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09911111111111111,
+      "grad_norm": 35.75,
+      "learning_rate": 6.6074074074074075e-06,
+      "loss": 0.9154,
+      "step": 1115
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 30.375,
+      "learning_rate": 6.613333333333334e-06,
+      "loss": 0.7358,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09928888888888888,
+      "grad_norm": 7.46875,
+      "learning_rate": 6.61925925925926e-06,
+      "loss": 0.9301,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09937777777777777,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.625185185185186e-06,
+      "loss": 0.92,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09946666666666666,
+      "grad_norm": 11.0,
+      "learning_rate": 6.631111111111111e-06,
+      "loss": 0.9036,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09955555555555555,
+      "grad_norm": 13.5,
+      "learning_rate": 6.637037037037037e-06,
+      "loss": 0.7808,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09964444444444444,
+      "grad_norm": 17.25,
+      "learning_rate": 6.642962962962964e-06,
+      "loss": 0.881,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09973333333333333,
+      "grad_norm": 11.3125,
+      "learning_rate": 6.648888888888889e-06,
+      "loss": 0.9252,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09982222222222223,
+      "grad_norm": 9.6875,
+      "learning_rate": 6.654814814814815e-06,
+      "loss": 0.8194,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09991111111111112,
+      "grad_norm": 8.0,
+      "learning_rate": 6.660740740740741e-06,
+      "loss": 1.0167,
+      "step": 1124
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.9235,
+      "step": 1125
+    },
+    {
+      "epoch": 0.1000888888888889,
+      "grad_norm": 7.15625,
+      "learning_rate": 6.672592592592592e-06,
+      "loss": 0.8409,
+      "step": 1126
+    },
+    {
+      "epoch": 0.10017777777777778,
+      "grad_norm": 13.5625,
+      "learning_rate": 6.678518518518519e-06,
+      "loss": 0.8438,
+      "step": 1127
+    },
+    {
+      "epoch": 0.10026666666666667,
+      "grad_norm": 19.0,
+      "learning_rate": 6.684444444444445e-06,
+      "loss": 0.9345,
+      "step": 1128
+    },
+    {
+      "epoch": 0.10035555555555556,
+      "grad_norm": 7.4375,
+      "learning_rate": 6.6903703703703705e-06,
+      "loss": 0.9212,
+      "step": 1129
+    },
+    {
+      "epoch": 0.10044444444444445,
+      "grad_norm": 6.0,
+      "learning_rate": 6.696296296296296e-06,
+      "loss": 0.8537,
+      "step": 1130
+    },
+    {
+      "epoch": 0.10053333333333334,
+      "grad_norm": 26.5,
+      "learning_rate": 6.702222222222224e-06,
+      "loss": 0.8952,
+      "step": 1131
+    },
+    {
+      "epoch": 0.10062222222222222,
+      "grad_norm": 26.75,
+      "learning_rate": 6.708148148148149e-06,
+      "loss": 0.8525,
+      "step": 1132
+    },
+    {
+      "epoch": 0.10071111111111111,
+      "grad_norm": 16.625,
+      "learning_rate": 6.714074074074075e-06,
+      "loss": 0.9141,
+      "step": 1133
+    },
+    {
+      "epoch": 0.1008,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.720000000000001e-06,
+      "loss": 0.7717,
+      "step": 1134
+    },
+    {
+      "epoch": 0.10088888888888889,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.725925925925927e-06,
+      "loss": 0.9763,
+      "step": 1135
+    },
+    {
+      "epoch": 0.10097777777777778,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.731851851851852e-06,
+      "loss": 0.9919,
+      "step": 1136
+    },
+    {
+      "epoch": 0.10106666666666667,
+      "grad_norm": 24.125,
+      "learning_rate": 6.737777777777779e-06,
+      "loss": 0.9144,
+      "step": 1137
+    },
+    {
+      "epoch": 0.10115555555555555,
+      "grad_norm": 20.75,
+      "learning_rate": 6.743703703703705e-06,
+      "loss": 0.8452,
+      "step": 1138
+    },
+    {
+      "epoch": 0.10124444444444444,
+      "grad_norm": 23.0,
+      "learning_rate": 6.7496296296296304e-06,
+      "loss": 1.1156,
+      "step": 1139
+    },
+    {
+      "epoch": 0.10133333333333333,
+      "grad_norm": 22.375,
+      "learning_rate": 6.755555555555556e-06,
+      "loss": 0.819,
+      "step": 1140
+    },
+    {
+      "epoch": 0.10142222222222222,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.761481481481482e-06,
+      "loss": 0.7988,
+      "step": 1141
+    },
+    {
+      "epoch": 0.10151111111111111,
+      "grad_norm": 19.25,
+      "learning_rate": 6.767407407407408e-06,
+      "loss": 0.889,
+      "step": 1142
+    },
+    {
+      "epoch": 0.1016,
+      "grad_norm": 19.0,
+      "learning_rate": 6.773333333333334e-06,
+      "loss": 1.0154,
+      "step": 1143
+    },
+    {
+      "epoch": 0.10168888888888888,
+      "grad_norm": 9.5625,
+      "learning_rate": 6.77925925925926e-06,
+      "loss": 0.7269,
+      "step": 1144
+    },
+    {
+      "epoch": 0.10177777777777777,
+      "grad_norm": 17.0,
+      "learning_rate": 6.785185185185186e-06,
+      "loss": 0.7763,
+      "step": 1145
+    },
+    {
+      "epoch": 0.10186666666666666,
+      "grad_norm": 9.0,
+      "learning_rate": 6.7911111111111115e-06,
+      "loss": 0.7606,
+      "step": 1146
+    },
+    {
+      "epoch": 0.10195555555555555,
+      "grad_norm": 15.1875,
+      "learning_rate": 6.797037037037037e-06,
+      "loss": 0.9226,
+      "step": 1147
+    },
+    {
+      "epoch": 0.10204444444444444,
+      "grad_norm": 9.6875,
+      "learning_rate": 6.802962962962964e-06,
+      "loss": 0.8795,
+      "step": 1148
+    },
+    {
+      "epoch": 0.10213333333333334,
+      "grad_norm": 17.75,
+      "learning_rate": 6.80888888888889e-06,
+      "loss": 0.911,
+      "step": 1149
+    },
+    {
+      "epoch": 0.10222222222222223,
+      "grad_norm": 27.375,
+      "learning_rate": 6.814814814814815e-06,
+      "loss": 0.8519,
+      "step": 1150
+    },
+    {
+      "epoch": 0.10231111111111112,
+      "grad_norm": 27.375,
+      "learning_rate": 6.820740740740741e-06,
+      "loss": 0.8917,
+      "step": 1151
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 14.9375,
+      "learning_rate": 6.826666666666667e-06,
+      "loss": 0.8839,
+      "step": 1152
+    },
+    {
+      "epoch": 0.1024888888888889,
+      "grad_norm": 11.4375,
+      "learning_rate": 6.8325925925925926e-06,
+      "loss": 0.8848,
+      "step": 1153
+    },
+    {
+      "epoch": 0.10257777777777778,
+      "grad_norm": 16.125,
+      "learning_rate": 6.838518518518519e-06,
+      "loss": 0.9181,
+      "step": 1154
+    },
+    {
+      "epoch": 0.10266666666666667,
+      "grad_norm": 34.75,
+      "learning_rate": 6.844444444444445e-06,
+      "loss": 0.8258,
+      "step": 1155
+    },
+    {
+      "epoch": 0.10275555555555556,
+      "grad_norm": 9.0,
+      "learning_rate": 6.850370370370371e-06,
+      "loss": 1.0891,
+      "step": 1156
+    },
+    {
+      "epoch": 0.10284444444444445,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.856296296296296e-06,
+      "loss": 0.7986,
+      "step": 1157
+    },
+    {
+      "epoch": 0.10293333333333334,
+      "grad_norm": 7.625,
+      "learning_rate": 6.862222222222222e-06,
+      "loss": 0.9662,
+      "step": 1158
+    },
+    {
+      "epoch": 0.10302222222222222,
+      "grad_norm": 7.75,
+      "learning_rate": 6.868148148148148e-06,
+      "loss": 0.9963,
+      "step": 1159
+    },
+    {
+      "epoch": 0.10311111111111111,
+      "grad_norm": 9.5,
+      "learning_rate": 6.8740740740740745e-06,
+      "loss": 0.9407,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1032,
+      "grad_norm": 19.625,
+      "learning_rate": 6.88e-06,
+      "loss": 1.0765,
+      "step": 1161
+    },
+    {
+      "epoch": 0.10328888888888889,
+      "grad_norm": 15.125,
+      "learning_rate": 6.885925925925927e-06,
+      "loss": 0.723,
+      "step": 1162
+    },
+    {
+      "epoch": 0.10337777777777778,
+      "grad_norm": 10.375,
+      "learning_rate": 6.8918518518518525e-06,
+      "loss": 0.7525,
+      "step": 1163
+    },
+    {
+      "epoch": 0.10346666666666667,
+      "grad_norm": 6.53125,
+      "learning_rate": 6.897777777777779e-06,
+      "loss": 0.8746,
+      "step": 1164
+    },
+    {
+      "epoch": 0.10355555555555555,
+      "grad_norm": 18.875,
+      "learning_rate": 6.903703703703705e-06,
+      "loss": 0.7874,
+      "step": 1165
+    },
+    {
+      "epoch": 0.10364444444444444,
+      "grad_norm": 18.5,
+      "learning_rate": 6.909629629629631e-06,
+      "loss": 0.7632,
+      "step": 1166
+    },
+    {
+      "epoch": 0.10373333333333333,
+      "grad_norm": 12.75,
+      "learning_rate": 6.915555555555556e-06,
+      "loss": 0.9784,
+      "step": 1167
+    },
+    {
+      "epoch": 0.10382222222222222,
+      "grad_norm": 17.75,
+      "learning_rate": 6.921481481481482e-06,
+      "loss": 0.82,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1039111111111111,
+      "grad_norm": 15.6875,
+      "learning_rate": 6.927407407407408e-06,
+      "loss": 0.6991,
+      "step": 1169
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.9333333333333344e-06,
+      "loss": 0.9267,
+      "step": 1170
+    },
+    {
+      "epoch": 0.10408888888888888,
+      "grad_norm": 7.875,
+      "learning_rate": 6.93925925925926e-06,
+      "loss": 0.7307,
+      "step": 1171
+    },
+    {
+      "epoch": 0.10417777777777777,
+      "grad_norm": 11.25,
+      "learning_rate": 6.945185185185186e-06,
+      "loss": 0.9462,
+      "step": 1172
+    },
+    {
+      "epoch": 0.10426666666666666,
+      "grad_norm": 10.4375,
+      "learning_rate": 6.951111111111112e-06,
+      "loss": 0.9953,
+      "step": 1173
+    },
+    {
+      "epoch": 0.10435555555555555,
+      "grad_norm": 5.84375,
+      "learning_rate": 6.957037037037037e-06,
+      "loss": 1.0094,
+      "step": 1174
+    },
+    {
+      "epoch": 0.10444444444444445,
+      "grad_norm": 23.125,
+      "learning_rate": 6.962962962962964e-06,
+      "loss": 0.8601,
+      "step": 1175
+    },
+    {
+      "epoch": 0.10453333333333334,
+      "grad_norm": 6.09375,
+      "learning_rate": 6.96888888888889e-06,
+      "loss": 0.6933,
+      "step": 1176
+    },
+    {
+      "epoch": 0.10462222222222223,
+      "grad_norm": 17.75,
+      "learning_rate": 6.9748148148148155e-06,
+      "loss": 0.9342,
+      "step": 1177
+    },
+    {
+      "epoch": 0.10471111111111112,
+      "grad_norm": 12.75,
+      "learning_rate": 6.980740740740741e-06,
+      "loss": 0.9157,
+      "step": 1178
+    },
+    {
+      "epoch": 0.1048,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.986666666666667e-06,
+      "loss": 0.9712,
+      "step": 1179
+    },
+    {
+      "epoch": 0.10488888888888889,
+      "grad_norm": 23.625,
+      "learning_rate": 6.992592592592593e-06,
+      "loss": 0.8507,
+      "step": 1180
+    },
+    {
+      "epoch": 0.10497777777777778,
+      "grad_norm": 24.5,
+      "learning_rate": 6.998518518518519e-06,
+      "loss": 0.8381,
+      "step": 1181
+    },
+    {
+      "epoch": 0.10506666666666667,
+      "grad_norm": 11.875,
+      "learning_rate": 7.004444444444445e-06,
+      "loss": 1.0284,
+      "step": 1182
+    },
+    {
+      "epoch": 0.10515555555555556,
+      "grad_norm": 24.125,
+      "learning_rate": 7.010370370370371e-06,
+      "loss": 0.9069,
+      "step": 1183
+    },
+    {
+      "epoch": 0.10524444444444445,
+      "grad_norm": 25.75,
+      "learning_rate": 7.0162962962962965e-06,
+      "loss": 0.8779,
+      "step": 1184
+    },
+    {
+      "epoch": 0.10533333333333333,
+      "grad_norm": 31.5,
+      "learning_rate": 7.022222222222222e-06,
+      "loss": 0.7875,
+      "step": 1185
+    },
+    {
+      "epoch": 0.10542222222222222,
+      "grad_norm": 11.8125,
+      "learning_rate": 7.028148148148148e-06,
+      "loss": 0.9332,
+      "step": 1186
+    },
+    {
+      "epoch": 0.10551111111111111,
+      "grad_norm": 11.5,
+      "learning_rate": 7.034074074074075e-06,
+      "loss": 0.9105,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 6.625,
+      "learning_rate": 7.04e-06,
+      "loss": 0.9609,
+      "step": 1188
+    },
+    {
+      "epoch": 0.10568888888888889,
+      "grad_norm": 25.0,
+      "learning_rate": 7.045925925925926e-06,
+      "loss": 0.8688,
+      "step": 1189
+    },
+    {
+      "epoch": 0.10577777777777778,
+      "grad_norm": 16.875,
+      "learning_rate": 7.051851851851852e-06,
+      "loss": 0.8308,
+      "step": 1190
+    },
+    {
+      "epoch": 0.10586666666666666,
+      "grad_norm": 9.625,
+      "learning_rate": 7.057777777777778e-06,
+      "loss": 0.7836,
+      "step": 1191
+    },
+    {
+      "epoch": 0.10595555555555555,
+      "grad_norm": 30.125,
+      "learning_rate": 7.063703703703703e-06,
+      "loss": 0.781,
+      "step": 1192
+    },
+    {
+      "epoch": 0.10604444444444444,
+      "grad_norm": 34.75,
+      "learning_rate": 7.069629629629631e-06,
+      "loss": 0.8113,
+      "step": 1193
+    },
+    {
+      "epoch": 0.10613333333333333,
+      "grad_norm": 34.25,
+      "learning_rate": 7.0755555555555565e-06,
+      "loss": 1.0321,
+      "step": 1194
+    },
+    {
+      "epoch": 0.10622222222222222,
+      "grad_norm": 6.71875,
+      "learning_rate": 7.081481481481482e-06,
+      "loss": 0.7164,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1063111111111111,
+      "grad_norm": 7.15625,
+      "learning_rate": 7.087407407407408e-06,
+      "loss": 0.8685,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1064,
+      "grad_norm": 19.0,
+      "learning_rate": 7.093333333333335e-06,
+      "loss": 0.8009,
+      "step": 1197
+    },
+    {
+      "epoch": 0.10648888888888888,
+      "grad_norm": 10.4375,
+      "learning_rate": 7.09925925925926e-06,
+      "loss": 0.8445,
+      "step": 1198
+    },
+    {
+      "epoch": 0.10657777777777777,
+      "grad_norm": 14.4375,
+      "learning_rate": 7.105185185185186e-06,
+      "loss": 0.8534,
+      "step": 1199
+    },
+    {
+      "epoch": 0.10666666666666667,
+      "grad_norm": 6.90625,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 0.8539,
+      "step": 1200
+    },
+    {
+      "epoch": 0.10675555555555556,
+      "grad_norm": 13.125,
+      "learning_rate": 7.1170370370370376e-06,
+      "loss": 0.5525,
+      "step": 1201
+    },
+    {
+      "epoch": 0.10684444444444445,
+      "grad_norm": 18.625,
+      "learning_rate": 7.122962962962964e-06,
+      "loss": 0.7004,
+      "step": 1202
+    },
+    {
+      "epoch": 0.10693333333333334,
+      "grad_norm": 12.375,
+      "learning_rate": 7.12888888888889e-06,
+      "loss": 0.8904,
+      "step": 1203
+    },
+    {
+      "epoch": 0.10702222222222223,
+      "grad_norm": 7.15625,
+      "learning_rate": 7.134814814814816e-06,
+      "loss": 0.8292,
+      "step": 1204
+    },
+    {
+      "epoch": 0.10711111111111112,
+      "grad_norm": 21.0,
+      "learning_rate": 7.140740740740741e-06,
+      "loss": 0.8406,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1072,
+      "grad_norm": 29.25,
+      "learning_rate": 7.146666666666667e-06,
+      "loss": 0.7008,
+      "step": 1206
+    },
+    {
+      "epoch": 0.10728888888888889,
+      "grad_norm": 21.0,
+      "learning_rate": 7.152592592592593e-06,
+      "loss": 0.7011,
+      "step": 1207
+    },
+    {
+      "epoch": 0.10737777777777778,
+      "grad_norm": 8.625,
+      "learning_rate": 7.1585185185185195e-06,
+      "loss": 0.7455,
+      "step": 1208
+    },
+    {
+      "epoch": 0.10746666666666667,
+      "grad_norm": 25.0,
+      "learning_rate": 7.164444444444445e-06,
+      "loss": 0.7435,
+      "step": 1209
+    },
+    {
+      "epoch": 0.10755555555555556,
+      "grad_norm": 14.0,
+      "learning_rate": 7.170370370370371e-06,
+      "loss": 0.8841,
+      "step": 1210
+    },
+    {
+      "epoch": 0.10764444444444445,
+      "grad_norm": 9.875,
+      "learning_rate": 7.176296296296297e-06,
+      "loss": 0.9084,
+      "step": 1211
+    },
+    {
+      "epoch": 0.10773333333333333,
+      "grad_norm": 10.625,
+      "learning_rate": 7.1822222222222224e-06,
+      "loss": 0.7414,
+      "step": 1212
+    },
+    {
+      "epoch": 0.10782222222222222,
+      "grad_norm": 7.9375,
+      "learning_rate": 7.188148148148148e-06,
+      "loss": 1.0047,
+      "step": 1213
+    },
+    {
+      "epoch": 0.10791111111111111,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.194074074074075e-06,
+      "loss": 0.8097,
+      "step": 1214
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 9.4375,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.909,
+      "step": 1215
+    },
+    {
+      "epoch": 0.10808888888888889,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.205925925925926e-06,
+      "loss": 0.7753,
+      "step": 1216
+    },
+    {
+      "epoch": 0.10817777777777778,
+      "grad_norm": 14.9375,
+      "learning_rate": 7.211851851851852e-06,
+      "loss": 0.7576,
+      "step": 1217
+    },
+    {
+      "epoch": 0.10826666666666666,
+      "grad_norm": 15.625,
+      "learning_rate": 7.217777777777778e-06,
+      "loss": 0.8713,
+      "step": 1218
+    },
+    {
+      "epoch": 0.10835555555555555,
+      "grad_norm": 7.59375,
+      "learning_rate": 7.2237037037037035e-06,
+      "loss": 0.875,
+      "step": 1219
+    },
+    {
+      "epoch": 0.10844444444444444,
+      "grad_norm": 15.125,
+      "learning_rate": 7.22962962962963e-06,
+      "loss": 0.8328,
+      "step": 1220
+    },
+    {
+      "epoch": 0.10853333333333333,
+      "grad_norm": 9.75,
+      "learning_rate": 7.235555555555556e-06,
+      "loss": 0.8398,
+      "step": 1221
+    },
+    {
+      "epoch": 0.10862222222222222,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.2414814814814816e-06,
+      "loss": 0.919,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1087111111111111,
+      "grad_norm": 5.375,
+      "learning_rate": 7.247407407407407e-06,
+      "loss": 0.8198,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 9.375,
+      "learning_rate": 7.253333333333335e-06,
+      "loss": 0.7871,
+      "step": 1224
+    },
+    {
+      "epoch": 0.10888888888888888,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.2592592592592605e-06,
+      "loss": 0.9701,
+      "step": 1225
+    },
+    {
+      "epoch": 0.10897777777777778,
+      "grad_norm": 10.5625,
+      "learning_rate": 7.265185185185186e-06,
+      "loss": 0.8649,
+      "step": 1226
+    },
+    {
+      "epoch": 0.10906666666666667,
+      "grad_norm": 9.125,
+      "learning_rate": 7.271111111111112e-06,
+      "loss": 0.9422,
+      "step": 1227
+    },
+    {
+      "epoch": 0.10915555555555556,
+      "grad_norm": 11.5625,
+      "learning_rate": 7.277037037037038e-06,
+      "loss": 0.8875,
+      "step": 1228
+    },
+    {
+      "epoch": 0.10924444444444445,
+      "grad_norm": 20.875,
+      "learning_rate": 7.282962962962964e-06,
+      "loss": 0.8457,
+      "step": 1229
+    },
+    {
+      "epoch": 0.10933333333333334,
+      "grad_norm": 12.9375,
+      "learning_rate": 7.28888888888889e-06,
+      "loss": 0.8127,
+      "step": 1230
+    },
+    {
+      "epoch": 0.10942222222222223,
+      "grad_norm": 14.0,
+      "learning_rate": 7.294814814814816e-06,
+      "loss": 0.9012,
+      "step": 1231
+    },
+    {
+      "epoch": 0.10951111111111111,
+      "grad_norm": 7.0,
+      "learning_rate": 7.3007407407407416e-06,
+      "loss": 0.6902,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1096,
+      "grad_norm": 10.75,
+      "learning_rate": 7.306666666666667e-06,
+      "loss": 0.858,
+      "step": 1233
+    },
+    {
+      "epoch": 0.10968888888888889,
+      "grad_norm": 10.0,
+      "learning_rate": 7.312592592592593e-06,
+      "loss": 0.725,
+      "step": 1234
+    },
+    {
+      "epoch": 0.10977777777777778,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.31851851851852e-06,
+      "loss": 1.0646,
+      "step": 1235
+    },
+    {
+      "epoch": 0.10986666666666667,
+      "grad_norm": 10.5,
+      "learning_rate": 7.324444444444445e-06,
+      "loss": 0.6865,
+      "step": 1236
+    },
+    {
+      "epoch": 0.10995555555555556,
+      "grad_norm": 12.5,
+      "learning_rate": 7.330370370370371e-06,
+      "loss": 0.8453,
+      "step": 1237
+    },
+    {
+      "epoch": 0.11004444444444444,
+      "grad_norm": 13.125,
+      "learning_rate": 7.336296296296297e-06,
+      "loss": 0.8852,
+      "step": 1238
+    },
+    {
+      "epoch": 0.11013333333333333,
+      "grad_norm": 16.0,
+      "learning_rate": 7.342222222222223e-06,
+      "loss": 0.9035,
+      "step": 1239
+    },
+    {
+      "epoch": 0.11022222222222222,
+      "grad_norm": 5.71875,
+      "learning_rate": 7.348148148148148e-06,
+      "loss": 0.8255,
+      "step": 1240
+    },
+    {
+      "epoch": 0.11031111111111111,
+      "grad_norm": 12.6875,
+      "learning_rate": 7.354074074074075e-06,
+      "loss": 0.8841,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1104,
+      "grad_norm": 10.0625,
+      "learning_rate": 7.360000000000001e-06,
+      "loss": 0.9327,
+      "step": 1242
+    },
+    {
+      "epoch": 0.11048888888888889,
+      "grad_norm": 8.75,
+      "learning_rate": 7.3659259259259264e-06,
+      "loss": 0.8033,
+      "step": 1243
+    },
+    {
+      "epoch": 0.11057777777777777,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.371851851851852e-06,
+      "loss": 0.7512,
+      "step": 1244
+    },
+    {
+      "epoch": 0.11066666666666666,
+      "grad_norm": 13.375,
+      "learning_rate": 7.377777777777778e-06,
+      "loss": 0.8582,
+      "step": 1245
+    },
+    {
+      "epoch": 0.11075555555555555,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.383703703703704e-06,
+      "loss": 0.8241,
+      "step": 1246
+    },
+    {
+      "epoch": 0.11084444444444444,
+      "grad_norm": 5.90625,
+      "learning_rate": 7.38962962962963e-06,
+      "loss": 0.8721,
+      "step": 1247
+    },
+    {
+      "epoch": 0.11093333333333333,
+      "grad_norm": 6.0625,
+      "learning_rate": 7.395555555555556e-06,
+      "loss": 0.7601,
+      "step": 1248
+    },
+    {
+      "epoch": 0.11102222222222222,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.401481481481482e-06,
+      "loss": 0.8948,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 19.625,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.8345,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1112,
+      "grad_norm": 13.875,
+      "learning_rate": 7.413333333333333e-06,
+      "loss": 0.849,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1112888888888889,
+      "grad_norm": 15.8125,
+      "learning_rate": 7.419259259259259e-06,
+      "loss": 0.7141,
+      "step": 1252
+    },
+    {
+      "epoch": 0.11137777777777778,
+      "grad_norm": 17.0,
+      "learning_rate": 7.4251851851851856e-06,
+      "loss": 0.5895,
+      "step": 1253
+    },
+    {
+      "epoch": 0.11146666666666667,
+      "grad_norm": 7.25,
+      "learning_rate": 7.431111111111111e-06,
+      "loss": 0.8366,
+      "step": 1254
+    },
+    {
+      "epoch": 0.11155555555555556,
+      "grad_norm": 6.0,
+      "learning_rate": 7.437037037037038e-06,
+      "loss": 0.8703,
+      "step": 1255
+    },
+    {
+      "epoch": 0.11164444444444445,
+      "grad_norm": 13.8125,
+      "learning_rate": 7.442962962962964e-06,
+      "loss": 0.8208,
+      "step": 1256
+    },
+    {
+      "epoch": 0.11173333333333334,
+      "grad_norm": 9.125,
+      "learning_rate": 7.44888888888889e-06,
+      "loss": 0.9369,
+      "step": 1257
+    },
+    {
+      "epoch": 0.11182222222222223,
+      "grad_norm": 6.1875,
+      "learning_rate": 7.454814814814816e-06,
+      "loss": 0.7371,
+      "step": 1258
+    },
+    {
+      "epoch": 0.11191111111111111,
+      "grad_norm": 7.375,
+      "learning_rate": 7.460740740740742e-06,
+      "loss": 0.6644,
+      "step": 1259
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 10.25,
+      "learning_rate": 7.4666666666666675e-06,
+      "loss": 0.836,
+      "step": 1260
+    },
+    {
+      "epoch": 0.11208888888888889,
+      "grad_norm": 11.6875,
+      "learning_rate": 7.472592592592593e-06,
+      "loss": 0.942,
+      "step": 1261
+    },
+    {
+      "epoch": 0.11217777777777778,
+      "grad_norm": 5.90625,
+      "learning_rate": 7.47851851851852e-06,
+      "loss": 0.8785,
+      "step": 1262
+    },
+    {
+      "epoch": 0.11226666666666667,
+      "grad_norm": 6.15625,
+      "learning_rate": 7.4844444444444455e-06,
+      "loss": 0.8412,
+      "step": 1263
+    },
+    {
+      "epoch": 0.11235555555555556,
+      "grad_norm": 11.375,
+      "learning_rate": 7.490370370370371e-06,
+      "loss": 0.8434,
+      "step": 1264
+    },
+    {
+      "epoch": 0.11244444444444444,
+      "grad_norm": 9.0,
+      "learning_rate": 7.496296296296297e-06,
+      "loss": 0.7112,
+      "step": 1265
+    },
+    {
+      "epoch": 0.11253333333333333,
+      "grad_norm": 6.65625,
+      "learning_rate": 7.502222222222223e-06,
+      "loss": 0.903,
+      "step": 1266
+    },
+    {
+      "epoch": 0.11262222222222222,
+      "grad_norm": 13.0,
+      "learning_rate": 7.5081481481481485e-06,
+      "loss": 0.8774,
+      "step": 1267
+    },
+    {
+      "epoch": 0.11271111111111111,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.514074074074075e-06,
+      "loss": 0.9165,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1128,
+      "grad_norm": 5.4375,
+      "learning_rate": 7.520000000000001e-06,
+      "loss": 0.9077,
+      "step": 1269
+    },
+    {
+      "epoch": 0.11288888888888889,
+      "grad_norm": 10.0,
+      "learning_rate": 7.525925925925927e-06,
+      "loss": 0.7027,
+      "step": 1270
+    },
+    {
+      "epoch": 0.11297777777777777,
+      "grad_norm": 6.65625,
+      "learning_rate": 7.531851851851852e-06,
+      "loss": 0.8017,
+      "step": 1271
+    },
+    {
+      "epoch": 0.11306666666666666,
+      "grad_norm": 8.125,
+      "learning_rate": 7.537777777777778e-06,
+      "loss": 0.8544,
+      "step": 1272
+    },
+    {
+      "epoch": 0.11315555555555555,
+      "grad_norm": 9.125,
+      "learning_rate": 7.543703703703704e-06,
+      "loss": 0.8285,
+      "step": 1273
+    },
+    {
+      "epoch": 0.11324444444444444,
+      "grad_norm": 7.875,
+      "learning_rate": 7.54962962962963e-06,
+      "loss": 0.8885,
+      "step": 1274
+    },
+    {
+      "epoch": 0.11333333333333333,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 0.8201,
+      "step": 1275
+    },
+    {
+      "epoch": 0.11342222222222222,
+      "grad_norm": 8.625,
+      "learning_rate": 7.561481481481482e-06,
+      "loss": 0.7593,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1135111111111111,
+      "grad_norm": 6.25,
+      "learning_rate": 7.567407407407408e-06,
+      "loss": 0.7734,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1136,
+      "grad_norm": 8.75,
+      "learning_rate": 7.573333333333333e-06,
+      "loss": 0.7114,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1136888888888889,
+      "grad_norm": 8.375,
+      "learning_rate": 7.579259259259259e-06,
+      "loss": 0.8204,
+      "step": 1279
+    },
+    {
+      "epoch": 0.11377777777777778,
+      "grad_norm": 5.25,
+      "learning_rate": 7.585185185185186e-06,
+      "loss": 0.9045,
+      "step": 1280
+    },
+    {
+      "epoch": 0.11386666666666667,
+      "grad_norm": 12.3125,
+      "learning_rate": 7.5911111111111115e-06,
+      "loss": 0.7439,
+      "step": 1281
+    },
+    {
+      "epoch": 0.11395555555555556,
+      "grad_norm": 8.5,
+      "learning_rate": 7.597037037037037e-06,
+      "loss": 0.7282,
+      "step": 1282
+    },
+    {
+      "epoch": 0.11404444444444445,
+      "grad_norm": 6.15625,
+      "learning_rate": 7.602962962962963e-06,
+      "loss": 0.8721,
+      "step": 1283
+    },
+    {
+      "epoch": 0.11413333333333334,
+      "grad_norm": 12.9375,
+      "learning_rate": 7.608888888888889e-06,
+      "loss": 0.8243,
+      "step": 1284
+    },
+    {
+      "epoch": 0.11422222222222222,
+      "grad_norm": 27.625,
+      "learning_rate": 7.614814814814816e-06,
+      "loss": 0.7299,
+      "step": 1285
+    },
+    {
+      "epoch": 0.11431111111111111,
+      "grad_norm": 8.8125,
+      "learning_rate": 7.620740740740742e-06,
+      "loss": 0.9335,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1144,
+      "grad_norm": 9.1875,
+      "learning_rate": 7.626666666666668e-06,
+      "loss": 0.7898,
+      "step": 1287
+    },
+    {
+      "epoch": 0.11448888888888889,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.632592592592593e-06,
+      "loss": 0.9031,
+      "step": 1288
+    },
+    {
+      "epoch": 0.11457777777777778,
+      "grad_norm": 13.5625,
+      "learning_rate": 7.638518518518519e-06,
+      "loss": 0.6925,
+      "step": 1289
+    },
+    {
+      "epoch": 0.11466666666666667,
+      "grad_norm": 9.125,
+      "learning_rate": 7.644444444444445e-06,
+      "loss": 0.8144,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11475555555555556,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.65037037037037e-06,
+      "loss": 0.7776,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11484444444444444,
+      "grad_norm": 13.875,
+      "learning_rate": 7.656296296296298e-06,
+      "loss": 0.7395,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11493333333333333,
+      "grad_norm": 10.625,
+      "learning_rate": 7.662222222222224e-06,
+      "loss": 0.8094,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11502222222222222,
+      "grad_norm": 7.03125,
+      "learning_rate": 7.66814814814815e-06,
+      "loss": 0.7831,
+      "step": 1294
+    },
+    {
+      "epoch": 0.11511111111111111,
+      "grad_norm": 7.53125,
+      "learning_rate": 7.674074074074075e-06,
+      "loss": 0.879,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 5.96875,
+      "learning_rate": 7.680000000000001e-06,
+      "loss": 0.8789,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11528888888888889,
+      "grad_norm": 27.75,
+      "learning_rate": 7.685925925925927e-06,
+      "loss": 0.8374,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11537777777777777,
+      "grad_norm": 10.9375,
+      "learning_rate": 7.691851851851852e-06,
+      "loss": 0.7157,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11546666666666666,
+      "grad_norm": 7.3125,
+      "learning_rate": 7.697777777777778e-06,
+      "loss": 0.8837,
+      "step": 1299
+    },
+    {
+      "epoch": 0.11555555555555555,
+      "grad_norm": 8.625,
+      "learning_rate": 7.703703703703704e-06,
+      "loss": 0.8037,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11564444444444444,
+      "grad_norm": 9.0625,
+      "learning_rate": 7.70962962962963e-06,
+      "loss": 0.7825,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11573333333333333,
+      "grad_norm": 10.9375,
+      "learning_rate": 7.715555555555555e-06,
+      "loss": 0.8615,
+      "step": 1302
+    },
+    {
+      "epoch": 0.11582222222222222,
+      "grad_norm": 5.71875,
+      "learning_rate": 7.721481481481481e-06,
+      "loss": 0.9494,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11591111111111112,
+      "grad_norm": 4.3125,
+      "learning_rate": 7.727407407407409e-06,
+      "loss": 0.9054,
+      "step": 1304
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.733333333333334e-06,
+      "loss": 0.9066,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1160888888888889,
+      "grad_norm": 8.25,
+      "learning_rate": 7.73925925925926e-06,
+      "loss": 0.745,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11617777777777778,
+      "grad_norm": 9.3125,
+      "learning_rate": 7.745185185185186e-06,
+      "loss": 1.003,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11626666666666667,
+      "grad_norm": 6.65625,
+      "learning_rate": 7.751111111111112e-06,
+      "loss": 0.8174,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11635555555555556,
+      "grad_norm": 6.375,
+      "learning_rate": 7.757037037037037e-06,
+      "loss": 0.7773,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11644444444444445,
+      "grad_norm": 53.5,
+      "learning_rate": 7.762962962962963e-06,
+      "loss": 0.8396,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11653333333333334,
+      "grad_norm": 34.0,
+      "learning_rate": 7.768888888888889e-06,
+      "loss": 0.6927,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11662222222222222,
+      "grad_norm": 58.5,
+      "learning_rate": 7.774814814814815e-06,
+      "loss": 0.7829,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11671111111111111,
+      "grad_norm": 20.25,
+      "learning_rate": 7.78074074074074e-06,
+      "loss": 0.9433,
+      "step": 1313
+    },
+    {
+      "epoch": 0.1168,
+      "grad_norm": 9.4375,
+      "learning_rate": 7.786666666666666e-06,
+      "loss": 0.816,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11688888888888889,
+      "grad_norm": 31.75,
+      "learning_rate": 7.792592592592594e-06,
+      "loss": 0.8541,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11697777777777778,
+      "grad_norm": 27.0,
+      "learning_rate": 7.79851851851852e-06,
+      "loss": 0.7087,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11706666666666667,
+      "grad_norm": 25.5,
+      "learning_rate": 7.804444444444445e-06,
+      "loss": 0.6958,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11715555555555555,
+      "grad_norm": 26.5,
+      "learning_rate": 7.81037037037037e-06,
+      "loss": 0.6955,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11724444444444444,
+      "grad_norm": 10.1875,
+      "learning_rate": 7.816296296296298e-06,
+      "loss": 0.722,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11733333333333333,
+      "grad_norm": 15.5625,
+      "learning_rate": 7.822222222222224e-06,
+      "loss": 0.6799,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11742222222222222,
+      "grad_norm": 10.0,
+      "learning_rate": 7.82814814814815e-06,
+      "loss": 0.7793,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11751111111111111,
+      "grad_norm": 6.46875,
+      "learning_rate": 7.834074074074075e-06,
+      "loss": 0.7437,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1176,
+      "grad_norm": 10.375,
+      "learning_rate": 7.840000000000001e-06,
+      "loss": 0.7464,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11768888888888888,
+      "grad_norm": 9.125,
+      "learning_rate": 7.845925925925927e-06,
+      "loss": 0.8849,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11777777777777777,
+      "grad_norm": 7.28125,
+      "learning_rate": 7.851851851851853e-06,
+      "loss": 0.9145,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11786666666666666,
+      "grad_norm": 14.25,
+      "learning_rate": 7.857777777777778e-06,
+      "loss": 0.6774,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11795555555555555,
+      "grad_norm": 9.9375,
+      "learning_rate": 7.863703703703704e-06,
+      "loss": 0.93,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11804444444444444,
+      "grad_norm": 11.8125,
+      "learning_rate": 7.86962962962963e-06,
+      "loss": 0.8051,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11813333333333334,
+      "grad_norm": 12.75,
+      "learning_rate": 7.875555555555556e-06,
+      "loss": 0.8184,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11822222222222223,
+      "grad_norm": 4.84375,
+      "learning_rate": 7.881481481481481e-06,
+      "loss": 0.8202,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11831111111111112,
+      "grad_norm": 19.75,
+      "learning_rate": 7.887407407407409e-06,
+      "loss": 0.8024,
+      "step": 1331
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 27.5,
+      "learning_rate": 7.893333333333335e-06,
+      "loss": 0.8174,
+      "step": 1332
+    },
+    {
+      "epoch": 0.1184888888888889,
+      "grad_norm": 25.125,
+      "learning_rate": 7.89925925925926e-06,
+      "loss": 0.8432,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11857777777777778,
+      "grad_norm": 9.4375,
+      "learning_rate": 7.905185185185186e-06,
+      "loss": 0.8528,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11866666666666667,
+      "grad_norm": 27.625,
+      "learning_rate": 7.911111111111112e-06,
+      "loss": 0.8119,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11875555555555556,
+      "grad_norm": 61.5,
+      "learning_rate": 7.917037037037038e-06,
+      "loss": 0.7115,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11884444444444445,
+      "grad_norm": 50.5,
+      "learning_rate": 7.922962962962963e-06,
+      "loss": 0.9371,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11893333333333334,
+      "grad_norm": 21.875,
+      "learning_rate": 7.928888888888889e-06,
+      "loss": 0.8263,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11902222222222222,
+      "grad_norm": 10.875,
+      "learning_rate": 7.934814814814815e-06,
+      "loss": 0.7817,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11911111111111111,
+      "grad_norm": 10.8125,
+      "learning_rate": 7.94074074074074e-06,
+      "loss": 0.8277,
+      "step": 1340
+    },
+    {
+      "epoch": 0.1192,
+      "grad_norm": 10.375,
+      "learning_rate": 7.946666666666666e-06,
+      "loss": 0.7695,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11928888888888889,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.952592592592594e-06,
+      "loss": 0.7538,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11937777777777778,
+      "grad_norm": 8.8125,
+      "learning_rate": 7.95851851851852e-06,
+      "loss": 0.811,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11946666666666667,
+      "grad_norm": 9.625,
+      "learning_rate": 7.964444444444445e-06,
+      "loss": 0.705,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11955555555555555,
+      "grad_norm": 5.625,
+      "learning_rate": 7.970370370370371e-06,
+      "loss": 0.6921,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11964444444444444,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.976296296296297e-06,
+      "loss": 0.7692,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11973333333333333,
+      "grad_norm": 9.375,
+      "learning_rate": 7.982222222222224e-06,
+      "loss": 0.7243,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11982222222222222,
+      "grad_norm": 7.0625,
+      "learning_rate": 7.98814814814815e-06,
+      "loss": 0.8506,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11991111111111111,
+      "grad_norm": 5.5,
+      "learning_rate": 7.994074074074076e-06,
+      "loss": 0.7501,
+      "step": 1349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.34375,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.7822,
+      "step": 1350
+    },
+    {
+      "epoch": 0.12008888888888888,
+      "grad_norm": 6.3125,
+      "learning_rate": 8.005925925925927e-06,
+      "loss": 0.8219,
+      "step": 1351
+    },
+    {
+      "epoch": 0.12017777777777777,
+      "grad_norm": 5.25,
+      "learning_rate": 8.011851851851853e-06,
+      "loss": 0.819,
+      "step": 1352
+    },
+    {
+      "epoch": 0.12026666666666666,
+      "grad_norm": 5.375,
+      "learning_rate": 8.017777777777779e-06,
+      "loss": 0.7614,
+      "step": 1353
+    },
+    {
+      "epoch": 0.12035555555555555,
+      "grad_norm": 5.125,
+      "learning_rate": 8.023703703703704e-06,
+      "loss": 0.7793,
+      "step": 1354
+    },
+    {
+      "epoch": 0.12044444444444445,
+      "grad_norm": 4.84375,
+      "learning_rate": 8.02962962962963e-06,
+      "loss": 0.7374,
+      "step": 1355
+    },
+    {
+      "epoch": 0.12053333333333334,
+      "grad_norm": 5.65625,
+      "learning_rate": 8.035555555555556e-06,
+      "loss": 0.7353,
+      "step": 1356
+    },
+    {
+      "epoch": 0.12062222222222223,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.041481481481482e-06,
+      "loss": 0.7345,
+      "step": 1357
+    },
+    {
+      "epoch": 0.12071111111111112,
+      "grad_norm": 13.25,
+      "learning_rate": 8.047407407407409e-06,
+      "loss": 0.7433,
+      "step": 1358
+    },
+    {
+      "epoch": 0.1208,
+      "grad_norm": 5.8125,
+      "learning_rate": 8.053333333333335e-06,
+      "loss": 0.8487,
+      "step": 1359
+    },
+    {
+      "epoch": 0.12088888888888889,
+      "grad_norm": 3.703125,
+      "learning_rate": 8.05925925925926e-06,
+      "loss": 0.8406,
+      "step": 1360
+    },
+    {
+      "epoch": 0.12097777777777778,
+      "grad_norm": 5.96875,
+      "learning_rate": 8.065185185185186e-06,
+      "loss": 0.7369,
+      "step": 1361
+    },
+    {
+      "epoch": 0.12106666666666667,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.071111111111112e-06,
+      "loss": 0.6819,
+      "step": 1362
+    },
+    {
+      "epoch": 0.12115555555555556,
+      "grad_norm": 5.34375,
+      "learning_rate": 8.077037037037038e-06,
+      "loss": 0.9302,
+      "step": 1363
+    },
+    {
+      "epoch": 0.12124444444444445,
+      "grad_norm": 10.375,
+      "learning_rate": 8.082962962962963e-06,
+      "loss": 0.7301,
+      "step": 1364
+    },
+    {
+      "epoch": 0.12133333333333333,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.08888888888889e-06,
+      "loss": 0.7602,
+      "step": 1365
+    },
+    {
+      "epoch": 0.12142222222222222,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.094814814814815e-06,
+      "loss": 0.919,
+      "step": 1366
+    },
+    {
+      "epoch": 0.12151111111111111,
+      "grad_norm": 28.375,
+      "learning_rate": 8.10074074074074e-06,
+      "loss": 0.7997,
+      "step": 1367
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 34.5,
+      "learning_rate": 8.106666666666666e-06,
+      "loss": 0.7739,
+      "step": 1368
+    },
+    {
+      "epoch": 0.12168888888888889,
+      "grad_norm": 29.125,
+      "learning_rate": 8.112592592592594e-06,
+      "loss": 0.6644,
+      "step": 1369
+    },
+    {
+      "epoch": 0.12177777777777778,
+      "grad_norm": 9.9375,
+      "learning_rate": 8.11851851851852e-06,
+      "loss": 0.779,
+      "step": 1370
+    },
+    {
+      "epoch": 0.12186666666666666,
+      "grad_norm": 10.5625,
+      "learning_rate": 8.124444444444445e-06,
+      "loss": 0.6236,
+      "step": 1371
+    },
+    {
+      "epoch": 0.12195555555555555,
+      "grad_norm": 11.0,
+      "learning_rate": 8.130370370370371e-06,
+      "loss": 0.6822,
+      "step": 1372
+    },
+    {
+      "epoch": 0.12204444444444444,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.136296296296297e-06,
+      "loss": 0.7178,
+      "step": 1373
+    },
+    {
+      "epoch": 0.12213333333333333,
+      "grad_norm": 7.4375,
+      "learning_rate": 8.142222222222223e-06,
+      "loss": 0.8446,
+      "step": 1374
+    },
+    {
+      "epoch": 0.12222222222222222,
+      "grad_norm": 5.34375,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.7028,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1223111111111111,
+      "grad_norm": 6.875,
+      "learning_rate": 8.154074074074074e-06,
+      "loss": 0.8365,
+      "step": 1376
+    },
+    {
+      "epoch": 0.1224,
+      "grad_norm": 7.75,
+      "learning_rate": 8.16e-06,
+      "loss": 0.7277,
+      "step": 1377
+    },
+    {
+      "epoch": 0.12248888888888888,
+      "grad_norm": 4.28125,
+      "learning_rate": 8.165925925925927e-06,
+      "loss": 0.7777,
+      "step": 1378
+    },
+    {
+      "epoch": 0.12257777777777777,
+      "grad_norm": 5.46875,
+      "learning_rate": 8.171851851851853e-06,
+      "loss": 0.8806,
+      "step": 1379
+    },
+    {
+      "epoch": 0.12266666666666666,
+      "grad_norm": 5.71875,
+      "learning_rate": 8.177777777777779e-06,
+      "loss": 0.8902,
+      "step": 1380
+    },
+    {
+      "epoch": 0.12275555555555556,
+      "grad_norm": 6.15625,
+      "learning_rate": 8.183703703703704e-06,
+      "loss": 0.5914,
+      "step": 1381
+    },
+    {
+      "epoch": 0.12284444444444445,
+      "grad_norm": 6.1875,
+      "learning_rate": 8.18962962962963e-06,
+      "loss": 0.7486,
+      "step": 1382
+    },
+    {
+      "epoch": 0.12293333333333334,
+      "grad_norm": 6.125,
+      "learning_rate": 8.195555555555556e-06,
+      "loss": 0.8102,
+      "step": 1383
+    },
+    {
+      "epoch": 0.12302222222222223,
+      "grad_norm": 6.25,
+      "learning_rate": 8.201481481481482e-06,
+      "loss": 0.8055,
+      "step": 1384
+    },
+    {
+      "epoch": 0.12311111111111112,
+      "grad_norm": 6.375,
+      "learning_rate": 8.207407407407409e-06,
+      "loss": 0.6933,
+      "step": 1385
+    },
+    {
+      "epoch": 0.1232,
+      "grad_norm": 3.9375,
+      "learning_rate": 8.213333333333335e-06,
+      "loss": 0.9121,
+      "step": 1386
+    },
+    {
+      "epoch": 0.12328888888888889,
+      "grad_norm": 5.0625,
+      "learning_rate": 8.21925925925926e-06,
+      "loss": 1.1197,
+      "step": 1387
+    },
+    {
+      "epoch": 0.12337777777777778,
+      "grad_norm": 6.15625,
+      "learning_rate": 8.225185185185186e-06,
+      "loss": 0.85,
+      "step": 1388
+    },
+    {
+      "epoch": 0.12346666666666667,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.231111111111112e-06,
+      "loss": 0.7155,
+      "step": 1389
+    },
+    {
+      "epoch": 0.12355555555555556,
+      "grad_norm": 4.40625,
+      "learning_rate": 8.237037037037038e-06,
+      "loss": 0.7878,
+      "step": 1390
+    },
+    {
+      "epoch": 0.12364444444444445,
+      "grad_norm": 6.40625,
+      "learning_rate": 8.242962962962964e-06,
+      "loss": 0.8274,
+      "step": 1391
+    },
+    {
+      "epoch": 0.12373333333333333,
+      "grad_norm": 11.5,
+      "learning_rate": 8.24888888888889e-06,
+      "loss": 0.8183,
+      "step": 1392
+    },
+    {
+      "epoch": 0.12382222222222222,
+      "grad_norm": 8.375,
+      "learning_rate": 8.254814814814815e-06,
+      "loss": 0.8736,
+      "step": 1393
+    },
+    {
+      "epoch": 0.12391111111111111,
+      "grad_norm": 6.46875,
+      "learning_rate": 8.26074074074074e-06,
+      "loss": 0.8099,
+      "step": 1394
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 5.3125,
+      "learning_rate": 8.266666666666667e-06,
+      "loss": 0.9381,
+      "step": 1395
+    },
+    {
+      "epoch": 0.12408888888888889,
+      "grad_norm": 6.40625,
+      "learning_rate": 8.272592592592594e-06,
+      "loss": 0.7539,
+      "step": 1396
+    },
+    {
+      "epoch": 0.12417777777777778,
+      "grad_norm": 7.0,
+      "learning_rate": 8.27851851851852e-06,
+      "loss": 0.8351,
+      "step": 1397
+    },
+    {
+      "epoch": 0.12426666666666666,
+      "grad_norm": 6.8125,
+      "learning_rate": 8.284444444444446e-06,
+      "loss": 1.0744,
+      "step": 1398
+    },
+    {
+      "epoch": 0.12435555555555555,
+      "grad_norm": 4.75,
+      "learning_rate": 8.290370370370371e-06,
+      "loss": 0.901,
+      "step": 1399
+    },
+    {
+      "epoch": 0.12444444444444444,
+      "grad_norm": 6.78125,
+      "learning_rate": 8.296296296296297e-06,
+      "loss": 0.7084,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12453333333333333,
+      "grad_norm": 5.875,
+      "learning_rate": 8.302222222222223e-06,
+      "loss": 0.8864,
+      "step": 1401
+    },
+    {
+      "epoch": 0.12462222222222222,
+      "grad_norm": 5.5,
+      "learning_rate": 8.308148148148148e-06,
+      "loss": 0.8054,
+      "step": 1402
+    },
+    {
+      "epoch": 0.1247111111111111,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.314074074074074e-06,
+      "loss": 0.7922,
+      "step": 1403
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 6.46875,
+      "learning_rate": 8.32e-06,
+      "loss": 0.6533,
+      "step": 1404
+    },
+    {
+      "epoch": 0.12488888888888888,
+      "grad_norm": 6.6875,
+      "learning_rate": 8.325925925925926e-06,
+      "loss": 0.7847,
+      "step": 1405
+    },
+    {
+      "epoch": 0.12497777777777777,
+      "grad_norm": 22.0,
+      "learning_rate": 8.331851851851851e-06,
+      "loss": 0.8431,
+      "step": 1406
+    },
+    {
+      "epoch": 0.12506666666666666,
+      "grad_norm": 12.625,
+      "learning_rate": 8.337777777777777e-06,
+      "loss": 0.6818,
+      "step": 1407
+    },
+    {
+      "epoch": 0.12515555555555555,
+      "grad_norm": 12.5625,
+      "learning_rate": 8.343703703703705e-06,
+      "loss": 0.8128,
+      "step": 1408
+    },
+    {
+      "epoch": 0.12524444444444444,
+      "grad_norm": 4.8125,
+      "learning_rate": 8.34962962962963e-06,
+      "loss": 0.7838,
+      "step": 1409
+    },
+    {
+      "epoch": 0.12533333333333332,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.355555555555556e-06,
+      "loss": 0.9634,
+      "step": 1410
+    },
+    {
+      "epoch": 0.1254222222222222,
+      "grad_norm": 5.46875,
+      "learning_rate": 8.361481481481482e-06,
+      "loss": 0.7863,
+      "step": 1411
+    },
+    {
+      "epoch": 0.1255111111111111,
+      "grad_norm": 5.53125,
+      "learning_rate": 8.36740740740741e-06,
+      "loss": 0.8197,
+      "step": 1412
+    },
+    {
+      "epoch": 0.1256,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.373333333333335e-06,
+      "loss": 0.955,
+      "step": 1413
+    },
+    {
+      "epoch": 0.12568888888888888,
+      "grad_norm": 7.53125,
+      "learning_rate": 8.37925925925926e-06,
+      "loss": 0.9475,
+      "step": 1414
+    },
+    {
+      "epoch": 0.12577777777777777,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.385185185185187e-06,
+      "loss": 0.7948,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12586666666666665,
+      "grad_norm": 6.0,
+      "learning_rate": 8.391111111111112e-06,
+      "loss": 0.75,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12595555555555554,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.397037037037038e-06,
+      "loss": 0.9336,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12604444444444443,
+      "grad_norm": 3.859375,
+      "learning_rate": 8.402962962962964e-06,
+      "loss": 0.9235,
+      "step": 1418
+    },
+    {
+      "epoch": 0.12613333333333332,
+      "grad_norm": 4.6875,
+      "learning_rate": 8.40888888888889e-06,
+      "loss": 0.8921,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12622222222222224,
+      "grad_norm": 5.5,
+      "learning_rate": 8.414814814814815e-06,
+      "loss": 0.7974,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12631111111111112,
+      "grad_norm": 4.375,
+      "learning_rate": 8.420740740740741e-06,
+      "loss": 0.7789,
+      "step": 1421
+    },
+    {
+      "epoch": 0.1264,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.426666666666667e-06,
+      "loss": 0.7468,
+      "step": 1422
+    },
+    {
+      "epoch": 0.1264888888888889,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.432592592592594e-06,
+      "loss": 0.8692,
+      "step": 1423
+    },
+    {
+      "epoch": 0.1265777777777778,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.43851851851852e-06,
+      "loss": 1.038,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12666666666666668,
+      "grad_norm": 5.40625,
+      "learning_rate": 8.444444444444446e-06,
+      "loss": 0.7993,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12675555555555557,
+      "grad_norm": 4.625,
+      "learning_rate": 8.450370370370371e-06,
+      "loss": 0.759,
+      "step": 1426
+    },
+    {
+      "epoch": 0.12684444444444445,
+      "grad_norm": 4.15625,
+      "learning_rate": 8.456296296296297e-06,
+      "loss": 0.8466,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12693333333333334,
+      "grad_norm": 5.71875,
+      "learning_rate": 8.462222222222223e-06,
+      "loss": 1.0064,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12702222222222223,
+      "grad_norm": 6.125,
+      "learning_rate": 8.468148148148149e-06,
+      "loss": 0.8555,
+      "step": 1429
+    },
+    {
+      "epoch": 0.12711111111111112,
+      "grad_norm": 5.59375,
+      "learning_rate": 8.474074074074074e-06,
+      "loss": 0.7991,
+      "step": 1430
+    },
+    {
+      "epoch": 0.1272,
+      "grad_norm": 4.125,
+      "learning_rate": 8.48e-06,
+      "loss": 0.989,
+      "step": 1431
+    },
+    {
+      "epoch": 0.1272888888888889,
+      "grad_norm": 6.09375,
+      "learning_rate": 8.485925925925926e-06,
+      "loss": 0.7069,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12737777777777778,
+      "grad_norm": 5.84375,
+      "learning_rate": 8.491851851851852e-06,
+      "loss": 0.8287,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12746666666666667,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.497777777777777e-06,
+      "loss": 0.7934,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12755555555555556,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.503703703703705e-06,
+      "loss": 0.5995,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12764444444444445,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.50962962962963e-06,
+      "loss": 0.6852,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12773333333333334,
+      "grad_norm": 5.34375,
+      "learning_rate": 8.515555555555556e-06,
+      "loss": 0.817,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12782222222222223,
+      "grad_norm": 7.0625,
+      "learning_rate": 8.521481481481482e-06,
+      "loss": 0.8188,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12791111111111111,
+      "grad_norm": 5.96875,
+      "learning_rate": 8.527407407407408e-06,
+      "loss": 0.698,
+      "step": 1439
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 6.875,
+      "learning_rate": 8.533333333333335e-06,
+      "loss": 0.9002,
+      "step": 1440
+    },
+    {
+      "epoch": 0.1280888888888889,
+      "grad_norm": 6.875,
+      "learning_rate": 8.539259259259261e-06,
+      "loss": 0.7785,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12817777777777778,
+      "grad_norm": 10.25,
+      "learning_rate": 8.545185185185187e-06,
+      "loss": 0.8418,
+      "step": 1442
+    },
+    {
+      "epoch": 0.12826666666666667,
+      "grad_norm": 3.5,
+      "learning_rate": 8.551111111111112e-06,
+      "loss": 0.7249,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12835555555555556,
+      "grad_norm": 4.0,
+      "learning_rate": 8.557037037037038e-06,
+      "loss": 0.7831,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12844444444444444,
+      "grad_norm": 7.21875,
+      "learning_rate": 8.562962962962964e-06,
+      "loss": 0.7326,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12853333333333333,
+      "grad_norm": 5.59375,
+      "learning_rate": 8.56888888888889e-06,
+      "loss": 0.6756,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12862222222222222,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.574814814814815e-06,
+      "loss": 0.8853,
+      "step": 1447
+    },
+    {
+      "epoch": 0.1287111111111111,
+      "grad_norm": 4.9375,
+      "learning_rate": 8.580740740740741e-06,
+      "loss": 0.847,
+      "step": 1448
+    },
+    {
+      "epoch": 0.1288,
+      "grad_norm": 3.828125,
+      "learning_rate": 8.586666666666667e-06,
+      "loss": 0.7924,
+      "step": 1449
+    },
+    {
+      "epoch": 0.1288888888888889,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.592592592592593e-06,
+      "loss": 0.7719,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12897777777777777,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.59851851851852e-06,
+      "loss": 0.722,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12906666666666666,
+      "grad_norm": 4.46875,
+      "learning_rate": 8.604444444444446e-06,
+      "loss": 0.7185,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12915555555555555,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.610370370370372e-06,
+      "loss": 0.7059,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12924444444444444,
+      "grad_norm": 8.125,
+      "learning_rate": 8.616296296296297e-06,
+      "loss": 0.7285,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12933333333333333,
+      "grad_norm": 5.4375,
+      "learning_rate": 8.622222222222223e-06,
+      "loss": 0.7066,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12942222222222222,
+      "grad_norm": 4.75,
+      "learning_rate": 8.628148148148149e-06,
+      "loss": 0.85,
+      "step": 1456
+    },
+    {
+      "epoch": 0.1295111111111111,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.634074074074075e-06,
+      "loss": 0.7463,
+      "step": 1457
+    },
+    {
+      "epoch": 0.1296,
+      "grad_norm": 3.25,
+      "learning_rate": 8.64e-06,
+      "loss": 0.8822,
+      "step": 1458
+    },
+    {
+      "epoch": 0.12968888888888888,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.645925925925926e-06,
+      "loss": 0.6991,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12977777777777777,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.651851851851852e-06,
+      "loss": 0.9137,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12986666666666666,
+      "grad_norm": 3.625,
+      "learning_rate": 8.657777777777778e-06,
+      "loss": 0.8432,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12995555555555555,
+      "grad_norm": 3.890625,
+      "learning_rate": 8.663703703703705e-06,
+      "loss": 0.7415,
+      "step": 1462
+    },
+    {
+      "epoch": 0.13004444444444443,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.66962962962963e-06,
+      "loss": 0.6009,
+      "step": 1463
+    },
+    {
+      "epoch": 0.13013333333333332,
+      "grad_norm": 5.09375,
+      "learning_rate": 8.675555555555556e-06,
+      "loss": 0.845,
+      "step": 1464
+    },
+    {
+      "epoch": 0.1302222222222222,
+      "grad_norm": 5.4375,
+      "learning_rate": 8.681481481481482e-06,
+      "loss": 0.7526,
+      "step": 1465
+    },
+    {
+      "epoch": 0.1303111111111111,
+      "grad_norm": 4.125,
+      "learning_rate": 8.687407407407408e-06,
+      "loss": 0.9874,
+      "step": 1466
+    },
+    {
+      "epoch": 0.1304,
+      "grad_norm": 4.9375,
+      "learning_rate": 8.693333333333334e-06,
+      "loss": 0.7788,
+      "step": 1467
+    },
+    {
+      "epoch": 0.13048888888888888,
+      "grad_norm": 8.375,
+      "learning_rate": 8.69925925925926e-06,
+      "loss": 0.6952,
+      "step": 1468
+    },
+    {
+      "epoch": 0.13057777777777776,
+      "grad_norm": 6.125,
+      "learning_rate": 8.705185185185185e-06,
+      "loss": 0.7767,
+      "step": 1469
+    },
+    {
+      "epoch": 0.13066666666666665,
+      "grad_norm": 8.125,
+      "learning_rate": 8.711111111111111e-06,
+      "loss": 0.8136,
+      "step": 1470
+    },
+    {
+      "epoch": 0.13075555555555557,
+      "grad_norm": 4.375,
+      "learning_rate": 8.717037037037038e-06,
+      "loss": 0.9013,
+      "step": 1471
+    },
+    {
+      "epoch": 0.13084444444444446,
+      "grad_norm": 3.5,
+      "learning_rate": 8.722962962962964e-06,
+      "loss": 0.7642,
+      "step": 1472
+    },
+    {
+      "epoch": 0.13093333333333335,
+      "grad_norm": 13.625,
+      "learning_rate": 8.72888888888889e-06,
+      "loss": 0.7549,
+      "step": 1473
+    },
+    {
+      "epoch": 0.13102222222222223,
+      "grad_norm": 6.3125,
+      "learning_rate": 8.734814814814816e-06,
+      "loss": 0.7763,
+      "step": 1474
+    },
+    {
+      "epoch": 0.13111111111111112,
+      "grad_norm": 6.3125,
+      "learning_rate": 8.740740740740741e-06,
+      "loss": 0.6656,
+      "step": 1475
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.746666666666667e-06,
+      "loss": 0.8732,
+      "step": 1476
+    },
+    {
+      "epoch": 0.1312888888888889,
+      "grad_norm": 5.78125,
+      "learning_rate": 8.752592592592593e-06,
+      "loss": 0.916,
+      "step": 1477
+    },
+    {
+      "epoch": 0.1313777777777778,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.75851851851852e-06,
+      "loss": 0.6464,
+      "step": 1478
+    },
+    {
+      "epoch": 0.13146666666666668,
+      "grad_norm": 5.46875,
+      "learning_rate": 8.764444444444446e-06,
+      "loss": 0.7317,
+      "step": 1479
+    },
+    {
+      "epoch": 0.13155555555555556,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.770370370370372e-06,
+      "loss": 0.8277,
+      "step": 1480
+    },
+    {
+      "epoch": 0.13164444444444445,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.776296296296297e-06,
+      "loss": 0.6836,
+      "step": 1481
+    },
+    {
+      "epoch": 0.13173333333333334,
+      "grad_norm": 5.5625,
+      "learning_rate": 8.782222222222223e-06,
+      "loss": 0.6896,
+      "step": 1482
+    },
+    {
+      "epoch": 0.13182222222222223,
+      "grad_norm": 4.8125,
+      "learning_rate": 8.788148148148149e-06,
+      "loss": 0.7507,
+      "step": 1483
+    },
+    {
+      "epoch": 0.13191111111111112,
+      "grad_norm": 5.3125,
+      "learning_rate": 8.794074074074075e-06,
+      "loss": 0.7122,
+      "step": 1484
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.8e-06,
+      "loss": 0.828,
+      "step": 1485
+    },
+    {
+      "epoch": 0.1320888888888889,
+      "grad_norm": 3.796875,
+      "learning_rate": 8.805925925925926e-06,
+      "loss": 0.9521,
+      "step": 1486
+    },
+    {
+      "epoch": 0.13217777777777778,
+      "grad_norm": 3.25,
+      "learning_rate": 8.811851851851852e-06,
+      "loss": 0.7329,
+      "step": 1487
+    },
+    {
+      "epoch": 0.13226666666666667,
+      "grad_norm": 6.40625,
+      "learning_rate": 8.817777777777778e-06,
+      "loss": 0.8431,
+      "step": 1488
+    },
+    {
+      "epoch": 0.13235555555555556,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.823703703703705e-06,
+      "loss": 0.6459,
+      "step": 1489
+    },
+    {
+      "epoch": 0.13244444444444445,
+      "grad_norm": 6.6875,
+      "learning_rate": 8.829629629629631e-06,
+      "loss": 0.6318,
+      "step": 1490
+    },
+    {
+      "epoch": 0.13253333333333334,
+      "grad_norm": 5.3125,
+      "learning_rate": 8.835555555555557e-06,
+      "loss": 0.7381,
+      "step": 1491
+    },
+    {
+      "epoch": 0.13262222222222222,
+      "grad_norm": 5.59375,
+      "learning_rate": 8.841481481481482e-06,
+      "loss": 0.7537,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1327111111111111,
+      "grad_norm": 4.65625,
+      "learning_rate": 8.847407407407408e-06,
+      "loss": 0.8871,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1328,
+      "grad_norm": 5.5,
+      "learning_rate": 8.853333333333334e-06,
+      "loss": 0.9171,
+      "step": 1494
+    },
+    {
+      "epoch": 0.1328888888888889,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.85925925925926e-06,
+      "loss": 0.8533,
+      "step": 1495
+    },
+    {
+      "epoch": 0.13297777777777778,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.865185185185185e-06,
+      "loss": 0.8414,
+      "step": 1496
+    },
+    {
+      "epoch": 0.13306666666666667,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.871111111111111e-06,
+      "loss": 0.9271,
+      "step": 1497
+    },
+    {
+      "epoch": 0.13315555555555555,
+      "grad_norm": 6.84375,
+      "learning_rate": 8.877037037037037e-06,
+      "loss": 0.7158,
+      "step": 1498
+    },
+    {
+      "epoch": 0.13324444444444444,
+      "grad_norm": 5.375,
+      "learning_rate": 8.882962962962963e-06,
+      "loss": 0.8504,
+      "step": 1499
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.6922,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13342222222222222,
+      "grad_norm": 4.15625,
+      "learning_rate": 8.894814814814816e-06,
+      "loss": 0.6699,
+      "step": 1501
+    },
+    {
+      "epoch": 0.1335111111111111,
+      "grad_norm": 4.5,
+      "learning_rate": 8.900740740740741e-06,
+      "loss": 0.7099,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1336,
+      "grad_norm": 5.375,
+      "learning_rate": 8.906666666666667e-06,
+      "loss": 0.8085,
+      "step": 1503
+    },
+    {
+      "epoch": 0.13368888888888888,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.912592592592593e-06,
+      "loss": 0.8424,
+      "step": 1504
+    },
+    {
+      "epoch": 0.13377777777777777,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.91851851851852e-06,
+      "loss": 0.8831,
+      "step": 1505
+    },
+    {
+      "epoch": 0.13386666666666666,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.924444444444446e-06,
+      "loss": 0.8065,
+      "step": 1506
+    },
+    {
+      "epoch": 0.13395555555555555,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.930370370370372e-06,
+      "loss": 0.6532,
+      "step": 1507
+    },
+    {
+      "epoch": 0.13404444444444444,
+      "grad_norm": 7.4375,
+      "learning_rate": 8.936296296296298e-06,
+      "loss": 0.7253,
+      "step": 1508
+    },
+    {
+      "epoch": 0.13413333333333333,
+      "grad_norm": 7.0,
+      "learning_rate": 8.942222222222223e-06,
+      "loss": 0.6877,
+      "step": 1509
+    },
+    {
+      "epoch": 0.13422222222222221,
+      "grad_norm": 5.9375,
+      "learning_rate": 8.948148148148149e-06,
+      "loss": 0.7786,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1343111111111111,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.954074074074075e-06,
+      "loss": 0.7366,
+      "step": 1511
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 4.9375,
+      "learning_rate": 8.96e-06,
+      "loss": 0.6963,
+      "step": 1512
+    },
+    {
+      "epoch": 0.13448888888888888,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.965925925925926e-06,
+      "loss": 0.8105,
+      "step": 1513
+    },
+    {
+      "epoch": 0.13457777777777777,
+      "grad_norm": 3.625,
+      "learning_rate": 8.971851851851852e-06,
+      "loss": 0.9328,
+      "step": 1514
+    },
+    {
+      "epoch": 0.13466666666666666,
+      "grad_norm": 5.96875,
+      "learning_rate": 8.977777777777778e-06,
+      "loss": 0.7627,
+      "step": 1515
+    },
+    {
+      "epoch": 0.13475555555555555,
+      "grad_norm": 3.21875,
+      "learning_rate": 8.983703703703705e-06,
+      "loss": 0.659,
+      "step": 1516
+    },
+    {
+      "epoch": 0.13484444444444443,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.989629629629631e-06,
+      "loss": 0.7357,
+      "step": 1517
+    },
+    {
+      "epoch": 0.13493333333333332,
+      "grad_norm": 2.375,
+      "learning_rate": 8.995555555555557e-06,
+      "loss": 0.7324,
+      "step": 1518
+    },
+    {
+      "epoch": 0.1350222222222222,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.001481481481483e-06,
+      "loss": 0.7532,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1351111111111111,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.007407407407408e-06,
+      "loss": 0.8675,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1352,
+      "grad_norm": 6.375,
+      "learning_rate": 9.013333333333334e-06,
+      "loss": 0.7614,
+      "step": 1521
+    },
+    {
+      "epoch": 0.13528888888888888,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.01925925925926e-06,
+      "loss": 0.5805,
+      "step": 1522
+    },
+    {
+      "epoch": 0.1353777777777778,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.025185185185185e-06,
+      "loss": 0.8683,
+      "step": 1523
+    },
+    {
+      "epoch": 0.13546666666666668,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.031111111111111e-06,
+      "loss": 0.7099,
+      "step": 1524
+    },
+    {
+      "epoch": 0.13555555555555557,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.037037037037037e-06,
+      "loss": 0.861,
+      "step": 1525
+    },
+    {
+      "epoch": 0.13564444444444446,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.042962962962963e-06,
+      "loss": 0.672,
+      "step": 1526
+    },
+    {
+      "epoch": 0.13573333333333334,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.048888888888888e-06,
+      "loss": 0.7888,
+      "step": 1527
+    },
+    {
+      "epoch": 0.13582222222222223,
+      "grad_norm": 4.625,
+      "learning_rate": 9.054814814814816e-06,
+      "loss": 0.6548,
+      "step": 1528
+    },
+    {
+      "epoch": 0.13591111111111112,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.060740740740742e-06,
+      "loss": 0.6114,
+      "step": 1529
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.066666666666667e-06,
+      "loss": 0.7872,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1360888888888889,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.072592592592593e-06,
+      "loss": 0.6641,
+      "step": 1531
+    },
+    {
+      "epoch": 0.1361777777777778,
+      "grad_norm": 6.375,
+      "learning_rate": 9.078518518518519e-06,
+      "loss": 0.8178,
+      "step": 1532
+    },
+    {
+      "epoch": 0.13626666666666667,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.084444444444446e-06,
+      "loss": 0.7842,
+      "step": 1533
+    },
+    {
+      "epoch": 0.13635555555555556,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.090370370370372e-06,
+      "loss": 0.7455,
+      "step": 1534
+    },
+    {
+      "epoch": 0.13644444444444445,
+      "grad_norm": 8.5,
+      "learning_rate": 9.096296296296298e-06,
+      "loss": 0.7259,
+      "step": 1535
+    },
+    {
+      "epoch": 0.13653333333333334,
+      "grad_norm": 3.625,
+      "learning_rate": 9.102222222222224e-06,
+      "loss": 0.8199,
+      "step": 1536
+    },
+    {
+      "epoch": 0.13662222222222223,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.10814814814815e-06,
+      "loss": 0.7669,
+      "step": 1537
+    },
+    {
+      "epoch": 0.13671111111111112,
+      "grad_norm": 4.5,
+      "learning_rate": 9.114074074074075e-06,
+      "loss": 0.6614,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1368,
+      "grad_norm": 4.5,
+      "learning_rate": 9.12e-06,
+      "loss": 0.7136,
+      "step": 1539
+    },
+    {
+      "epoch": 0.1368888888888889,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.125925925925927e-06,
+      "loss": 0.8265,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13697777777777778,
+      "grad_norm": 8.25,
+      "learning_rate": 9.131851851851852e-06,
+      "loss": 0.7495,
+      "step": 1541
+    },
+    {
+      "epoch": 0.13706666666666667,
+      "grad_norm": 10.625,
+      "learning_rate": 9.137777777777778e-06,
+      "loss": 0.5651,
+      "step": 1542
+    },
+    {
+      "epoch": 0.13715555555555556,
+      "grad_norm": 5.375,
+      "learning_rate": 9.143703703703705e-06,
+      "loss": 0.8749,
+      "step": 1543
+    },
+    {
+      "epoch": 0.13724444444444445,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.149629629629631e-06,
+      "loss": 0.7709,
+      "step": 1544
+    },
+    {
+      "epoch": 0.13733333333333334,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.155555555555557e-06,
+      "loss": 0.8744,
+      "step": 1545
+    },
+    {
+      "epoch": 0.13742222222222222,
+      "grad_norm": 14.6875,
+      "learning_rate": 9.161481481481483e-06,
+      "loss": 0.958,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1375111111111111,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.167407407407408e-06,
+      "loss": 0.615,
+      "step": 1547
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.173333333333334e-06,
+      "loss": 0.7555,
+      "step": 1548
+    },
+    {
+      "epoch": 0.1376888888888889,
+      "grad_norm": 7.65625,
+      "learning_rate": 9.17925925925926e-06,
+      "loss": 0.8189,
+      "step": 1549
+    },
+    {
+      "epoch": 0.13777777777777778,
+      "grad_norm": 6.53125,
+      "learning_rate": 9.185185185185186e-06,
+      "loss": 0.8245,
+      "step": 1550
+    },
+    {
+      "epoch": 0.13786666666666667,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.191111111111111e-06,
+      "loss": 0.8174,
+      "step": 1551
+    },
+    {
+      "epoch": 0.13795555555555555,
+      "grad_norm": 4.0,
+      "learning_rate": 9.197037037037037e-06,
+      "loss": 0.9012,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13804444444444444,
+      "grad_norm": 5.78125,
+      "learning_rate": 9.202962962962963e-06,
+      "loss": 0.76,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13813333333333333,
+      "grad_norm": 4.25,
+      "learning_rate": 9.208888888888889e-06,
+      "loss": 0.8186,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13822222222222222,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.214814814814816e-06,
+      "loss": 0.8016,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1383111111111111,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.220740740740742e-06,
+      "loss": 0.8522,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1384,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.226666666666668e-06,
+      "loss": 0.7723,
+      "step": 1557
+    },
+    {
+      "epoch": 0.13848888888888888,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.232592592592593e-06,
+      "loss": 0.8057,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13857777777777777,
+      "grad_norm": 6.8125,
+      "learning_rate": 9.238518518518519e-06,
+      "loss": 0.6628,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13866666666666666,
+      "grad_norm": 8.75,
+      "learning_rate": 9.244444444444445e-06,
+      "loss": 0.5118,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13875555555555555,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.25037037037037e-06,
+      "loss": 0.6723,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13884444444444444,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.256296296296296e-06,
+      "loss": 0.9005,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13893333333333333,
+      "grad_norm": 7.46875,
+      "learning_rate": 9.262222222222222e-06,
+      "loss": 0.6981,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1390222222222222,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.26814814814815e-06,
+      "loss": 0.7076,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1391111111111111,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.274074074074075e-06,
+      "loss": 1.0034,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1392,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.280000000000001e-06,
+      "loss": 0.7524,
+      "step": 1566
+    },
+    {
+      "epoch": 0.13928888888888888,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.285925925925927e-06,
+      "loss": 0.7901,
+      "step": 1567
+    },
+    {
+      "epoch": 0.13937777777777777,
+      "grad_norm": 7.28125,
+      "learning_rate": 9.291851851851852e-06,
+      "loss": 0.7349,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13946666666666666,
+      "grad_norm": 5.71875,
+      "learning_rate": 9.297777777777778e-06,
+      "loss": 0.9307,
+      "step": 1569
+    },
+    {
+      "epoch": 0.13955555555555554,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.303703703703706e-06,
+      "loss": 0.7269,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13964444444444443,
+      "grad_norm": 8.875,
+      "learning_rate": 9.309629629629631e-06,
+      "loss": 0.6346,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13973333333333332,
+      "grad_norm": 9.875,
+      "learning_rate": 9.315555555555557e-06,
+      "loss": 0.7779,
+      "step": 1572
+    },
+    {
+      "epoch": 0.1398222222222222,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.321481481481483e-06,
+      "loss": 0.6874,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13991111111111112,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.327407407407409e-06,
+      "loss": 0.8507,
+      "step": 1574
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.7904,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1400888888888889,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.33925925925926e-06,
+      "loss": 0.8847,
+      "step": 1576
+    },
+    {
+      "epoch": 0.1401777777777778,
+      "grad_norm": 4.25,
+      "learning_rate": 9.345185185185186e-06,
+      "loss": 0.6988,
+      "step": 1577
+    },
+    {
+      "epoch": 0.14026666666666668,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.351111111111112e-06,
+      "loss": 0.8017,
+      "step": 1578
+    },
+    {
+      "epoch": 0.14035555555555557,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.357037037037037e-06,
+      "loss": 0.7888,
+      "step": 1579
+    },
+    {
+      "epoch": 0.14044444444444446,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.362962962962963e-06,
+      "loss": 0.9292,
+      "step": 1580
+    },
+    {
+      "epoch": 0.14053333333333334,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.368888888888889e-06,
+      "loss": 0.7662,
+      "step": 1581
+    },
+    {
+      "epoch": 0.14062222222222223,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.374814814814816e-06,
+      "loss": 0.669,
+      "step": 1582
+    },
+    {
+      "epoch": 0.14071111111111112,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.380740740740742e-06,
+      "loss": 0.7666,
+      "step": 1583
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.386666666666668e-06,
+      "loss": 0.6963,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1408888888888889,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.392592592592593e-06,
+      "loss": 0.8019,
+      "step": 1585
+    },
+    {
+      "epoch": 0.14097777777777779,
+      "grad_norm": 6.53125,
+      "learning_rate": 9.39851851851852e-06,
+      "loss": 0.6553,
+      "step": 1586
+    },
+    {
+      "epoch": 0.14106666666666667,
+      "grad_norm": 7.21875,
+      "learning_rate": 9.404444444444445e-06,
+      "loss": 0.8108,
+      "step": 1587
+    },
+    {
+      "epoch": 0.14115555555555556,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.41037037037037e-06,
+      "loss": 0.7784,
+      "step": 1588
+    },
+    {
+      "epoch": 0.14124444444444445,
+      "grad_norm": 4.0,
+      "learning_rate": 9.416296296296296e-06,
+      "loss": 0.8524,
+      "step": 1589
+    },
+    {
+      "epoch": 0.14133333333333334,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.422222222222222e-06,
+      "loss": 0.7443,
+      "step": 1590
+    },
+    {
+      "epoch": 0.14142222222222223,
+      "grad_norm": 6.125,
+      "learning_rate": 9.428148148148148e-06,
+      "loss": 0.7194,
+      "step": 1591
+    },
+    {
+      "epoch": 0.14151111111111112,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.434074074074074e-06,
+      "loss": 0.7349,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1416,
+      "grad_norm": 5.78125,
+      "learning_rate": 9.440000000000001e-06,
+      "loss": 0.7504,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1416888888888889,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.445925925925927e-06,
+      "loss": 0.8087,
+      "step": 1594
+    },
+    {
+      "epoch": 0.14177777777777778,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.451851851851853e-06,
+      "loss": 0.883,
+      "step": 1595
+    },
+    {
+      "epoch": 0.14186666666666667,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.457777777777778e-06,
+      "loss": 0.6453,
+      "step": 1596
+    },
+    {
+      "epoch": 0.14195555555555556,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.463703703703706e-06,
+      "loss": 0.7482,
+      "step": 1597
+    },
+    {
+      "epoch": 0.14204444444444445,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.469629629629632e-06,
+      "loss": 0.7364,
+      "step": 1598
+    },
+    {
+      "epoch": 0.14213333333333333,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.475555555555557e-06,
+      "loss": 0.8343,
+      "step": 1599
+    },
+    {
+      "epoch": 0.14222222222222222,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.481481481481483e-06,
+      "loss": 0.679,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1423111111111111,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.487407407407409e-06,
+      "loss": 0.688,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1424,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.493333333333334e-06,
+      "loss": 0.7233,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1424888888888889,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.49925925925926e-06,
+      "loss": 0.7972,
+      "step": 1603
+    },
+    {
+      "epoch": 0.14257777777777778,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.505185185185186e-06,
+      "loss": 0.7901,
+      "step": 1604
+    },
+    {
+      "epoch": 0.14266666666666666,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.511111111111112e-06,
+      "loss": 0.6012,
+      "step": 1605
+    },
+    {
+      "epoch": 0.14275555555555555,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.517037037037037e-06,
+      "loss": 0.7178,
+      "step": 1606
+    },
+    {
+      "epoch": 0.14284444444444444,
+      "grad_norm": 5.8125,
+      "learning_rate": 9.522962962962963e-06,
+      "loss": 0.7855,
+      "step": 1607
+    },
+    {
+      "epoch": 0.14293333333333333,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.528888888888889e-06,
+      "loss": 0.7047,
+      "step": 1608
+    },
+    {
+      "epoch": 0.14302222222222222,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.534814814814816e-06,
+      "loss": 0.8049,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1431111111111111,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.540740740740742e-06,
+      "loss": 0.7721,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1432,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.546666666666668e-06,
+      "loss": 0.7959,
+      "step": 1611
+    },
+    {
+      "epoch": 0.14328888888888888,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.552592592592594e-06,
+      "loss": 0.6168,
+      "step": 1612
+    },
+    {
+      "epoch": 0.14337777777777777,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.55851851851852e-06,
+      "loss": 0.7532,
+      "step": 1613
+    },
+    {
+      "epoch": 0.14346666666666666,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.564444444444445e-06,
+      "loss": 0.6836,
+      "step": 1614
+    },
+    {
+      "epoch": 0.14355555555555555,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.570370370370371e-06,
+      "loss": 0.8073,
+      "step": 1615
+    },
+    {
+      "epoch": 0.14364444444444444,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.576296296296297e-06,
+      "loss": 0.7168,
+      "step": 1616
+    },
+    {
+      "epoch": 0.14373333333333332,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.582222222222222e-06,
+      "loss": 0.998,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1438222222222222,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.588148148148148e-06,
+      "loss": 0.8352,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1439111111111111,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.594074074074074e-06,
+      "loss": 0.7926,
+      "step": 1619
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.7456,
+      "step": 1620
+    },
+    {
+      "epoch": 0.14408888888888888,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.605925925925927e-06,
+      "loss": 0.8201,
+      "step": 1621
+    },
+    {
+      "epoch": 0.14417777777777777,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.611851851851853e-06,
+      "loss": 0.7872,
+      "step": 1622
+    },
+    {
+      "epoch": 0.14426666666666665,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.617777777777778e-06,
+      "loss": 0.8034,
+      "step": 1623
+    },
+    {
+      "epoch": 0.14435555555555554,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.623703703703704e-06,
+      "loss": 0.8028,
+      "step": 1624
+    },
+    {
+      "epoch": 0.14444444444444443,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.6899,
+      "step": 1625
+    },
+    {
+      "epoch": 0.14453333333333335,
+      "grad_norm": 5.75,
+      "learning_rate": 9.635555555555557e-06,
+      "loss": 0.7391,
+      "step": 1626
+    },
+    {
+      "epoch": 0.14462222222222224,
+      "grad_norm": 5.875,
+      "learning_rate": 9.641481481481483e-06,
+      "loss": 0.684,
+      "step": 1627
+    },
+    {
+      "epoch": 0.14471111111111112,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.647407407407409e-06,
+      "loss": 0.8181,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1448,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.653333333333335e-06,
+      "loss": 0.7883,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1448888888888889,
+      "grad_norm": 3.0,
+      "learning_rate": 9.65925925925926e-06,
+      "loss": 0.8272,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1449777777777778,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.665185185185186e-06,
+      "loss": 0.7674,
+      "step": 1631
+    },
+    {
+      "epoch": 0.14506666666666668,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.671111111111112e-06,
+      "loss": 0.7116,
+      "step": 1632
+    },
+    {
+      "epoch": 0.14515555555555557,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.677037037037038e-06,
+      "loss": 0.7902,
+      "step": 1633
+    },
+    {
+      "epoch": 0.14524444444444445,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.682962962962963e-06,
+      "loss": 0.713,
+      "step": 1634
+    },
+    {
+      "epoch": 0.14533333333333334,
+      "grad_norm": 3.875,
+      "learning_rate": 9.688888888888889e-06,
+      "loss": 0.7146,
+      "step": 1635
+    },
+    {
+      "epoch": 0.14542222222222223,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.694814814814817e-06,
+      "loss": 0.779,
+      "step": 1636
+    },
+    {
+      "epoch": 0.14551111111111112,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.700740740740742e-06,
+      "loss": 0.6502,
+      "step": 1637
+    },
+    {
+      "epoch": 0.1456,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.706666666666668e-06,
+      "loss": 0.6987,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1456888888888889,
+      "grad_norm": 5.625,
+      "learning_rate": 9.712592592592594e-06,
+      "loss": 0.9131,
+      "step": 1639
+    },
+    {
+      "epoch": 0.14577777777777778,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.71851851851852e-06,
+      "loss": 0.7467,
+      "step": 1640
+    },
+    {
+      "epoch": 0.14586666666666667,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.724444444444445e-06,
+      "loss": 0.6565,
+      "step": 1641
+    },
+    {
+      "epoch": 0.14595555555555556,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.730370370370371e-06,
+      "loss": 0.7589,
+      "step": 1642
+    },
+    {
+      "epoch": 0.14604444444444445,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.736296296296297e-06,
+      "loss": 0.7783,
+      "step": 1643
+    },
+    {
+      "epoch": 0.14613333333333334,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.742222222222222e-06,
+      "loss": 0.7171,
+      "step": 1644
+    },
+    {
+      "epoch": 0.14622222222222223,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.748148148148148e-06,
+      "loss": 0.8249,
+      "step": 1645
+    },
+    {
+      "epoch": 0.14631111111111111,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.754074074074074e-06,
+      "loss": 0.8164,
+      "step": 1646
+    },
+    {
+      "epoch": 0.1464,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.760000000000001e-06,
+      "loss": 0.5781,
+      "step": 1647
+    },
+    {
+      "epoch": 0.1464888888888889,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.765925925925927e-06,
+      "loss": 0.7683,
+      "step": 1648
+    },
+    {
+      "epoch": 0.14657777777777778,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.771851851851853e-06,
+      "loss": 0.7717,
+      "step": 1649
+    },
+    {
+      "epoch": 0.14666666666666667,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 0.8227,
+      "step": 1650
+    },
+    {
+      "epoch": 0.14675555555555556,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.783703703703704e-06,
+      "loss": 0.842,
+      "step": 1651
+    },
+    {
+      "epoch": 0.14684444444444444,
+      "grad_norm": 3.75,
+      "learning_rate": 9.78962962962963e-06,
+      "loss": 0.6487,
+      "step": 1652
+    },
+    {
+      "epoch": 0.14693333333333333,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.795555555555556e-06,
+      "loss": 0.8168,
+      "step": 1653
+    },
+    {
+      "epoch": 0.14702222222222222,
+      "grad_norm": 5.0,
+      "learning_rate": 9.801481481481482e-06,
+      "loss": 0.7469,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1471111111111111,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.807407407407407e-06,
+      "loss": 0.8849,
+      "step": 1655
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.813333333333333e-06,
+      "loss": 0.6741,
+      "step": 1656
+    },
+    {
+      "epoch": 0.14728888888888889,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.81925925925926e-06,
+      "loss": 0.6945,
+      "step": 1657
+    },
+    {
+      "epoch": 0.14737777777777777,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.825185185185186e-06,
+      "loss": 0.673,
+      "step": 1658
+    },
+    {
+      "epoch": 0.14746666666666666,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.831111111111112e-06,
+      "loss": 0.7909,
+      "step": 1659
+    },
+    {
+      "epoch": 0.14755555555555555,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.837037037037038e-06,
+      "loss": 0.673,
+      "step": 1660
+    },
+    {
+      "epoch": 0.14764444444444444,
+      "grad_norm": 2.75,
+      "learning_rate": 9.842962962962964e-06,
+      "loss": 0.7362,
+      "step": 1661
+    },
+    {
+      "epoch": 0.14773333333333333,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.84888888888889e-06,
+      "loss": 1.031,
+      "step": 1662
+    },
+    {
+      "epoch": 0.14782222222222222,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.854814814814817e-06,
+      "loss": 0.9457,
+      "step": 1663
+    },
+    {
+      "epoch": 0.1479111111111111,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.860740740740742e-06,
+      "loss": 0.7299,
+      "step": 1664
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 3.375,
+      "learning_rate": 9.866666666666668e-06,
+      "loss": 0.6971,
+      "step": 1665
+    },
+    {
+      "epoch": 0.14808888888888888,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.872592592592594e-06,
+      "loss": 0.704,
+      "step": 1666
+    },
+    {
+      "epoch": 0.14817777777777777,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.87851851851852e-06,
+      "loss": 0.8684,
+      "step": 1667
+    },
+    {
+      "epoch": 0.14826666666666666,
+      "grad_norm": 3.25,
+      "learning_rate": 9.884444444444445e-06,
+      "loss": 0.8124,
+      "step": 1668
+    },
+    {
+      "epoch": 0.14835555555555555,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.890370370370371e-06,
+      "loss": 0.73,
+      "step": 1669
+    },
+    {
+      "epoch": 0.14844444444444443,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.896296296296297e-06,
+      "loss": 0.8538,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14853333333333332,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.902222222222223e-06,
+      "loss": 0.7255,
+      "step": 1671
+    },
+    {
+      "epoch": 0.1486222222222222,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.908148148148148e-06,
+      "loss": 0.7967,
+      "step": 1672
+    },
+    {
+      "epoch": 0.1487111111111111,
+      "grad_norm": 3.75,
+      "learning_rate": 9.914074074074074e-06,
+      "loss": 0.7862,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1488,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.920000000000002e-06,
+      "loss": 0.825,
+      "step": 1674
+    },
+    {
+      "epoch": 0.14888888888888888,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.925925925925927e-06,
+      "loss": 0.6566,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14897777777777776,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.931851851851853e-06,
+      "loss": 0.7165,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14906666666666665,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.937777777777779e-06,
+      "loss": 0.7416,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14915555555555557,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.943703703703705e-06,
+      "loss": 0.6813,
+      "step": 1678
+    },
+    {
+      "epoch": 0.14924444444444446,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.94962962962963e-06,
+      "loss": 0.6014,
+      "step": 1679
+    },
+    {
+      "epoch": 0.14933333333333335,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.955555555555556e-06,
+      "loss": 0.783,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14942222222222223,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.961481481481482e-06,
+      "loss": 0.7657,
+      "step": 1681
+    },
+    {
+      "epoch": 0.14951111111111112,
+      "grad_norm": 4.25,
+      "learning_rate": 9.967407407407408e-06,
+      "loss": 0.7342,
+      "step": 1682
+    },
+    {
+      "epoch": 0.1496,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.973333333333333e-06,
+      "loss": 0.7946,
+      "step": 1683
+    },
+    {
+      "epoch": 0.1496888888888889,
+      "grad_norm": 3.5,
+      "learning_rate": 9.979259259259259e-06,
+      "loss": 0.6433,
+      "step": 1684
+    },
+    {
+      "epoch": 0.1497777777777778,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.985185185185185e-06,
+      "loss": 0.8824,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14986666666666668,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.991111111111112e-06,
+      "loss": 0.7308,
+      "step": 1686
+    },
+    {
+      "epoch": 0.14995555555555556,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.997037037037038e-06,
+      "loss": 0.8611,
+      "step": 1687
+    },
+    {
+      "epoch": 0.15004444444444445,
+      "grad_norm": 4.5,
+      "learning_rate": 1.0002962962962964e-05,
+      "loss": 0.7534,
+      "step": 1688
+    },
+    {
+      "epoch": 0.15013333333333334,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.000888888888889e-05,
+      "loss": 0.7346,
+      "step": 1689
+    },
+    {
+      "epoch": 0.15022222222222223,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.0014814814814815e-05,
+      "loss": 0.6456,
+      "step": 1690
+    },
+    {
+      "epoch": 0.15031111111111112,
+      "grad_norm": 3.75,
+      "learning_rate": 1.0020740740740743e-05,
+      "loss": 0.7262,
+      "step": 1691
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.0026666666666667e-05,
+      "loss": 0.57,
+      "step": 1692
+    },
+    {
+      "epoch": 0.1504888888888889,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.0032592592592594e-05,
+      "loss": 0.7864,
+      "step": 1693
+    },
+    {
+      "epoch": 0.15057777777777778,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.0038518518518518e-05,
+      "loss": 0.7271,
+      "step": 1694
+    },
+    {
+      "epoch": 0.15066666666666667,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.0044444444444446e-05,
+      "loss": 0.9409,
+      "step": 1695
+    },
+    {
+      "epoch": 0.15075555555555556,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.005037037037037e-05,
+      "loss": 0.8279,
+      "step": 1696
+    },
+    {
+      "epoch": 0.15084444444444445,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.0056296296296297e-05,
+      "loss": 0.6442,
+      "step": 1697
+    },
+    {
+      "epoch": 0.15093333333333334,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.0062222222222223e-05,
+      "loss": 0.7818,
+      "step": 1698
+    },
+    {
+      "epoch": 0.15102222222222222,
+      "grad_norm": 2.125,
+      "learning_rate": 1.0068148148148149e-05,
+      "loss": 0.7192,
+      "step": 1699
+    },
+    {
+      "epoch": 0.1511111111111111,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.0074074074074074e-05,
+      "loss": 0.6755,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1512,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.008e-05,
+      "loss": 0.8746,
+      "step": 1701
+    },
+    {
+      "epoch": 0.1512888888888889,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.0085925925925926e-05,
+      "loss": 0.9031,
+      "step": 1702
+    },
+    {
+      "epoch": 0.15137777777777778,
+      "grad_norm": 5.15625,
+      "learning_rate": 1.0091851851851853e-05,
+      "loss": 0.9105,
+      "step": 1703
+    },
+    {
+      "epoch": 0.15146666666666667,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.0097777777777779e-05,
+      "loss": 0.7264,
+      "step": 1704
+    },
+    {
+      "epoch": 0.15155555555555555,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.0103703703703705e-05,
+      "loss": 0.7961,
+      "step": 1705
+    },
+    {
+      "epoch": 0.15164444444444444,
+      "grad_norm": 5.40625,
+      "learning_rate": 1.0109629629629632e-05,
+      "loss": 0.7619,
+      "step": 1706
+    },
+    {
+      "epoch": 0.15173333333333333,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.0115555555555556e-05,
+      "loss": 0.552,
+      "step": 1707
+    },
+    {
+      "epoch": 0.15182222222222222,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.0121481481481484e-05,
+      "loss": 0.671,
+      "step": 1708
+    },
+    {
+      "epoch": 0.1519111111111111,
+      "grad_norm": 2.25,
+      "learning_rate": 1.0127407407407408e-05,
+      "loss": 0.8181,
+      "step": 1709
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 4.53125,
+      "learning_rate": 1.0133333333333335e-05,
+      "loss": 0.7452,
+      "step": 1710
+    },
+    {
+      "epoch": 0.15208888888888888,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.013925925925926e-05,
+      "loss": 0.7225,
+      "step": 1711
+    },
+    {
+      "epoch": 0.15217777777777777,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.0145185185185187e-05,
+      "loss": 0.8622,
+      "step": 1712
+    },
+    {
+      "epoch": 0.15226666666666666,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.0151111111111112e-05,
+      "loss": 0.7069,
+      "step": 1713
+    },
+    {
+      "epoch": 0.15235555555555555,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.0157037037037038e-05,
+      "loss": 0.7656,
+      "step": 1714
+    },
+    {
+      "epoch": 0.15244444444444444,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.0162962962962964e-05,
+      "loss": 0.7002,
+      "step": 1715
+    },
+    {
+      "epoch": 0.15253333333333333,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.016888888888889e-05,
+      "loss": 0.8045,
+      "step": 1716
+    },
+    {
+      "epoch": 0.15262222222222221,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.0174814814814815e-05,
+      "loss": 0.7145,
+      "step": 1717
+    },
+    {
+      "epoch": 0.1527111111111111,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.0180740740740743e-05,
+      "loss": 0.8503,
+      "step": 1718
+    },
+    {
+      "epoch": 0.1528,
+      "grad_norm": 2.5,
+      "learning_rate": 1.0186666666666667e-05,
+      "loss": 0.7853,
+      "step": 1719
+    },
+    {
+      "epoch": 0.15288888888888888,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.0192592592592594e-05,
+      "loss": 0.7126,
+      "step": 1720
+    },
+    {
+      "epoch": 0.15297777777777777,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.0198518518518518e-05,
+      "loss": 0.8294,
+      "step": 1721
+    },
+    {
+      "epoch": 0.15306666666666666,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.0204444444444446e-05,
+      "loss": 0.6747,
+      "step": 1722
+    },
+    {
+      "epoch": 0.15315555555555554,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.021037037037037e-05,
+      "loss": 0.9036,
+      "step": 1723
+    },
+    {
+      "epoch": 0.15324444444444443,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.0216296296296297e-05,
+      "loss": 0.8673,
+      "step": 1724
+    },
+    {
+      "epoch": 0.15333333333333332,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.0222222222222223e-05,
+      "loss": 0.7601,
+      "step": 1725
+    },
+    {
+      "epoch": 0.1534222222222222,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.0228148148148149e-05,
+      "loss": 0.6036,
+      "step": 1726
+    },
+    {
+      "epoch": 0.1535111111111111,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.0234074074074074e-05,
+      "loss": 0.7267,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.024e-05,
+      "loss": 0.6909,
+      "step": 1728
+    },
+    {
+      "epoch": 0.1536888888888889,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.0245925925925926e-05,
+      "loss": 0.659,
+      "step": 1729
+    },
+    {
+      "epoch": 0.1537777777777778,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.0251851851851853e-05,
+      "loss": 0.8074,
+      "step": 1730
+    },
+    {
+      "epoch": 0.15386666666666668,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.0257777777777777e-05,
+      "loss": 0.5704,
+      "step": 1731
+    },
+    {
+      "epoch": 0.15395555555555557,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.0263703703703705e-05,
+      "loss": 0.7046,
+      "step": 1732
+    },
+    {
+      "epoch": 0.15404444444444446,
+      "grad_norm": 2.375,
+      "learning_rate": 1.0269629629629629e-05,
+      "loss": 0.5703,
+      "step": 1733
+    },
+    {
+      "epoch": 0.15413333333333334,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0275555555555556e-05,
+      "loss": 0.7596,
+      "step": 1734
+    },
+    {
+      "epoch": 0.15422222222222223,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.0281481481481484e-05,
+      "loss": 0.6832,
+      "step": 1735
+    },
+    {
+      "epoch": 0.15431111111111112,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.0287407407407408e-05,
+      "loss": 0.6784,
+      "step": 1736
+    },
+    {
+      "epoch": 0.1544,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.0293333333333335e-05,
+      "loss": 0.7871,
+      "step": 1737
+    },
+    {
+      "epoch": 0.1544888888888889,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.029925925925926e-05,
+      "loss": 0.8033,
+      "step": 1738
+    },
+    {
+      "epoch": 0.1545777777777778,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.0305185185185187e-05,
+      "loss": 0.9218,
+      "step": 1739
+    },
+    {
+      "epoch": 0.15466666666666667,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.0311111111111113e-05,
+      "loss": 0.6645,
+      "step": 1740
+    },
+    {
+      "epoch": 0.15475555555555556,
+      "grad_norm": 5.71875,
+      "learning_rate": 1.0317037037037038e-05,
+      "loss": 0.7112,
+      "step": 1741
+    },
+    {
+      "epoch": 0.15484444444444445,
+      "grad_norm": 6.875,
+      "learning_rate": 1.0322962962962964e-05,
+      "loss": 0.8668,
+      "step": 1742
+    },
+    {
+      "epoch": 0.15493333333333334,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.032888888888889e-05,
+      "loss": 0.7538,
+      "step": 1743
+    },
+    {
+      "epoch": 0.15502222222222223,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.0334814814814816e-05,
+      "loss": 0.8147,
+      "step": 1744
+    },
+    {
+      "epoch": 0.15511111111111112,
+      "grad_norm": 3.25,
+      "learning_rate": 1.0340740740740743e-05,
+      "loss": 0.7203,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1552,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.0346666666666667e-05,
+      "loss": 0.791,
+      "step": 1746
+    },
+    {
+      "epoch": 0.1552888888888889,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.0352592592592594e-05,
+      "loss": 0.868,
+      "step": 1747
+    },
+    {
+      "epoch": 0.15537777777777778,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.0358518518518518e-05,
+      "loss": 0.6926,
+      "step": 1748
+    },
+    {
+      "epoch": 0.15546666666666667,
+      "grad_norm": 6.125,
+      "learning_rate": 1.0364444444444446e-05,
+      "loss": 0.6318,
+      "step": 1749
+    },
+    {
+      "epoch": 0.15555555555555556,
+      "grad_norm": 5.5,
+      "learning_rate": 1.037037037037037e-05,
+      "loss": 0.7863,
+      "step": 1750
+    },
+    {
+      "epoch": 0.15564444444444445,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.0376296296296297e-05,
+      "loss": 0.6637,
+      "step": 1751
+    },
+    {
+      "epoch": 0.15573333333333333,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.0382222222222223e-05,
+      "loss": 0.7641,
+      "step": 1752
+    },
+    {
+      "epoch": 0.15582222222222222,
+      "grad_norm": 5.25,
+      "learning_rate": 1.0388148148148149e-05,
+      "loss": 0.6834,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1559111111111111,
+      "grad_norm": 7.125,
+      "learning_rate": 1.0394074074074075e-05,
+      "loss": 0.781,
+      "step": 1754
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 5.15625,
+      "learning_rate": 1.04e-05,
+      "loss": 0.733,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1560888888888889,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.0405925925925926e-05,
+      "loss": 0.5979,
+      "step": 1756
+    },
+    {
+      "epoch": 0.15617777777777778,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.0411851851851854e-05,
+      "loss": 0.5918,
+      "step": 1757
+    },
+    {
+      "epoch": 0.15626666666666666,
+      "grad_norm": 3.5,
+      "learning_rate": 1.0417777777777778e-05,
+      "loss": 0.6635,
+      "step": 1758
+    },
+    {
+      "epoch": 0.15635555555555555,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.0423703703703705e-05,
+      "loss": 0.753,
+      "step": 1759
+    },
+    {
+      "epoch": 0.15644444444444444,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.0429629629629629e-05,
+      "loss": 0.8238,
+      "step": 1760
+    },
+    {
+      "epoch": 0.15653333333333333,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.0435555555555557e-05,
+      "loss": 0.742,
+      "step": 1761
+    },
+    {
+      "epoch": 0.15662222222222222,
+      "grad_norm": 2.75,
+      "learning_rate": 1.044148148148148e-05,
+      "loss": 0.607,
+      "step": 1762
+    },
+    {
+      "epoch": 0.1567111111111111,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.0447407407407408e-05,
+      "loss": 0.7209,
+      "step": 1763
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.0453333333333334e-05,
+      "loss": 0.8241,
+      "step": 1764
+    },
+    {
+      "epoch": 0.15688888888888888,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.045925925925926e-05,
+      "loss": 0.7169,
+      "step": 1765
+    },
+    {
+      "epoch": 0.15697777777777777,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.0465185185185187e-05,
+      "loss": 0.7454,
+      "step": 1766
+    },
+    {
+      "epoch": 0.15706666666666666,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.0471111111111113e-05,
+      "loss": 0.6747,
+      "step": 1767
+    },
+    {
+      "epoch": 0.15715555555555555,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.0477037037037038e-05,
+      "loss": 0.8164,
+      "step": 1768
+    },
+    {
+      "epoch": 0.15724444444444444,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.0482962962962964e-05,
+      "loss": 0.7371,
+      "step": 1769
+    },
+    {
+      "epoch": 0.15733333333333333,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.048888888888889e-05,
+      "loss": 0.7606,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1574222222222222,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.0494814814814816e-05,
+      "loss": 0.883,
+      "step": 1771
+    },
+    {
+      "epoch": 0.1575111111111111,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.0500740740740743e-05,
+      "loss": 0.68,
+      "step": 1772
+    },
+    {
+      "epoch": 0.1576,
+      "grad_norm": 3.125,
+      "learning_rate": 1.0506666666666667e-05,
+      "loss": 0.9207,
+      "step": 1773
+    },
+    {
+      "epoch": 0.15768888888888888,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0512592592592595e-05,
+      "loss": 0.6703,
+      "step": 1774
+    },
+    {
+      "epoch": 0.15777777777777777,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.0518518518518519e-05,
+      "loss": 0.7338,
+      "step": 1775
+    },
+    {
+      "epoch": 0.15786666666666666,
+      "grad_norm": 3.0,
+      "learning_rate": 1.0524444444444446e-05,
+      "loss": 0.7904,
+      "step": 1776
+    },
+    {
+      "epoch": 0.15795555555555554,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.053037037037037e-05,
+      "loss": 0.6436,
+      "step": 1777
+    },
+    {
+      "epoch": 0.15804444444444443,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.0536296296296298e-05,
+      "loss": 0.658,
+      "step": 1778
+    },
+    {
+      "epoch": 0.15813333333333332,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.0542222222222223e-05,
+      "loss": 0.5237,
+      "step": 1779
+    },
+    {
+      "epoch": 0.1582222222222222,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.0548148148148149e-05,
+      "loss": 0.6674,
+      "step": 1780
+    },
+    {
+      "epoch": 0.15831111111111112,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.0554074074074075e-05,
+      "loss": 0.8016,
+      "step": 1781
+    },
+    {
+      "epoch": 0.1584,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.056e-05,
+      "loss": 0.5742,
+      "step": 1782
+    },
+    {
+      "epoch": 0.1584888888888889,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.0565925925925926e-05,
+      "loss": 0.6906,
+      "step": 1783
+    },
+    {
+      "epoch": 0.1585777777777778,
+      "grad_norm": 2.875,
+      "learning_rate": 1.0571851851851854e-05,
+      "loss": 0.675,
+      "step": 1784
+    },
+    {
+      "epoch": 0.15866666666666668,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.0577777777777778e-05,
+      "loss": 0.8251,
+      "step": 1785
+    },
+    {
+      "epoch": 0.15875555555555557,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.0583703703703705e-05,
+      "loss": 0.8081,
+      "step": 1786
+    },
+    {
+      "epoch": 0.15884444444444445,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.058962962962963e-05,
+      "loss": 0.8168,
+      "step": 1787
+    },
+    {
+      "epoch": 0.15893333333333334,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.0595555555555557e-05,
+      "loss": 0.6035,
+      "step": 1788
+    },
+    {
+      "epoch": 0.15902222222222223,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.060148148148148e-05,
+      "loss": 0.8018,
+      "step": 1789
+    },
+    {
+      "epoch": 0.15911111111111112,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.0607407407407408e-05,
+      "loss": 0.7484,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1592,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.0613333333333334e-05,
+      "loss": 0.7052,
+      "step": 1791
+    },
+    {
+      "epoch": 0.1592888888888889,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.061925925925926e-05,
+      "loss": 0.7408,
+      "step": 1792
+    },
+    {
+      "epoch": 0.15937777777777778,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.0625185185185185e-05,
+      "loss": 0.7783,
+      "step": 1793
+    },
+    {
+      "epoch": 0.15946666666666667,
+      "grad_norm": 3.0,
+      "learning_rate": 1.0631111111111113e-05,
+      "loss": 0.6797,
+      "step": 1794
+    },
+    {
+      "epoch": 0.15955555555555556,
+      "grad_norm": 4.375,
+      "learning_rate": 1.0637037037037037e-05,
+      "loss": 0.6319,
+      "step": 1795
+    },
+    {
+      "epoch": 0.15964444444444445,
+      "grad_norm": 4.71875,
+      "learning_rate": 1.0642962962962964e-05,
+      "loss": 0.5485,
+      "step": 1796
+    },
+    {
+      "epoch": 0.15973333333333334,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.064888888888889e-05,
+      "loss": 0.6699,
+      "step": 1797
+    },
+    {
+      "epoch": 0.15982222222222223,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.0654814814814816e-05,
+      "loss": 0.4824,
+      "step": 1798
+    },
+    {
+      "epoch": 0.15991111111111111,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.0660740740740743e-05,
+      "loss": 0.7749,
+      "step": 1799
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 0.7506,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1600888888888889,
+      "grad_norm": 2.875,
+      "learning_rate": 1.0672592592592595e-05,
+      "loss": 0.8534,
+      "step": 1801
+    },
+    {
+      "epoch": 0.16017777777777778,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.0678518518518519e-05,
+      "loss": 0.6544,
+      "step": 1802
+    },
+    {
+      "epoch": 0.16026666666666667,
+      "grad_norm": 3.0,
+      "learning_rate": 1.0684444444444446e-05,
+      "loss": 0.8301,
+      "step": 1803
+    },
+    {
+      "epoch": 0.16035555555555556,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.069037037037037e-05,
+      "loss": 0.7839,
+      "step": 1804
+    },
+    {
+      "epoch": 0.16044444444444445,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.0696296296296298e-05,
+      "loss": 0.5878,
+      "step": 1805
+    },
+    {
+      "epoch": 0.16053333333333333,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0702222222222223e-05,
+      "loss": 0.837,
+      "step": 1806
+    },
+    {
+      "epoch": 0.16062222222222222,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.070814814814815e-05,
+      "loss": 0.7531,
+      "step": 1807
+    },
+    {
+      "epoch": 0.1607111111111111,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.0714074074074075e-05,
+      "loss": 0.722,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1608,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.072e-05,
+      "loss": 0.8268,
+      "step": 1809
+    },
+    {
+      "epoch": 0.1608888888888889,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.0725925925925926e-05,
+      "loss": 0.7601,
+      "step": 1810
+    },
+    {
+      "epoch": 0.16097777777777778,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0731851851851854e-05,
+      "loss": 0.735,
+      "step": 1811
+    },
+    {
+      "epoch": 0.16106666666666666,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.0737777777777778e-05,
+      "loss": 0.7806,
+      "step": 1812
+    },
+    {
+      "epoch": 0.16115555555555555,
+      "grad_norm": 4.5,
+      "learning_rate": 1.0743703703703705e-05,
+      "loss": 0.6439,
+      "step": 1813
+    },
+    {
+      "epoch": 0.16124444444444444,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.074962962962963e-05,
+      "loss": 0.8079,
+      "step": 1814
+    },
+    {
+      "epoch": 0.16133333333333333,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.0755555555555557e-05,
+      "loss": 0.5211,
+      "step": 1815
+    },
+    {
+      "epoch": 0.16142222222222222,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.0761481481481481e-05,
+      "loss": 0.7321,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1615111111111111,
+      "grad_norm": 2.875,
+      "learning_rate": 1.0767407407407408e-05,
+      "loss": 0.654,
+      "step": 1817
+    },
+    {
+      "epoch": 0.1616,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.0773333333333334e-05,
+      "loss": 0.7571,
+      "step": 1818
+    },
+    {
+      "epoch": 0.16168888888888888,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.077925925925926e-05,
+      "loss": 0.6469,
+      "step": 1819
+    },
+    {
+      "epoch": 0.16177777777777777,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.0785185185185186e-05,
+      "loss": 0.7068,
+      "step": 1820
+    },
+    {
+      "epoch": 0.16186666666666666,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.0791111111111113e-05,
+      "loss": 0.7009,
+      "step": 1821
+    },
+    {
+      "epoch": 0.16195555555555555,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0797037037037037e-05,
+      "loss": 0.7075,
+      "step": 1822
+    },
+    {
+      "epoch": 0.16204444444444444,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.0802962962962965e-05,
+      "loss": 0.7907,
+      "step": 1823
+    },
+    {
+      "epoch": 0.16213333333333332,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.0808888888888889e-05,
+      "loss": 0.7529,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1622222222222222,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.0814814814814816e-05,
+      "loss": 0.7559,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1623111111111111,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.082074074074074e-05,
+      "loss": 0.6665,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1624,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.0826666666666667e-05,
+      "loss": 0.6955,
+      "step": 1827
+    },
+    {
+      "epoch": 0.16248888888888888,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.0832592592592595e-05,
+      "loss": 0.6752,
+      "step": 1828
+    },
+    {
+      "epoch": 0.16257777777777777,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.0838518518518519e-05,
+      "loss": 0.8722,
+      "step": 1829
+    },
+    {
+      "epoch": 0.16266666666666665,
+      "grad_norm": 2.625,
+      "learning_rate": 1.0844444444444446e-05,
+      "loss": 0.585,
+      "step": 1830
+    },
+    {
+      "epoch": 0.16275555555555554,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.085037037037037e-05,
+      "loss": 0.7412,
+      "step": 1831
+    },
+    {
+      "epoch": 0.16284444444444446,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.0856296296296298e-05,
+      "loss": 0.652,
+      "step": 1832
+    },
+    {
+      "epoch": 0.16293333333333335,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.0862222222222224e-05,
+      "loss": 0.8299,
+      "step": 1833
+    },
+    {
+      "epoch": 0.16302222222222224,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.086814814814815e-05,
+      "loss": 0.9755,
+      "step": 1834
+    },
+    {
+      "epoch": 0.16311111111111112,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0874074074074075e-05,
+      "loss": 0.6988,
+      "step": 1835
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 2.75,
+      "learning_rate": 1.0880000000000001e-05,
+      "loss": 0.6095,
+      "step": 1836
+    },
+    {
+      "epoch": 0.1632888888888889,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.0885925925925927e-05,
+      "loss": 0.7751,
+      "step": 1837
+    },
+    {
+      "epoch": 0.1633777777777778,
+      "grad_norm": 2.375,
+      "learning_rate": 1.0891851851851854e-05,
+      "loss": 0.7083,
+      "step": 1838
+    },
+    {
+      "epoch": 0.16346666666666668,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.0897777777777778e-05,
+      "loss": 0.6313,
+      "step": 1839
+    },
+    {
+      "epoch": 0.16355555555555557,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.0903703703703706e-05,
+      "loss": 0.7132,
+      "step": 1840
+    },
+    {
+      "epoch": 0.16364444444444445,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.090962962962963e-05,
+      "loss": 0.8355,
+      "step": 1841
+    },
+    {
+      "epoch": 0.16373333333333334,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.0915555555555557e-05,
+      "loss": 0.5993,
+      "step": 1842
+    },
+    {
+      "epoch": 0.16382222222222223,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.0921481481481481e-05,
+      "loss": 0.6459,
+      "step": 1843
+    },
+    {
+      "epoch": 0.16391111111111112,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.0927407407407409e-05,
+      "loss": 0.6992,
+      "step": 1844
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.0933333333333334e-05,
+      "loss": 0.7578,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1640888888888889,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.093925925925926e-05,
+      "loss": 0.7682,
+      "step": 1846
+    },
+    {
+      "epoch": 0.16417777777777778,
+      "grad_norm": 2.5,
+      "learning_rate": 1.0945185185185186e-05,
+      "loss": 0.7224,
+      "step": 1847
+    },
+    {
+      "epoch": 0.16426666666666667,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.0951111111111113e-05,
+      "loss": 0.7987,
+      "step": 1848
+    },
+    {
+      "epoch": 0.16435555555555556,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.0957037037037037e-05,
+      "loss": 0.7723,
+      "step": 1849
+    },
+    {
+      "epoch": 0.16444444444444445,
+      "grad_norm": 2.875,
+      "learning_rate": 1.0962962962962965e-05,
+      "loss": 0.687,
+      "step": 1850
+    },
+    {
+      "epoch": 0.16453333333333334,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.0968888888888889e-05,
+      "loss": 0.6654,
+      "step": 1851
+    },
+    {
+      "epoch": 0.16462222222222223,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.0974814814814816e-05,
+      "loss": 0.6643,
+      "step": 1852
+    },
+    {
+      "epoch": 0.1647111111111111,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.098074074074074e-05,
+      "loss": 0.6751,
+      "step": 1853
+    },
+    {
+      "epoch": 0.1648,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.0986666666666668e-05,
+      "loss": 0.7347,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1648888888888889,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.0992592592592592e-05,
+      "loss": 0.6674,
+      "step": 1855
+    },
+    {
+      "epoch": 0.16497777777777778,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.0998518518518519e-05,
+      "loss": 0.6475,
+      "step": 1856
+    },
+    {
+      "epoch": 0.16506666666666667,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1004444444444445e-05,
+      "loss": 0.7685,
+      "step": 1857
+    },
+    {
+      "epoch": 0.16515555555555556,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.101037037037037e-05,
+      "loss": 0.7037,
+      "step": 1858
+    },
+    {
+      "epoch": 0.16524444444444444,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.1016296296296298e-05,
+      "loss": 0.8107,
+      "step": 1859
+    },
+    {
+      "epoch": 0.16533333333333333,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.1022222222222224e-05,
+      "loss": 0.8436,
+      "step": 1860
+    },
+    {
+      "epoch": 0.16542222222222222,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.102814814814815e-05,
+      "loss": 0.7061,
+      "step": 1861
+    },
+    {
+      "epoch": 0.1655111111111111,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.1034074074074075e-05,
+      "loss": 0.6734,
+      "step": 1862
+    },
+    {
+      "epoch": 0.1656,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.1040000000000001e-05,
+      "loss": 0.6807,
+      "step": 1863
+    },
+    {
+      "epoch": 0.16568888888888889,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.1045925925925927e-05,
+      "loss": 0.876,
+      "step": 1864
+    },
+    {
+      "epoch": 0.16577777777777777,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.1051851851851854e-05,
+      "loss": 0.8206,
+      "step": 1865
+    },
+    {
+      "epoch": 0.16586666666666666,
+      "grad_norm": 2.5,
+      "learning_rate": 1.1057777777777778e-05,
+      "loss": 0.7485,
+      "step": 1866
+    },
+    {
+      "epoch": 0.16595555555555555,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.1063703703703706e-05,
+      "loss": 0.8619,
+      "step": 1867
+    },
+    {
+      "epoch": 0.16604444444444444,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.106962962962963e-05,
+      "loss": 0.5414,
+      "step": 1868
+    },
+    {
+      "epoch": 0.16613333333333333,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.1075555555555557e-05,
+      "loss": 0.5873,
+      "step": 1869
+    },
+    {
+      "epoch": 0.16622222222222222,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.1081481481481481e-05,
+      "loss": 0.7519,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1663111111111111,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.1087407407407409e-05,
+      "loss": 0.7773,
+      "step": 1871
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.1093333333333334e-05,
+      "loss": 0.61,
+      "step": 1872
+    },
+    {
+      "epoch": 0.16648888888888888,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.109925925925926e-05,
+      "loss": 0.6885,
+      "step": 1873
+    },
+    {
+      "epoch": 0.16657777777777777,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.1105185185185186e-05,
+      "loss": 0.7482,
+      "step": 1874
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 0.7506,
+      "step": 1875
+    },
+    {
+      "epoch": 0.16675555555555555,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.1117037037037037e-05,
+      "loss": 0.7707,
+      "step": 1876
+    },
+    {
+      "epoch": 0.16684444444444443,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.1122962962962965e-05,
+      "loss": 0.7803,
+      "step": 1877
+    },
+    {
+      "epoch": 0.16693333333333332,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.1128888888888889e-05,
+      "loss": 0.7515,
+      "step": 1878
+    },
+    {
+      "epoch": 0.1670222222222222,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.1134814814814816e-05,
+      "loss": 0.6518,
+      "step": 1879
+    },
+    {
+      "epoch": 0.1671111111111111,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.114074074074074e-05,
+      "loss": 0.8495,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1672,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.1146666666666668e-05,
+      "loss": 0.7195,
+      "step": 1881
+    },
+    {
+      "epoch": 0.16728888888888888,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.1152592592592592e-05,
+      "loss": 0.8034,
+      "step": 1882
+    },
+    {
+      "epoch": 0.16737777777777776,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.115851851851852e-05,
+      "loss": 0.8104,
+      "step": 1883
+    },
+    {
+      "epoch": 0.16746666666666668,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.1164444444444445e-05,
+      "loss": 0.7475,
+      "step": 1884
+    },
+    {
+      "epoch": 0.16755555555555557,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.117037037037037e-05,
+      "loss": 0.7125,
+      "step": 1885
+    },
+    {
+      "epoch": 0.16764444444444446,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.1176296296296297e-05,
+      "loss": 0.8055,
+      "step": 1886
+    },
+    {
+      "epoch": 0.16773333333333335,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.1182222222222224e-05,
+      "loss": 0.7931,
+      "step": 1887
+    },
+    {
+      "epoch": 0.16782222222222223,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.1188148148148148e-05,
+      "loss": 0.7557,
+      "step": 1888
+    },
+    {
+      "epoch": 0.16791111111111112,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.1194074074074075e-05,
+      "loss": 0.6997,
+      "step": 1889
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.8313,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1680888888888889,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1205925925925927e-05,
+      "loss": 0.6363,
+      "step": 1891
+    },
+    {
+      "epoch": 0.1681777777777778,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.1211851851851854e-05,
+      "loss": 0.6687,
+      "step": 1892
+    },
+    {
+      "epoch": 0.16826666666666668,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.1217777777777778e-05,
+      "loss": 0.7659,
+      "step": 1893
+    },
+    {
+      "epoch": 0.16835555555555556,
+      "grad_norm": 2.375,
+      "learning_rate": 1.1223703703703706e-05,
+      "loss": 0.7122,
+      "step": 1894
+    },
+    {
+      "epoch": 0.16844444444444445,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.122962962962963e-05,
+      "loss": 0.7449,
+      "step": 1895
+    },
+    {
+      "epoch": 0.16853333333333334,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1235555555555557e-05,
+      "loss": 0.6902,
+      "step": 1896
+    },
+    {
+      "epoch": 0.16862222222222223,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.1241481481481481e-05,
+      "loss": 0.7799,
+      "step": 1897
+    },
+    {
+      "epoch": 0.16871111111111112,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.1247407407407409e-05,
+      "loss": 0.6739,
+      "step": 1898
+    },
+    {
+      "epoch": 0.1688,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.1253333333333335e-05,
+      "loss": 0.8144,
+      "step": 1899
+    },
+    {
+      "epoch": 0.1688888888888889,
+      "grad_norm": 2.25,
+      "learning_rate": 1.125925925925926e-05,
+      "loss": 0.9376,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16897777777777778,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.1265185185185186e-05,
+      "loss": 0.6991,
+      "step": 1901
+    },
+    {
+      "epoch": 0.16906666666666667,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.1271111111111113e-05,
+      "loss": 0.7731,
+      "step": 1902
+    },
+    {
+      "epoch": 0.16915555555555556,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.1277037037037038e-05,
+      "loss": 0.6108,
+      "step": 1903
+    },
+    {
+      "epoch": 0.16924444444444445,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.1282962962962965e-05,
+      "loss": 0.5859,
+      "step": 1904
+    },
+    {
+      "epoch": 0.16933333333333334,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.1288888888888889e-05,
+      "loss": 0.7762,
+      "step": 1905
+    },
+    {
+      "epoch": 0.16942222222222222,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.1294814814814816e-05,
+      "loss": 0.7076,
+      "step": 1906
+    },
+    {
+      "epoch": 0.1695111111111111,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.130074074074074e-05,
+      "loss": 0.7859,
+      "step": 1907
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.1306666666666668e-05,
+      "loss": 0.5584,
+      "step": 1908
+    },
+    {
+      "epoch": 0.1696888888888889,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.1312592592592592e-05,
+      "loss": 0.8488,
+      "step": 1909
+    },
+    {
+      "epoch": 0.16977777777777778,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.131851851851852e-05,
+      "loss": 0.7748,
+      "step": 1910
+    },
+    {
+      "epoch": 0.16986666666666667,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.1324444444444445e-05,
+      "loss": 0.8084,
+      "step": 1911
+    },
+    {
+      "epoch": 0.16995555555555555,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.1330370370370371e-05,
+      "loss": 0.6368,
+      "step": 1912
+    },
+    {
+      "epoch": 0.17004444444444444,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.1336296296296297e-05,
+      "loss": 0.9459,
+      "step": 1913
+    },
+    {
+      "epoch": 0.17013333333333333,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.1342222222222224e-05,
+      "loss": 0.7337,
+      "step": 1914
+    },
+    {
+      "epoch": 0.17022222222222222,
+      "grad_norm": 2.5,
+      "learning_rate": 1.1348148148148148e-05,
+      "loss": 0.6572,
+      "step": 1915
+    },
+    {
+      "epoch": 0.1703111111111111,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.1354074074074076e-05,
+      "loss": 0.7695,
+      "step": 1916
+    },
+    {
+      "epoch": 0.1704,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.136e-05,
+      "loss": 0.6999,
+      "step": 1917
+    },
+    {
+      "epoch": 0.17048888888888888,
+      "grad_norm": 3.125,
+      "learning_rate": 1.1365925925925927e-05,
+      "loss": 0.8115,
+      "step": 1918
+    },
+    {
+      "epoch": 0.17057777777777777,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.1371851851851851e-05,
+      "loss": 0.663,
+      "step": 1919
+    },
+    {
+      "epoch": 0.17066666666666666,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.1377777777777779e-05,
+      "loss": 0.5869,
+      "step": 1920
+    },
+    {
+      "epoch": 0.17075555555555555,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.1383703703703706e-05,
+      "loss": 0.8221,
+      "step": 1921
+    },
+    {
+      "epoch": 0.17084444444444444,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.138962962962963e-05,
+      "loss": 0.5984,
+      "step": 1922
+    },
+    {
+      "epoch": 0.17093333333333333,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.1395555555555558e-05,
+      "loss": 0.7635,
+      "step": 1923
+    },
+    {
+      "epoch": 0.17102222222222221,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.1401481481481482e-05,
+      "loss": 0.5726,
+      "step": 1924
+    },
+    {
+      "epoch": 0.1711111111111111,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.1407407407407409e-05,
+      "loss": 0.5862,
+      "step": 1925
+    },
+    {
+      "epoch": 0.1712,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.1413333333333335e-05,
+      "loss": 0.7014,
+      "step": 1926
+    },
+    {
+      "epoch": 0.17128888888888888,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.141925925925926e-05,
+      "loss": 0.7276,
+      "step": 1927
+    },
+    {
+      "epoch": 0.17137777777777777,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1425185185185186e-05,
+      "loss": 0.8141,
+      "step": 1928
+    },
+    {
+      "epoch": 0.17146666666666666,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.1431111111111114e-05,
+      "loss": 0.6771,
+      "step": 1929
+    },
+    {
+      "epoch": 0.17155555555555554,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.1437037037037038e-05,
+      "loss": 0.8293,
+      "step": 1930
+    },
+    {
+      "epoch": 0.17164444444444443,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.1442962962962965e-05,
+      "loss": 0.7688,
+      "step": 1931
+    },
+    {
+      "epoch": 0.17173333333333332,
+      "grad_norm": 3.5,
+      "learning_rate": 1.144888888888889e-05,
+      "loss": 0.7816,
+      "step": 1932
+    },
+    {
+      "epoch": 0.1718222222222222,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.1454814814814817e-05,
+      "loss": 0.7257,
+      "step": 1933
+    },
+    {
+      "epoch": 0.1719111111111111,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.146074074074074e-05,
+      "loss": 0.7506,
+      "step": 1934
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.1466666666666668e-05,
+      "loss": 0.6895,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1720888888888889,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.1472592592592592e-05,
+      "loss": 0.7996,
+      "step": 1936
+    },
+    {
+      "epoch": 0.1721777777777778,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.147851851851852e-05,
+      "loss": 0.8152,
+      "step": 1937
+    },
+    {
+      "epoch": 0.17226666666666668,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.1484444444444445e-05,
+      "loss": 0.7622,
+      "step": 1938
+    },
+    {
+      "epoch": 0.17235555555555557,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1490370370370371e-05,
+      "loss": 0.7777,
+      "step": 1939
+    },
+    {
+      "epoch": 0.17244444444444446,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.1496296296296297e-05,
+      "loss": 0.7226,
+      "step": 1940
+    },
+    {
+      "epoch": 0.17253333333333334,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.1502222222222224e-05,
+      "loss": 0.7677,
+      "step": 1941
+    },
+    {
+      "epoch": 0.17262222222222223,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.1508148148148148e-05,
+      "loss": 0.7543,
+      "step": 1942
+    },
+    {
+      "epoch": 0.17271111111111112,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1514074074074076e-05,
+      "loss": 0.7511,
+      "step": 1943
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.152e-05,
+      "loss": 0.6526,
+      "step": 1944
+    },
+    {
+      "epoch": 0.1728888888888889,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.1525925925925927e-05,
+      "loss": 0.6155,
+      "step": 1945
+    },
+    {
+      "epoch": 0.17297777777777779,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.1531851851851851e-05,
+      "loss": 0.6553,
+      "step": 1946
+    },
+    {
+      "epoch": 0.17306666666666667,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.1537777777777779e-05,
+      "loss": 0.7265,
+      "step": 1947
+    },
+    {
+      "epoch": 0.17315555555555556,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.1543703703703703e-05,
+      "loss": 0.6962,
+      "step": 1948
+    },
+    {
+      "epoch": 0.17324444444444445,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.154962962962963e-05,
+      "loss": 0.7239,
+      "step": 1949
+    },
+    {
+      "epoch": 0.17333333333333334,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.1555555555555556e-05,
+      "loss": 0.9371,
+      "step": 1950
+    },
+    {
+      "epoch": 0.17342222222222223,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.1561481481481482e-05,
+      "loss": 0.7096,
+      "step": 1951
+    },
+    {
+      "epoch": 0.17351111111111112,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.156740740740741e-05,
+      "loss": 0.7146,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1736,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.1573333333333335e-05,
+      "loss": 0.8311,
+      "step": 1953
+    },
+    {
+      "epoch": 0.1736888888888889,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.157925925925926e-05,
+      "loss": 0.765,
+      "step": 1954
+    },
+    {
+      "epoch": 0.17377777777777778,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.1585185185185186e-05,
+      "loss": 0.7599,
+      "step": 1955
+    },
+    {
+      "epoch": 0.17386666666666667,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.1591111111111114e-05,
+      "loss": 0.7476,
+      "step": 1956
+    },
+    {
+      "epoch": 0.17395555555555556,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.1597037037037038e-05,
+      "loss": 0.7309,
+      "step": 1957
+    },
+    {
+      "epoch": 0.17404444444444445,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.1602962962962965e-05,
+      "loss": 0.7168,
+      "step": 1958
+    },
+    {
+      "epoch": 0.17413333333333333,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.160888888888889e-05,
+      "loss": 0.6896,
+      "step": 1959
+    },
+    {
+      "epoch": 0.17422222222222222,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.1614814814814817e-05,
+      "loss": 0.8648,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1743111111111111,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.1620740740740741e-05,
+      "loss": 0.8963,
+      "step": 1961
+    },
+    {
+      "epoch": 0.1744,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.1626666666666668e-05,
+      "loss": 0.8538,
+      "step": 1962
+    },
+    {
+      "epoch": 0.1744888888888889,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.1632592592592592e-05,
+      "loss": 0.7284,
+      "step": 1963
+    },
+    {
+      "epoch": 0.17457777777777778,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.163851851851852e-05,
+      "loss": 0.6605,
+      "step": 1964
+    },
+    {
+      "epoch": 0.17466666666666666,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.1644444444444446e-05,
+      "loss": 0.778,
+      "step": 1965
+    },
+    {
+      "epoch": 0.17475555555555555,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.1650370370370371e-05,
+      "loss": 0.606,
+      "step": 1966
+    },
+    {
+      "epoch": 0.17484444444444444,
+      "grad_norm": 2.375,
+      "learning_rate": 1.1656296296296297e-05,
+      "loss": 0.7786,
+      "step": 1967
+    },
+    {
+      "epoch": 0.17493333333333333,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.1662222222222224e-05,
+      "loss": 0.7435,
+      "step": 1968
+    },
+    {
+      "epoch": 0.17502222222222222,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.1668148148148148e-05,
+      "loss": 0.7689,
+      "step": 1969
+    },
+    {
+      "epoch": 0.1751111111111111,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.1674074074074076e-05,
+      "loss": 0.737,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1752,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.168e-05,
+      "loss": 0.6222,
+      "step": 1971
+    },
+    {
+      "epoch": 0.17528888888888888,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.1685925925925927e-05,
+      "loss": 0.73,
+      "step": 1972
+    },
+    {
+      "epoch": 0.17537777777777777,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.1691851851851851e-05,
+      "loss": 0.8118,
+      "step": 1973
+    },
+    {
+      "epoch": 0.17546666666666666,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1697777777777779e-05,
+      "loss": 0.6254,
+      "step": 1974
+    },
+    {
+      "epoch": 0.17555555555555555,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1703703703703703e-05,
+      "loss": 0.5514,
+      "step": 1975
+    },
+    {
+      "epoch": 0.17564444444444444,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.170962962962963e-05,
+      "loss": 0.7424,
+      "step": 1976
+    },
+    {
+      "epoch": 0.17573333333333332,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.1715555555555556e-05,
+      "loss": 0.6563,
+      "step": 1977
+    },
+    {
+      "epoch": 0.1758222222222222,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.1721481481481482e-05,
+      "loss": 0.7151,
+      "step": 1978
+    },
+    {
+      "epoch": 0.1759111111111111,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.1727407407407408e-05,
+      "loss": 0.6307,
+      "step": 1979
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 2.875,
+      "learning_rate": 1.1733333333333335e-05,
+      "loss": 0.7671,
+      "step": 1980
+    },
+    {
+      "epoch": 0.17608888888888888,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.1739259259259259e-05,
+      "loss": 0.9124,
+      "step": 1981
+    },
+    {
+      "epoch": 0.17617777777777777,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.1745185185185187e-05,
+      "loss": 0.8286,
+      "step": 1982
+    },
+    {
+      "epoch": 0.17626666666666665,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.1751111111111112e-05,
+      "loss": 0.8236,
+      "step": 1983
+    },
+    {
+      "epoch": 0.17635555555555554,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.1757037037037038e-05,
+      "loss": 0.7355,
+      "step": 1984
+    },
+    {
+      "epoch": 0.17644444444444443,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.1762962962962965e-05,
+      "loss": 0.8531,
+      "step": 1985
+    },
+    {
+      "epoch": 0.17653333333333332,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.176888888888889e-05,
+      "loss": 0.5262,
+      "step": 1986
+    },
+    {
+      "epoch": 0.17662222222222224,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.1774814814814817e-05,
+      "loss": 0.6114,
+      "step": 1987
+    },
+    {
+      "epoch": 0.17671111111111112,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.1780740740740741e-05,
+      "loss": 0.744,
+      "step": 1988
+    },
+    {
+      "epoch": 0.1768,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.1786666666666668e-05,
+      "loss": 0.7228,
+      "step": 1989
+    },
+    {
+      "epoch": 0.1768888888888889,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1792592592592592e-05,
+      "loss": 0.686,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1769777777777778,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.179851851851852e-05,
+      "loss": 0.6016,
+      "step": 1991
+    },
+    {
+      "epoch": 0.17706666666666668,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.1804444444444446e-05,
+      "loss": 0.7217,
+      "step": 1992
+    },
+    {
+      "epoch": 0.17715555555555557,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.1810370370370371e-05,
+      "loss": 0.6431,
+      "step": 1993
+    },
+    {
+      "epoch": 0.17724444444444445,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.1816296296296297e-05,
+      "loss": 0.7143,
+      "step": 1994
+    },
+    {
+      "epoch": 0.17733333333333334,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.1822222222222225e-05,
+      "loss": 0.7122,
+      "step": 1995
+    },
+    {
+      "epoch": 0.17742222222222223,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.1828148148148149e-05,
+      "loss": 0.6779,
+      "step": 1996
+    },
+    {
+      "epoch": 0.17751111111111112,
+      "grad_norm": 4.0,
+      "learning_rate": 1.1834074074074076e-05,
+      "loss": 0.586,
+      "step": 1997
+    },
+    {
+      "epoch": 0.1776,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.184e-05,
+      "loss": 0.8364,
+      "step": 1998
+    },
+    {
+      "epoch": 0.1776888888888889,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.1845925925925928e-05,
+      "loss": 0.8718,
+      "step": 1999
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.1851851851851852e-05,
+      "loss": 0.6612,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17786666666666667,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.1857777777777779e-05,
+      "loss": 0.8356,
+      "step": 2001
+    },
+    {
+      "epoch": 0.17795555555555556,
+      "grad_norm": 2.875,
+      "learning_rate": 1.1863703703703703e-05,
+      "loss": 0.735,
+      "step": 2002
+    },
+    {
+      "epoch": 0.17804444444444445,
+      "grad_norm": 2.0,
+      "learning_rate": 1.186962962962963e-05,
+      "loss": 0.6491,
+      "step": 2003
+    },
+    {
+      "epoch": 0.17813333333333334,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.1875555555555556e-05,
+      "loss": 0.7939,
+      "step": 2004
+    },
+    {
+      "epoch": 0.17822222222222223,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.1881481481481482e-05,
+      "loss": 0.696,
+      "step": 2005
+    },
+    {
+      "epoch": 0.17831111111111111,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.1887407407407408e-05,
+      "loss": 0.7329,
+      "step": 2006
+    },
+    {
+      "epoch": 0.1784,
+      "grad_norm": 3.0,
+      "learning_rate": 1.1893333333333335e-05,
+      "loss": 0.6224,
+      "step": 2007
+    },
+    {
+      "epoch": 0.1784888888888889,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.189925925925926e-05,
+      "loss": 0.7038,
+      "step": 2008
+    },
+    {
+      "epoch": 0.17857777777777778,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.1905185185185187e-05,
+      "loss": 0.75,
+      "step": 2009
+    },
+    {
+      "epoch": 0.17866666666666667,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.191111111111111e-05,
+      "loss": 0.9179,
+      "step": 2010
+    },
+    {
+      "epoch": 0.17875555555555556,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.1917037037037038e-05,
+      "loss": 0.7332,
+      "step": 2011
+    },
+    {
+      "epoch": 0.17884444444444444,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.1922962962962962e-05,
+      "loss": 0.6642,
+      "step": 2012
+    },
+    {
+      "epoch": 0.17893333333333333,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.192888888888889e-05,
+      "loss": 0.62,
+      "step": 2013
+    },
+    {
+      "epoch": 0.17902222222222222,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.1934814814814817e-05,
+      "loss": 0.919,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1791111111111111,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.1940740740740741e-05,
+      "loss": 0.753,
+      "step": 2015
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.1946666666666669e-05,
+      "loss": 0.6746,
+      "step": 2016
+    },
+    {
+      "epoch": 0.1792888888888889,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.1952592592592593e-05,
+      "loss": 0.7196,
+      "step": 2017
+    },
+    {
+      "epoch": 0.17937777777777777,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.195851851851852e-05,
+      "loss": 0.6258,
+      "step": 2018
+    },
+    {
+      "epoch": 0.17946666666666666,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.1964444444444446e-05,
+      "loss": 0.528,
+      "step": 2019
+    },
+    {
+      "epoch": 0.17955555555555555,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.1970370370370372e-05,
+      "loss": 0.6508,
+      "step": 2020
+    },
+    {
+      "epoch": 0.17964444444444444,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.1976296296296297e-05,
+      "loss": 0.6372,
+      "step": 2021
+    },
+    {
+      "epoch": 0.17973333333333333,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1982222222222225e-05,
+      "loss": 0.8969,
+      "step": 2022
+    },
+    {
+      "epoch": 0.17982222222222222,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.1988148148148149e-05,
+      "loss": 0.9344,
+      "step": 2023
+    },
+    {
+      "epoch": 0.1799111111111111,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.1994074074074076e-05,
+      "loss": 0.6303,
+      "step": 2024
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.2e-05,
+      "loss": 0.8072,
+      "step": 2025
+    },
+    {
+      "epoch": 0.18008888888888888,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2005925925925928e-05,
+      "loss": 0.7423,
+      "step": 2026
+    },
+    {
+      "epoch": 0.18017777777777777,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.2011851851851852e-05,
+      "loss": 0.8647,
+      "step": 2027
+    },
+    {
+      "epoch": 0.18026666666666666,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.201777777777778e-05,
+      "loss": 0.9974,
+      "step": 2028
+    },
+    {
+      "epoch": 0.18035555555555555,
+      "grad_norm": 2.625,
+      "learning_rate": 1.2023703703703703e-05,
+      "loss": 0.5809,
+      "step": 2029
+    },
+    {
+      "epoch": 0.18044444444444444,
+      "grad_norm": 2.5,
+      "learning_rate": 1.202962962962963e-05,
+      "loss": 0.7404,
+      "step": 2030
+    },
+    {
+      "epoch": 0.18053333333333332,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.2035555555555556e-05,
+      "loss": 0.7942,
+      "step": 2031
+    },
+    {
+      "epoch": 0.1806222222222222,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2041481481481482e-05,
+      "loss": 0.8038,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1807111111111111,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.2047407407407408e-05,
+      "loss": 0.7825,
+      "step": 2033
+    },
+    {
+      "epoch": 0.1808,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2053333333333335e-05,
+      "loss": 0.6792,
+      "step": 2034
+    },
+    {
+      "epoch": 0.18088888888888888,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.205925925925926e-05,
+      "loss": 0.7187,
+      "step": 2035
+    },
+    {
+      "epoch": 0.18097777777777777,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.2065185185185187e-05,
+      "loss": 0.5529,
+      "step": 2036
+    },
+    {
+      "epoch": 0.18106666666666665,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.2071111111111111e-05,
+      "loss": 0.6668,
+      "step": 2037
+    },
+    {
+      "epoch": 0.18115555555555554,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2077037037037038e-05,
+      "loss": 0.6218,
+      "step": 2038
+    },
+    {
+      "epoch": 0.18124444444444446,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.2082962962962962e-05,
+      "loss": 0.8833,
+      "step": 2039
+    },
+    {
+      "epoch": 0.18133333333333335,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.208888888888889e-05,
+      "loss": 0.769,
+      "step": 2040
+    },
+    {
+      "epoch": 0.18142222222222223,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.2094814814814816e-05,
+      "loss": 0.7568,
+      "step": 2041
+    },
+    {
+      "epoch": 0.18151111111111112,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.2100740740740741e-05,
+      "loss": 0.8049,
+      "step": 2042
+    },
+    {
+      "epoch": 0.1816,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.2106666666666667e-05,
+      "loss": 0.8017,
+      "step": 2043
+    },
+    {
+      "epoch": 0.1816888888888889,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.2112592592592593e-05,
+      "loss": 0.8376,
+      "step": 2044
+    },
+    {
+      "epoch": 0.1817777777777778,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.211851851851852e-05,
+      "loss": 0.8639,
+      "step": 2045
+    },
+    {
+      "epoch": 0.18186666666666668,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.2124444444444446e-05,
+      "loss": 0.7449,
+      "step": 2046
+    },
+    {
+      "epoch": 0.18195555555555556,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.2130370370370372e-05,
+      "loss": 0.7785,
+      "step": 2047
+    },
+    {
+      "epoch": 0.18204444444444445,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.2136296296296297e-05,
+      "loss": 0.6647,
+      "step": 2048
+    },
+    {
+      "epoch": 0.18213333333333334,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.2142222222222225e-05,
+      "loss": 0.7537,
+      "step": 2049
+    },
+    {
+      "epoch": 0.18222222222222223,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.2148148148148149e-05,
+      "loss": 0.5977,
+      "step": 2050
+    },
+    {
+      "epoch": 0.18231111111111112,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.2154074074074076e-05,
+      "loss": 0.7479,
+      "step": 2051
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.216e-05,
+      "loss": 0.7912,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1824888888888889,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.2165925925925928e-05,
+      "loss": 0.6879,
+      "step": 2053
+    },
+    {
+      "epoch": 0.18257777777777778,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2171851851851852e-05,
+      "loss": 0.6356,
+      "step": 2054
+    },
+    {
+      "epoch": 0.18266666666666667,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.217777777777778e-05,
+      "loss": 0.6544,
+      "step": 2055
+    },
+    {
+      "epoch": 0.18275555555555556,
+      "grad_norm": 2.25,
+      "learning_rate": 1.2183703703703703e-05,
+      "loss": 0.6433,
+      "step": 2056
+    },
+    {
+      "epoch": 0.18284444444444445,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.2189629629629631e-05,
+      "loss": 0.7882,
+      "step": 2057
+    },
+    {
+      "epoch": 0.18293333333333334,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.2195555555555557e-05,
+      "loss": 0.6902,
+      "step": 2058
+    },
+    {
+      "epoch": 0.18302222222222223,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.2201481481481482e-05,
+      "loss": 0.7035,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1831111111111111,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.2207407407407408e-05,
+      "loss": 0.6597,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1832,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.2213333333333336e-05,
+      "loss": 0.6107,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1832888888888889,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.221925925925926e-05,
+      "loss": 0.6959,
+      "step": 2062
+    },
+    {
+      "epoch": 0.18337777777777778,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.2225185185185187e-05,
+      "loss": 0.6065,
+      "step": 2063
+    },
+    {
+      "epoch": 0.18346666666666667,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.2231111111111111e-05,
+      "loss": 0.8065,
+      "step": 2064
+    },
+    {
+      "epoch": 0.18355555555555556,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.2237037037037039e-05,
+      "loss": 0.7826,
+      "step": 2065
+    },
+    {
+      "epoch": 0.18364444444444444,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.2242962962962963e-05,
+      "loss": 0.7684,
+      "step": 2066
+    },
+    {
+      "epoch": 0.18373333333333333,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.224888888888889e-05,
+      "loss": 0.562,
+      "step": 2067
+    },
+    {
+      "epoch": 0.18382222222222222,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.2254814814814816e-05,
+      "loss": 0.7293,
+      "step": 2068
+    },
+    {
+      "epoch": 0.1839111111111111,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.2260740740740741e-05,
+      "loss": 0.6859,
+      "step": 2069
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.2266666666666667e-05,
+      "loss": 0.6894,
+      "step": 2070
+    },
+    {
+      "epoch": 0.18408888888888889,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.2272592592592593e-05,
+      "loss": 0.5793,
+      "step": 2071
+    },
+    {
+      "epoch": 0.18417777777777777,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.2278518518518519e-05,
+      "loss": 0.7087,
+      "step": 2072
+    },
+    {
+      "epoch": 0.18426666666666666,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.2284444444444446e-05,
+      "loss": 0.8282,
+      "step": 2073
+    },
+    {
+      "epoch": 0.18435555555555555,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.229037037037037e-05,
+      "loss": 0.6908,
+      "step": 2074
+    },
+    {
+      "epoch": 0.18444444444444444,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.2296296296296298e-05,
+      "loss": 0.5919,
+      "step": 2075
+    },
+    {
+      "epoch": 0.18453333333333333,
+      "grad_norm": 5.1875,
+      "learning_rate": 1.2302222222222225e-05,
+      "loss": 0.7845,
+      "step": 2076
+    },
+    {
+      "epoch": 0.18462222222222222,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.2308148148148149e-05,
+      "loss": 0.7966,
+      "step": 2077
+    },
+    {
+      "epoch": 0.1847111111111111,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.2314074074074077e-05,
+      "loss": 0.6386,
+      "step": 2078
+    },
+    {
+      "epoch": 0.1848,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.232e-05,
+      "loss": 0.6549,
+      "step": 2079
+    },
+    {
+      "epoch": 0.18488888888888888,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.2325925925925928e-05,
+      "loss": 0.7775,
+      "step": 2080
+    },
+    {
+      "epoch": 0.18497777777777777,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.2331851851851852e-05,
+      "loss": 0.6775,
+      "step": 2081
+    },
+    {
+      "epoch": 0.18506666666666666,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.233777777777778e-05,
+      "loss": 0.6919,
+      "step": 2082
+    },
+    {
+      "epoch": 0.18515555555555555,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2343703703703704e-05,
+      "loss": 0.7097,
+      "step": 2083
+    },
+    {
+      "epoch": 0.18524444444444443,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.2349629629629631e-05,
+      "loss": 0.7755,
+      "step": 2084
+    },
+    {
+      "epoch": 0.18533333333333332,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2355555555555557e-05,
+      "loss": 0.7108,
+      "step": 2085
+    },
+    {
+      "epoch": 0.1854222222222222,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.2361481481481483e-05,
+      "loss": 0.6909,
+      "step": 2086
+    },
+    {
+      "epoch": 0.1855111111111111,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.2367407407407408e-05,
+      "loss": 0.713,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.2373333333333336e-05,
+      "loss": 0.6271,
+      "step": 2088
+    },
+    {
+      "epoch": 0.18568888888888888,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.237925925925926e-05,
+      "loss": 0.6847,
+      "step": 2089
+    },
+    {
+      "epoch": 0.18577777777777776,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.2385185185185187e-05,
+      "loss": 0.8376,
+      "step": 2090
+    },
+    {
+      "epoch": 0.18586666666666668,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.2391111111111111e-05,
+      "loss": 0.7049,
+      "step": 2091
+    },
+    {
+      "epoch": 0.18595555555555557,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.2397037037037039e-05,
+      "loss": 0.8674,
+      "step": 2092
+    },
+    {
+      "epoch": 0.18604444444444446,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.2402962962962963e-05,
+      "loss": 0.7446,
+      "step": 2093
+    },
+    {
+      "epoch": 0.18613333333333335,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.240888888888889e-05,
+      "loss": 0.5435,
+      "step": 2094
+    },
+    {
+      "epoch": 0.18622222222222223,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.2414814814814816e-05,
+      "loss": 0.7997,
+      "step": 2095
+    },
+    {
+      "epoch": 0.18631111111111112,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.2420740740740742e-05,
+      "loss": 0.574,
+      "step": 2096
+    },
+    {
+      "epoch": 0.1864,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.2426666666666667e-05,
+      "loss": 0.8149,
+      "step": 2097
+    },
+    {
+      "epoch": 0.1864888888888889,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2432592592592593e-05,
+      "loss": 0.9422,
+      "step": 2098
+    },
+    {
+      "epoch": 0.1865777777777778,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.2438518518518519e-05,
+      "loss": 0.7149,
+      "step": 2099
+    },
+    {
+      "epoch": 0.18666666666666668,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.2444444444444446e-05,
+      "loss": 0.7115,
+      "step": 2100
+    },
+    {
+      "epoch": 0.18675555555555556,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.245037037037037e-05,
+      "loss": 0.7109,
+      "step": 2101
+    },
+    {
+      "epoch": 0.18684444444444445,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.2456296296296298e-05,
+      "loss": 0.6197,
+      "step": 2102
+    },
+    {
+      "epoch": 0.18693333333333334,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2462222222222222e-05,
+      "loss": 0.4676,
+      "step": 2103
+    },
+    {
+      "epoch": 0.18702222222222223,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.246814814814815e-05,
+      "loss": 0.6566,
+      "step": 2104
+    },
+    {
+      "epoch": 0.18711111111111112,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.2474074074074073e-05,
+      "loss": 0.6979,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1872,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.248e-05,
+      "loss": 0.7176,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1872888888888889,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.2485925925925928e-05,
+      "loss": 0.674,
+      "step": 2107
+    },
+    {
+      "epoch": 0.18737777777777778,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.2491851851851852e-05,
+      "loss": 0.7182,
+      "step": 2108
+    },
+    {
+      "epoch": 0.18746666666666667,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.249777777777778e-05,
+      "loss": 0.7151,
+      "step": 2109
+    },
+    {
+      "epoch": 0.18755555555555556,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.2503703703703704e-05,
+      "loss": 0.6906,
+      "step": 2110
+    },
+    {
+      "epoch": 0.18764444444444445,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.2509629629629631e-05,
+      "loss": 0.7601,
+      "step": 2111
+    },
+    {
+      "epoch": 0.18773333333333334,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.2515555555555557e-05,
+      "loss": 0.5937,
+      "step": 2112
+    },
+    {
+      "epoch": 0.18782222222222222,
+      "grad_norm": 2.125,
+      "learning_rate": 1.2521481481481483e-05,
+      "loss": 0.7074,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1879111111111111,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.2527407407407408e-05,
+      "loss": 0.6048,
+      "step": 2114
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.2533333333333336e-05,
+      "loss": 0.8789,
+      "step": 2115
+    },
+    {
+      "epoch": 0.1880888888888889,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.253925925925926e-05,
+      "loss": 0.6329,
+      "step": 2116
+    },
+    {
+      "epoch": 0.18817777777777778,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.2545185185185187e-05,
+      "loss": 0.7206,
+      "step": 2117
+    },
+    {
+      "epoch": 0.18826666666666667,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.2551111111111111e-05,
+      "loss": 0.5747,
+      "step": 2118
+    },
+    {
+      "epoch": 0.18835555555555555,
+      "grad_norm": 2.375,
+      "learning_rate": 1.2557037037037039e-05,
+      "loss": 0.7397,
+      "step": 2119
+    },
+    {
+      "epoch": 0.18844444444444444,
+      "grad_norm": 2.375,
+      "learning_rate": 1.2562962962962963e-05,
+      "loss": 0.808,
+      "step": 2120
+    },
+    {
+      "epoch": 0.18853333333333333,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.256888888888889e-05,
+      "loss": 0.7184,
+      "step": 2121
+    },
+    {
+      "epoch": 0.18862222222222222,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.2574814814814816e-05,
+      "loss": 0.7081,
+      "step": 2122
+    },
+    {
+      "epoch": 0.1887111111111111,
+      "grad_norm": 2.25,
+      "learning_rate": 1.2580740740740742e-05,
+      "loss": 0.7576,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.2586666666666668e-05,
+      "loss": 0.8216,
+      "step": 2124
+    },
+    {
+      "epoch": 0.18888888888888888,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.2592592592592593e-05,
+      "loss": 0.7546,
+      "step": 2125
+    },
+    {
+      "epoch": 0.18897777777777777,
+      "grad_norm": 2.75,
+      "learning_rate": 1.2598518518518519e-05,
+      "loss": 0.6315,
+      "step": 2126
+    },
+    {
+      "epoch": 0.18906666666666666,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2604444444444446e-05,
+      "loss": 0.6417,
+      "step": 2127
+    },
+    {
+      "epoch": 0.18915555555555555,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.261037037037037e-05,
+      "loss": 0.7933,
+      "step": 2128
+    },
+    {
+      "epoch": 0.18924444444444444,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.2616296296296298e-05,
+      "loss": 0.6749,
+      "step": 2129
+    },
+    {
+      "epoch": 0.18933333333333333,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2622222222222222e-05,
+      "loss": 0.7572,
+      "step": 2130
+    },
+    {
+      "epoch": 0.18942222222222221,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.262814814814815e-05,
+      "loss": 0.7509,
+      "step": 2131
+    },
+    {
+      "epoch": 0.1895111111111111,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.2634074074074074e-05,
+      "loss": 0.6379,
+      "step": 2132
+    },
+    {
+      "epoch": 0.1896,
+      "grad_norm": 2.5,
+      "learning_rate": 1.2640000000000001e-05,
+      "loss": 0.7821,
+      "step": 2133
+    },
+    {
+      "epoch": 0.18968888888888888,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.2645925925925927e-05,
+      "loss": 0.8964,
+      "step": 2134
+    },
+    {
+      "epoch": 0.18977777777777777,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.2651851851851852e-05,
+      "loss": 0.7162,
+      "step": 2135
+    },
+    {
+      "epoch": 0.18986666666666666,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.2657777777777778e-05,
+      "loss": 0.7842,
+      "step": 2136
+    },
+    {
+      "epoch": 0.18995555555555554,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.2663703703703704e-05,
+      "loss": 0.5984,
+      "step": 2137
+    },
+    {
+      "epoch": 0.19004444444444443,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.2669629629629631e-05,
+      "loss": 0.7721,
+      "step": 2138
+    },
+    {
+      "epoch": 0.19013333333333332,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.2675555555555557e-05,
+      "loss": 0.763,
+      "step": 2139
+    },
+    {
+      "epoch": 0.1902222222222222,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.2681481481481483e-05,
+      "loss": 0.7279,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1903111111111111,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.2687407407407409e-05,
+      "loss": 0.6319,
+      "step": 2141
+    },
+    {
+      "epoch": 0.1904,
+      "grad_norm": 2.25,
+      "learning_rate": 1.2693333333333336e-05,
+      "loss": 0.7267,
+      "step": 2142
+    },
+    {
+      "epoch": 0.1904888888888889,
+      "grad_norm": 2.625,
+      "learning_rate": 1.269925925925926e-05,
+      "loss": 0.8294,
+      "step": 2143
+    },
+    {
+      "epoch": 0.1905777777777778,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.2705185185185188e-05,
+      "loss": 0.6405,
+      "step": 2144
+    },
+    {
+      "epoch": 0.19066666666666668,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.2711111111111112e-05,
+      "loss": 0.8243,
+      "step": 2145
+    },
+    {
+      "epoch": 0.19075555555555557,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.2717037037037039e-05,
+      "loss": 0.7613,
+      "step": 2146
+    },
+    {
+      "epoch": 0.19084444444444446,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.2722962962962963e-05,
+      "loss": 0.7816,
+      "step": 2147
+    },
+    {
+      "epoch": 0.19093333333333334,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.272888888888889e-05,
+      "loss": 0.8182,
+      "step": 2148
+    },
+    {
+      "epoch": 0.19102222222222223,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.2734814814814816e-05,
+      "loss": 0.8297,
+      "step": 2149
+    },
+    {
+      "epoch": 0.19111111111111112,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.2740740740740742e-05,
+      "loss": 0.7173,
+      "step": 2150
+    },
+    {
+      "epoch": 0.1912,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.2746666666666668e-05,
+      "loss": 0.7718,
+      "step": 2151
+    },
+    {
+      "epoch": 0.1912888888888889,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.2752592592592593e-05,
+      "loss": 0.7258,
+      "step": 2152
+    },
+    {
+      "epoch": 0.19137777777777779,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.275851851851852e-05,
+      "loss": 0.7726,
+      "step": 2153
+    },
+    {
+      "epoch": 0.19146666666666667,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.2764444444444447e-05,
+      "loss": 0.7607,
+      "step": 2154
+    },
+    {
+      "epoch": 0.19155555555555556,
+      "grad_norm": 2.375,
+      "learning_rate": 1.277037037037037e-05,
+      "loss": 0.771,
+      "step": 2155
+    },
+    {
+      "epoch": 0.19164444444444445,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.2776296296296298e-05,
+      "loss": 0.8735,
+      "step": 2156
+    },
+    {
+      "epoch": 0.19173333333333334,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.2782222222222222e-05,
+      "loss": 0.7427,
+      "step": 2157
+    },
+    {
+      "epoch": 0.19182222222222223,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.278814814814815e-05,
+      "loss": 0.6484,
+      "step": 2158
+    },
+    {
+      "epoch": 0.19191111111111112,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.2794074074074074e-05,
+      "loss": 0.7419,
+      "step": 2159
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.6693,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1920888888888889,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.2805925925925927e-05,
+      "loss": 0.7113,
+      "step": 2161
+    },
+    {
+      "epoch": 0.19217777777777778,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.2811851851851853e-05,
+      "loss": 0.7579,
+      "step": 2162
+    },
+    {
+      "epoch": 0.19226666666666667,
+      "grad_norm": 2.25,
+      "learning_rate": 1.2817777777777778e-05,
+      "loss": 0.6772,
+      "step": 2163
+    },
+    {
+      "epoch": 0.19235555555555556,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.2823703703703704e-05,
+      "loss": 0.7088,
+      "step": 2164
+    },
+    {
+      "epoch": 0.19244444444444445,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.282962962962963e-05,
+      "loss": 0.837,
+      "step": 2165
+    },
+    {
+      "epoch": 0.19253333333333333,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2835555555555557e-05,
+      "loss": 0.7039,
+      "step": 2166
+    },
+    {
+      "epoch": 0.19262222222222222,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.2841481481481481e-05,
+      "loss": 0.6275,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1927111111111111,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.2847407407407409e-05,
+      "loss": 0.7681,
+      "step": 2168
+    },
+    {
+      "epoch": 0.1928,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.2853333333333336e-05,
+      "loss": 0.792,
+      "step": 2169
+    },
+    {
+      "epoch": 0.1928888888888889,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.285925925925926e-05,
+      "loss": 0.7594,
+      "step": 2170
+    },
+    {
+      "epoch": 0.19297777777777778,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.2865185185185188e-05,
+      "loss": 0.5996,
+      "step": 2171
+    },
+    {
+      "epoch": 0.19306666666666666,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.2871111111111112e-05,
+      "loss": 0.8231,
+      "step": 2172
+    },
+    {
+      "epoch": 0.19315555555555555,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.287703703703704e-05,
+      "loss": 0.632,
+      "step": 2173
+    },
+    {
+      "epoch": 0.19324444444444444,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.2882962962962963e-05,
+      "loss": 0.7241,
+      "step": 2174
+    },
+    {
+      "epoch": 0.19333333333333333,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.288888888888889e-05,
+      "loss": 0.7411,
+      "step": 2175
+    },
+    {
+      "epoch": 0.19342222222222222,
+      "grad_norm": 3.125,
+      "learning_rate": 1.2894814814814816e-05,
+      "loss": 0.6544,
+      "step": 2176
+    },
+    {
+      "epoch": 0.1935111111111111,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.2900740740740742e-05,
+      "loss": 0.8018,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1936,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.2906666666666668e-05,
+      "loss": 0.5644,
+      "step": 2178
+    },
+    {
+      "epoch": 0.19368888888888888,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.2912592592592594e-05,
+      "loss": 0.825,
+      "step": 2179
+    },
+    {
+      "epoch": 0.19377777777777777,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.291851851851852e-05,
+      "loss": 0.6815,
+      "step": 2180
+    },
+    {
+      "epoch": 0.19386666666666666,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.2924444444444447e-05,
+      "loss": 0.6226,
+      "step": 2181
+    },
+    {
+      "epoch": 0.19395555555555555,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.2930370370370371e-05,
+      "loss": 0.8381,
+      "step": 2182
+    },
+    {
+      "epoch": 0.19404444444444444,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.2936296296296298e-05,
+      "loss": 0.7918,
+      "step": 2183
+    },
+    {
+      "epoch": 0.19413333333333332,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.2942222222222222e-05,
+      "loss": 0.8351,
+      "step": 2184
+    },
+    {
+      "epoch": 0.1942222222222222,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.294814814814815e-05,
+      "loss": 0.7598,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1943111111111111,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.2954074074074074e-05,
+      "loss": 0.8007,
+      "step": 2186
+    },
+    {
+      "epoch": 0.1944,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.2960000000000001e-05,
+      "loss": 0.8151,
+      "step": 2187
+    },
+    {
+      "epoch": 0.19448888888888888,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.2965925925925927e-05,
+      "loss": 0.601,
+      "step": 2188
+    },
+    {
+      "epoch": 0.19457777777777777,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.2971851851851853e-05,
+      "loss": 0.7854,
+      "step": 2189
+    },
+    {
+      "epoch": 0.19466666666666665,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.2977777777777779e-05,
+      "loss": 0.8119,
+      "step": 2190
+    },
+    {
+      "epoch": 0.19475555555555554,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.2983703703703704e-05,
+      "loss": 0.6886,
+      "step": 2191
+    },
+    {
+      "epoch": 0.19484444444444443,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.298962962962963e-05,
+      "loss": 0.7778,
+      "step": 2192
+    },
+    {
+      "epoch": 0.19493333333333332,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.2995555555555557e-05,
+      "loss": 0.5938,
+      "step": 2193
+    },
+    {
+      "epoch": 0.19502222222222224,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.3001481481481481e-05,
+      "loss": 0.7072,
+      "step": 2194
+    },
+    {
+      "epoch": 0.19511111111111112,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.3007407407407409e-05,
+      "loss": 0.8449,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3013333333333333e-05,
+      "loss": 0.5959,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1952888888888889,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.301925925925926e-05,
+      "loss": 0.6924,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1953777777777778,
+      "grad_norm": 2.125,
+      "learning_rate": 1.3025185185185184e-05,
+      "loss": 0.7932,
+      "step": 2198
+    },
+    {
+      "epoch": 0.19546666666666668,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.3031111111111112e-05,
+      "loss": 0.7395,
+      "step": 2199
+    },
+    {
+      "epoch": 0.19555555555555557,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.303703703703704e-05,
+      "loss": 0.7737,
+      "step": 2200
+    },
+    {
+      "epoch": 0.19564444444444445,
+      "grad_norm": 2.125,
+      "learning_rate": 1.3042962962962963e-05,
+      "loss": 0.6811,
+      "step": 2201
+    },
+    {
+      "epoch": 0.19573333333333334,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.304888888888889e-05,
+      "loss": 0.5961,
+      "step": 2202
+    },
+    {
+      "epoch": 0.19582222222222223,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3054814814814817e-05,
+      "loss": 0.6551,
+      "step": 2203
+    },
+    {
+      "epoch": 0.19591111111111112,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.3060740740740742e-05,
+      "loss": 0.6615,
+      "step": 2204
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.3066666666666668e-05,
+      "loss": 0.6609,
+      "step": 2205
+    },
+    {
+      "epoch": 0.1960888888888889,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.3072592592592594e-05,
+      "loss": 0.7645,
+      "step": 2206
+    },
+    {
+      "epoch": 0.19617777777777778,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.307851851851852e-05,
+      "loss": 0.6968,
+      "step": 2207
+    },
+    {
+      "epoch": 0.19626666666666667,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.3084444444444447e-05,
+      "loss": 0.76,
+      "step": 2208
+    },
+    {
+      "epoch": 0.19635555555555556,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.3090370370370371e-05,
+      "loss": 0.8555,
+      "step": 2209
+    },
+    {
+      "epoch": 0.19644444444444445,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.3096296296296298e-05,
+      "loss": 0.7316,
+      "step": 2210
+    },
+    {
+      "epoch": 0.19653333333333334,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.3102222222222223e-05,
+      "loss": 0.7128,
+      "step": 2211
+    },
+    {
+      "epoch": 0.19662222222222223,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.310814814814815e-05,
+      "loss": 0.927,
+      "step": 2212
+    },
+    {
+      "epoch": 0.19671111111111111,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3114074074074074e-05,
+      "loss": 0.9199,
+      "step": 2213
+    },
+    {
+      "epoch": 0.1968,
+      "grad_norm": 3.125,
+      "learning_rate": 1.3120000000000001e-05,
+      "loss": 0.9017,
+      "step": 2214
+    },
+    {
+      "epoch": 0.1968888888888889,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.3125925925925927e-05,
+      "loss": 0.7193,
+      "step": 2215
+    },
+    {
+      "epoch": 0.19697777777777778,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.3131851851851853e-05,
+      "loss": 0.8485,
+      "step": 2216
+    },
+    {
+      "epoch": 0.19706666666666667,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3137777777777779e-05,
+      "loss": 0.7813,
+      "step": 2217
+    },
+    {
+      "epoch": 0.19715555555555556,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.3143703703703704e-05,
+      "loss": 0.8824,
+      "step": 2218
+    },
+    {
+      "epoch": 0.19724444444444444,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.314962962962963e-05,
+      "loss": 0.9063,
+      "step": 2219
+    },
+    {
+      "epoch": 0.19733333333333333,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.3155555555555558e-05,
+      "loss": 0.8049,
+      "step": 2220
+    },
+    {
+      "epoch": 0.19742222222222222,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.3161481481481482e-05,
+      "loss": 0.6204,
+      "step": 2221
+    },
+    {
+      "epoch": 0.1975111111111111,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.3167407407407409e-05,
+      "loss": 0.7557,
+      "step": 2222
+    },
+    {
+      "epoch": 0.1976,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.3173333333333333e-05,
+      "loss": 0.7106,
+      "step": 2223
+    },
+    {
+      "epoch": 0.1976888888888889,
+      "grad_norm": 2.875,
+      "learning_rate": 1.317925925925926e-05,
+      "loss": 0.7122,
+      "step": 2224
+    },
+    {
+      "epoch": 0.19777777777777777,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.3185185185185185e-05,
+      "loss": 0.5351,
+      "step": 2225
+    },
+    {
+      "epoch": 0.19786666666666666,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.3191111111111112e-05,
+      "loss": 0.7508,
+      "step": 2226
+    },
+    {
+      "epoch": 0.19795555555555555,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.3197037037037038e-05,
+      "loss": 0.7187,
+      "step": 2227
+    },
+    {
+      "epoch": 0.19804444444444444,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.3202962962962964e-05,
+      "loss": 0.7552,
+      "step": 2228
+    },
+    {
+      "epoch": 0.19813333333333333,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.320888888888889e-05,
+      "loss": 0.9607,
+      "step": 2229
+    },
+    {
+      "epoch": 0.19822222222222222,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.3214814814814815e-05,
+      "loss": 0.6735,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1983111111111111,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.3220740740740742e-05,
+      "loss": 0.6138,
+      "step": 2231
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.3226666666666668e-05,
+      "loss": 0.8307,
+      "step": 2232
+    },
+    {
+      "epoch": 0.19848888888888888,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.3232592592592594e-05,
+      "loss": 0.8257,
+      "step": 2233
+    },
+    {
+      "epoch": 0.19857777777777777,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.323851851851852e-05,
+      "loss": 0.6496,
+      "step": 2234
+    },
+    {
+      "epoch": 0.19866666666666666,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3244444444444447e-05,
+      "loss": 0.7025,
+      "step": 2235
+    },
+    {
+      "epoch": 0.19875555555555555,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.3250370370370371e-05,
+      "loss": 0.8191,
+      "step": 2236
+    },
+    {
+      "epoch": 0.19884444444444443,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.3256296296296299e-05,
+      "loss": 0.6295,
+      "step": 2237
+    },
+    {
+      "epoch": 0.19893333333333332,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3262222222222223e-05,
+      "loss": 0.7768,
+      "step": 2238
+    },
+    {
+      "epoch": 0.1990222222222222,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.326814814814815e-05,
+      "loss": 0.6739,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1991111111111111,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.3274074074074074e-05,
+      "loss": 0.6679,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1992,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.3280000000000002e-05,
+      "loss": 0.6743,
+      "step": 2241
+    },
+    {
+      "epoch": 0.19928888888888888,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.3285925925925927e-05,
+      "loss": 0.6864,
+      "step": 2242
+    },
+    {
+      "epoch": 0.19937777777777776,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.3291851851851853e-05,
+      "loss": 0.6327,
+      "step": 2243
+    },
+    {
+      "epoch": 0.19946666666666665,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3297777777777779e-05,
+      "loss": 0.8043,
+      "step": 2244
+    },
+    {
+      "epoch": 0.19955555555555557,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.3303703703703705e-05,
+      "loss": 0.6861,
+      "step": 2245
+    },
+    {
+      "epoch": 0.19964444444444446,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.330962962962963e-05,
+      "loss": 0.8832,
+      "step": 2246
+    },
+    {
+      "epoch": 0.19973333333333335,
+      "grad_norm": 2.5,
+      "learning_rate": 1.3315555555555558e-05,
+      "loss": 0.6775,
+      "step": 2247
+    },
+    {
+      "epoch": 0.19982222222222223,
+      "grad_norm": 2.75,
+      "learning_rate": 1.3321481481481482e-05,
+      "loss": 0.7535,
+      "step": 2248
+    },
+    {
+      "epoch": 0.19991111111111112,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.332740740740741e-05,
+      "loss": 0.658,
+      "step": 2249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7006,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2000888888888889,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.333925925925926e-05,
+      "loss": 0.6506,
+      "step": 2251
+    },
+    {
+      "epoch": 0.2001777777777778,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.3345185185185185e-05,
+      "loss": 0.7118,
+      "step": 2252
+    },
+    {
+      "epoch": 0.20026666666666668,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.3351111111111112e-05,
+      "loss": 0.6557,
+      "step": 2253
+    },
+    {
+      "epoch": 0.20035555555555556,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.3357037037037038e-05,
+      "loss": 0.693,
+      "step": 2254
+    },
+    {
+      "epoch": 0.20044444444444445,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.3362962962962964e-05,
+      "loss": 0.6625,
+      "step": 2255
+    },
+    {
+      "epoch": 0.20053333333333334,
+      "grad_norm": 3.375,
+      "learning_rate": 1.336888888888889e-05,
+      "loss": 0.8192,
+      "step": 2256
+    },
+    {
+      "epoch": 0.20062222222222223,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.3374814814814815e-05,
+      "loss": 0.666,
+      "step": 2257
+    },
+    {
+      "epoch": 0.20071111111111112,
+      "grad_norm": 2.375,
+      "learning_rate": 1.3380740740740741e-05,
+      "loss": 0.8205,
+      "step": 2258
+    },
+    {
+      "epoch": 0.2008,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3386666666666668e-05,
+      "loss": 0.7588,
+      "step": 2259
+    },
+    {
+      "epoch": 0.2008888888888889,
+      "grad_norm": 2.375,
+      "learning_rate": 1.3392592592592592e-05,
+      "loss": 0.668,
+      "step": 2260
+    },
+    {
+      "epoch": 0.20097777777777778,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.339851851851852e-05,
+      "loss": 0.6353,
+      "step": 2261
+    },
+    {
+      "epoch": 0.20106666666666667,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.3404444444444447e-05,
+      "loss": 0.7129,
+      "step": 2262
+    },
+    {
+      "epoch": 0.20115555555555556,
+      "grad_norm": 2.25,
+      "learning_rate": 1.3410370370370371e-05,
+      "loss": 0.7848,
+      "step": 2263
+    },
+    {
+      "epoch": 0.20124444444444445,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.3416296296296299e-05,
+      "loss": 0.7614,
+      "step": 2264
+    },
+    {
+      "epoch": 0.20133333333333334,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3422222222222223e-05,
+      "loss": 0.6063,
+      "step": 2265
+    },
+    {
+      "epoch": 0.20142222222222222,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.342814814814815e-05,
+      "loss": 0.7205,
+      "step": 2266
+    },
+    {
+      "epoch": 0.2015111111111111,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3434074074074074e-05,
+      "loss": 0.6663,
+      "step": 2267
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.3440000000000002e-05,
+      "loss": 0.7539,
+      "step": 2268
+    },
+    {
+      "epoch": 0.2016888888888889,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.3445925925925928e-05,
+      "loss": 0.7217,
+      "step": 2269
+    },
+    {
+      "epoch": 0.20177777777777778,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.3451851851851853e-05,
+      "loss": 0.71,
+      "step": 2270
+    },
+    {
+      "epoch": 0.20186666666666667,
+      "grad_norm": 2.5,
+      "learning_rate": 1.3457777777777779e-05,
+      "loss": 0.7718,
+      "step": 2271
+    },
+    {
+      "epoch": 0.20195555555555555,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3463703703703705e-05,
+      "loss": 0.6165,
+      "step": 2272
+    },
+    {
+      "epoch": 0.20204444444444444,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.346962962962963e-05,
+      "loss": 0.5134,
+      "step": 2273
+    },
+    {
+      "epoch": 0.20213333333333333,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.3475555555555558e-05,
+      "loss": 0.6601,
+      "step": 2274
+    },
+    {
+      "epoch": 0.20222222222222222,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.3481481481481482e-05,
+      "loss": 0.6742,
+      "step": 2275
+    },
+    {
+      "epoch": 0.2023111111111111,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.348740740740741e-05,
+      "loss": 0.6619,
+      "step": 2276
+    },
+    {
+      "epoch": 0.2024,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.3493333333333333e-05,
+      "loss": 0.6698,
+      "step": 2277
+    },
+    {
+      "epoch": 0.20248888888888888,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.3499259259259261e-05,
+      "loss": 0.6843,
+      "step": 2278
+    },
+    {
+      "epoch": 0.20257777777777777,
+      "grad_norm": 4.34375,
+      "learning_rate": 1.3505185185185185e-05,
+      "loss": 0.6638,
+      "step": 2279
+    },
+    {
+      "epoch": 0.20266666666666666,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.3511111111111112e-05,
+      "loss": 0.7048,
+      "step": 2280
+    },
+    {
+      "epoch": 0.20275555555555555,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3517037037037038e-05,
+      "loss": 0.6515,
+      "step": 2281
+    },
+    {
+      "epoch": 0.20284444444444444,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.3522962962962964e-05,
+      "loss": 0.7991,
+      "step": 2282
+    },
+    {
+      "epoch": 0.20293333333333333,
+      "grad_norm": 2.0,
+      "learning_rate": 1.352888888888889e-05,
+      "loss": 0.678,
+      "step": 2283
+    },
+    {
+      "epoch": 0.20302222222222222,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3534814814814815e-05,
+      "loss": 0.6611,
+      "step": 2284
+    },
+    {
+      "epoch": 0.2031111111111111,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.3540740740740741e-05,
+      "loss": 0.6813,
+      "step": 2285
+    },
+    {
+      "epoch": 0.2032,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3546666666666669e-05,
+      "loss": 0.7773,
+      "step": 2286
+    },
+    {
+      "epoch": 0.20328888888888888,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.3552592592592593e-05,
+      "loss": 0.8115,
+      "step": 2287
+    },
+    {
+      "epoch": 0.20337777777777777,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.355851851851852e-05,
+      "loss": 0.8381,
+      "step": 2288
+    },
+    {
+      "epoch": 0.20346666666666666,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.3564444444444444e-05,
+      "loss": 0.5738,
+      "step": 2289
+    },
+    {
+      "epoch": 0.20355555555555555,
+      "grad_norm": 2.25,
+      "learning_rate": 1.3570370370370372e-05,
+      "loss": 0.6205,
+      "step": 2290
+    },
+    {
+      "epoch": 0.20364444444444443,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.3576296296296296e-05,
+      "loss": 0.6962,
+      "step": 2291
+    },
+    {
+      "epoch": 0.20373333333333332,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.3582222222222223e-05,
+      "loss": 0.7102,
+      "step": 2292
+    },
+    {
+      "epoch": 0.2038222222222222,
+      "grad_norm": 2.125,
+      "learning_rate": 1.358814814814815e-05,
+      "loss": 0.5722,
+      "step": 2293
+    },
+    {
+      "epoch": 0.2039111111111111,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.3594074074074074e-05,
+      "loss": 0.6713,
+      "step": 2294
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 2.125,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.5757,
+      "step": 2295
+    },
+    {
+      "epoch": 0.20408888888888888,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.3605925925925928e-05,
+      "loss": 0.8761,
+      "step": 2296
+    },
+    {
+      "epoch": 0.2041777777777778,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.3611851851851853e-05,
+      "loss": 0.6991,
+      "step": 2297
+    },
+    {
+      "epoch": 0.20426666666666668,
+      "grad_norm": 5.96875,
+      "learning_rate": 1.361777777777778e-05,
+      "loss": 0.8005,
+      "step": 2298
+    },
+    {
+      "epoch": 0.20435555555555557,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.3623703703703705e-05,
+      "loss": 0.6876,
+      "step": 2299
+    },
+    {
+      "epoch": 0.20444444444444446,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.362962962962963e-05,
+      "loss": 0.631,
+      "step": 2300
+    },
+    {
+      "epoch": 0.20453333333333334,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.3635555555555558e-05,
+      "loss": 0.6615,
+      "step": 2301
+    },
+    {
+      "epoch": 0.20462222222222223,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3641481481481482e-05,
+      "loss": 0.6808,
+      "step": 2302
+    },
+    {
+      "epoch": 0.20471111111111112,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.364740740740741e-05,
+      "loss": 0.4803,
+      "step": 2303
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.3653333333333334e-05,
+      "loss": 0.6191,
+      "step": 2304
+    },
+    {
+      "epoch": 0.2048888888888889,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3659259259259261e-05,
+      "loss": 0.6756,
+      "step": 2305
+    },
+    {
+      "epoch": 0.2049777777777778,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.3665185185185185e-05,
+      "loss": 0.6196,
+      "step": 2306
+    },
+    {
+      "epoch": 0.20506666666666667,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.3671111111111113e-05,
+      "loss": 0.6297,
+      "step": 2307
+    },
+    {
+      "epoch": 0.20515555555555556,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.3677037037037038e-05,
+      "loss": 0.6238,
+      "step": 2308
+    },
+    {
+      "epoch": 0.20524444444444445,
+      "grad_norm": 1.875,
+      "learning_rate": 1.3682962962962964e-05,
+      "loss": 0.6726,
+      "step": 2309
+    },
+    {
+      "epoch": 0.20533333333333334,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.368888888888889e-05,
+      "loss": 0.7283,
+      "step": 2310
+    },
+    {
+      "epoch": 0.20542222222222223,
+      "grad_norm": 3.0,
+      "learning_rate": 1.3694814814814816e-05,
+      "loss": 0.6085,
+      "step": 2311
+    },
+    {
+      "epoch": 0.20551111111111112,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.3700740740740741e-05,
+      "loss": 0.6974,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2056,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.3706666666666669e-05,
+      "loss": 0.6569,
+      "step": 2313
+    },
+    {
+      "epoch": 0.2056888888888889,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.3712592592592593e-05,
+      "loss": 0.7058,
+      "step": 2314
+    },
+    {
+      "epoch": 0.20577777777777778,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.371851851851852e-05,
+      "loss": 0.7616,
+      "step": 2315
+    },
+    {
+      "epoch": 0.20586666666666667,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.3724444444444444e-05,
+      "loss": 0.5405,
+      "step": 2316
+    },
+    {
+      "epoch": 0.20595555555555556,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.3730370370370372e-05,
+      "loss": 0.6303,
+      "step": 2317
+    },
+    {
+      "epoch": 0.20604444444444445,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3736296296296296e-05,
+      "loss": 0.7295,
+      "step": 2318
+    },
+    {
+      "epoch": 0.20613333333333334,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.3742222222222223e-05,
+      "loss": 0.6332,
+      "step": 2319
+    },
+    {
+      "epoch": 0.20622222222222222,
+      "grad_norm": 2.375,
+      "learning_rate": 1.3748148148148149e-05,
+      "loss": 0.6485,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2063111111111111,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.3754074074074075e-05,
+      "loss": 0.7407,
+      "step": 2321
+    },
+    {
+      "epoch": 0.2064,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.376e-05,
+      "loss": 0.7963,
+      "step": 2322
+    },
+    {
+      "epoch": 0.2064888888888889,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.3765925925925928e-05,
+      "loss": 0.7886,
+      "step": 2323
+    },
+    {
+      "epoch": 0.20657777777777778,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.3771851851851854e-05,
+      "loss": 0.6129,
+      "step": 2324
+    },
+    {
+      "epoch": 0.20666666666666667,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.377777777777778e-05,
+      "loss": 0.7906,
+      "step": 2325
+    },
+    {
+      "epoch": 0.20675555555555555,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3783703703703705e-05,
+      "loss": 0.7451,
+      "step": 2326
+    },
+    {
+      "epoch": 0.20684444444444444,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.378962962962963e-05,
+      "loss": 0.7269,
+      "step": 2327
+    },
+    {
+      "epoch": 0.20693333333333333,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.3795555555555558e-05,
+      "loss": 0.7971,
+      "step": 2328
+    },
+    {
+      "epoch": 0.20702222222222222,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.3801481481481482e-05,
+      "loss": 0.7294,
+      "step": 2329
+    },
+    {
+      "epoch": 0.2071111111111111,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.380740740740741e-05,
+      "loss": 0.7382,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2072,
+      "grad_norm": 3.0,
+      "learning_rate": 1.3813333333333334e-05,
+      "loss": 0.768,
+      "step": 2331
+    },
+    {
+      "epoch": 0.20728888888888888,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.3819259259259261e-05,
+      "loss": 0.6567,
+      "step": 2332
+    },
+    {
+      "epoch": 0.20737777777777777,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.3825185185185185e-05,
+      "loss": 0.604,
+      "step": 2333
+    },
+    {
+      "epoch": 0.20746666666666666,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.3831111111111113e-05,
+      "loss": 0.8908,
+      "step": 2334
+    },
+    {
+      "epoch": 0.20755555555555555,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.3837037037037038e-05,
+      "loss": 0.7534,
+      "step": 2335
+    },
+    {
+      "epoch": 0.20764444444444444,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.3842962962962964e-05,
+      "loss": 0.7161,
+      "step": 2336
+    },
+    {
+      "epoch": 0.20773333333333333,
+      "grad_norm": 2.25,
+      "learning_rate": 1.384888888888889e-05,
+      "loss": 0.6686,
+      "step": 2337
+    },
+    {
+      "epoch": 0.2078222222222222,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.3854814814814816e-05,
+      "loss": 0.7432,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2079111111111111,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.3860740740740741e-05,
+      "loss": 0.8076,
+      "step": 2339
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.3866666666666669e-05,
+      "loss": 0.4632,
+      "step": 2340
+    },
+    {
+      "epoch": 0.20808888888888888,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.3872592592592593e-05,
+      "loss": 0.6296,
+      "step": 2341
+    },
+    {
+      "epoch": 0.20817777777777777,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.387851851851852e-05,
+      "loss": 0.6416,
+      "step": 2342
+    },
+    {
+      "epoch": 0.20826666666666666,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.3884444444444444e-05,
+      "loss": 0.8936,
+      "step": 2343
+    },
+    {
+      "epoch": 0.20835555555555554,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.3890370370370372e-05,
+      "loss": 0.7293,
+      "step": 2344
+    },
+    {
+      "epoch": 0.20844444444444443,
+      "grad_norm": 3.625,
+      "learning_rate": 1.3896296296296296e-05,
+      "loss": 1.0136,
+      "step": 2345
+    },
+    {
+      "epoch": 0.20853333333333332,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3902222222222223e-05,
+      "loss": 0.6707,
+      "step": 2346
+    },
+    {
+      "epoch": 0.2086222222222222,
+      "grad_norm": 2.0,
+      "learning_rate": 1.3908148148148149e-05,
+      "loss": 0.6675,
+      "step": 2347
+    },
+    {
+      "epoch": 0.2087111111111111,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.3914074074074075e-05,
+      "loss": 0.6744,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2088,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.392e-05,
+      "loss": 0.8201,
+      "step": 2349
+    },
+    {
+      "epoch": 0.2088888888888889,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.3925925925925928e-05,
+      "loss": 0.5337,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2089777777777778,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.3931851851851852e-05,
+      "loss": 0.6256,
+      "step": 2351
+    },
+    {
+      "epoch": 0.20906666666666668,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.393777777777778e-05,
+      "loss": 0.798,
+      "step": 2352
+    },
+    {
+      "epoch": 0.20915555555555557,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.3943703703703704e-05,
+      "loss": 0.7338,
+      "step": 2353
+    },
+    {
+      "epoch": 0.20924444444444446,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.3949629629629631e-05,
+      "loss": 0.6361,
+      "step": 2354
+    },
+    {
+      "epoch": 0.20933333333333334,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.3955555555555558e-05,
+      "loss": 0.7563,
+      "step": 2355
+    },
+    {
+      "epoch": 0.20942222222222223,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.3961481481481482e-05,
+      "loss": 0.7107,
+      "step": 2356
+    },
+    {
+      "epoch": 0.20951111111111112,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.396740740740741e-05,
+      "loss": 0.7382,
+      "step": 2357
+    },
+    {
+      "epoch": 0.2096,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3973333333333334e-05,
+      "loss": 0.597,
+      "step": 2358
+    },
+    {
+      "epoch": 0.2096888888888889,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.3979259259259261e-05,
+      "loss": 0.704,
+      "step": 2359
+    },
+    {
+      "epoch": 0.20977777777777779,
+      "grad_norm": 2.25,
+      "learning_rate": 1.3985185185185185e-05,
+      "loss": 0.735,
+      "step": 2360
+    },
+    {
+      "epoch": 0.20986666666666667,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.3991111111111113e-05,
+      "loss": 0.7314,
+      "step": 2361
+    },
+    {
+      "epoch": 0.20995555555555556,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.3997037037037039e-05,
+      "loss": 0.7515,
+      "step": 2362
+    },
+    {
+      "epoch": 0.21004444444444445,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.4002962962962964e-05,
+      "loss": 0.7325,
+      "step": 2363
+    },
+    {
+      "epoch": 0.21013333333333334,
+      "grad_norm": 2.25,
+      "learning_rate": 1.400888888888889e-05,
+      "loss": 0.6398,
+      "step": 2364
+    },
+    {
+      "epoch": 0.21022222222222223,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.4014814814814816e-05,
+      "loss": 0.6531,
+      "step": 2365
+    },
+    {
+      "epoch": 0.21031111111111112,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.4020740740740742e-05,
+      "loss": 0.7253,
+      "step": 2366
+    },
+    {
+      "epoch": 0.2104,
+      "grad_norm": 2.25,
+      "learning_rate": 1.4026666666666669e-05,
+      "loss": 0.75,
+      "step": 2367
+    },
+    {
+      "epoch": 0.2104888888888889,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4032592592592593e-05,
+      "loss": 0.7531,
+      "step": 2368
+    },
+    {
+      "epoch": 0.21057777777777778,
+      "grad_norm": 2.875,
+      "learning_rate": 1.403851851851852e-05,
+      "loss": 0.5676,
+      "step": 2369
+    },
+    {
+      "epoch": 0.21066666666666667,
+      "grad_norm": 2.25,
+      "learning_rate": 1.4044444444444445e-05,
+      "loss": 0.6524,
+      "step": 2370
+    },
+    {
+      "epoch": 0.21075555555555556,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.4050370370370372e-05,
+      "loss": 0.7017,
+      "step": 2371
+    },
+    {
+      "epoch": 0.21084444444444445,
+      "grad_norm": 2.25,
+      "learning_rate": 1.4056296296296296e-05,
+      "loss": 0.6139,
+      "step": 2372
+    },
+    {
+      "epoch": 0.21093333333333333,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.4062222222222223e-05,
+      "loss": 0.5778,
+      "step": 2373
+    },
+    {
+      "epoch": 0.21102222222222222,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.406814814814815e-05,
+      "loss": 0.6964,
+      "step": 2374
+    },
+    {
+      "epoch": 0.2111111111111111,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.4074074074074075e-05,
+      "loss": 0.6955,
+      "step": 2375
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.408e-05,
+      "loss": 0.6234,
+      "step": 2376
+    },
+    {
+      "epoch": 0.2112888888888889,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4085925925925928e-05,
+      "loss": 0.7611,
+      "step": 2377
+    },
+    {
+      "epoch": 0.21137777777777778,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.4091851851851852e-05,
+      "loss": 0.748,
+      "step": 2378
+    },
+    {
+      "epoch": 0.21146666666666666,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.409777777777778e-05,
+      "loss": 0.7326,
+      "step": 2379
+    },
+    {
+      "epoch": 0.21155555555555555,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.4103703703703704e-05,
+      "loss": 0.6059,
+      "step": 2380
+    },
+    {
+      "epoch": 0.21164444444444444,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.4109629629629631e-05,
+      "loss": 0.6068,
+      "step": 2381
+    },
+    {
+      "epoch": 0.21173333333333333,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.4115555555555555e-05,
+      "loss": 0.8832,
+      "step": 2382
+    },
+    {
+      "epoch": 0.21182222222222222,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.4121481481481483e-05,
+      "loss": 0.8033,
+      "step": 2383
+    },
+    {
+      "epoch": 0.2119111111111111,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.4127407407407407e-05,
+      "loss": 0.6359,
+      "step": 2384
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.4133333333333334e-05,
+      "loss": 0.77,
+      "step": 2385
+    },
+    {
+      "epoch": 0.21208888888888888,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.4139259259259262e-05,
+      "loss": 0.7701,
+      "step": 2386
+    },
+    {
+      "epoch": 0.21217777777777777,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.4145185185185186e-05,
+      "loss": 0.7674,
+      "step": 2387
+    },
+    {
+      "epoch": 0.21226666666666666,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.4151111111111113e-05,
+      "loss": 0.6724,
+      "step": 2388
+    },
+    {
+      "epoch": 0.21235555555555555,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.4157037037037039e-05,
+      "loss": 0.6515,
+      "step": 2389
+    },
+    {
+      "epoch": 0.21244444444444444,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.4162962962962965e-05,
+      "loss": 0.6868,
+      "step": 2390
+    },
+    {
+      "epoch": 0.21253333333333332,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.416888888888889e-05,
+      "loss": 0.6838,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2126222222222222,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.4174814814814816e-05,
+      "loss": 0.6863,
+      "step": 2392
+    },
+    {
+      "epoch": 0.2127111111111111,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4180740740740742e-05,
+      "loss": 0.8506,
+      "step": 2393
+    },
+    {
+      "epoch": 0.2128,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.418666666666667e-05,
+      "loss": 0.8352,
+      "step": 2394
+    },
+    {
+      "epoch": 0.21288888888888888,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.4192592592592593e-05,
+      "loss": 0.6546,
+      "step": 2395
+    },
+    {
+      "epoch": 0.21297777777777777,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.419851851851852e-05,
+      "loss": 0.8414,
+      "step": 2396
+    },
+    {
+      "epoch": 0.21306666666666665,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.4204444444444445e-05,
+      "loss": 0.6448,
+      "step": 2397
+    },
+    {
+      "epoch": 0.21315555555555554,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.4210370370370372e-05,
+      "loss": 0.7776,
+      "step": 2398
+    },
+    {
+      "epoch": 0.21324444444444443,
+      "grad_norm": 2.75,
+      "learning_rate": 1.4216296296296296e-05,
+      "loss": 0.7134,
+      "step": 2399
+    },
+    {
+      "epoch": 0.21333333333333335,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.4222222222222224e-05,
+      "loss": 0.8323,
+      "step": 2400
+    },
+    {
+      "epoch": 0.21342222222222224,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.422814814814815e-05,
+      "loss": 0.9121,
+      "step": 2401
+    },
+    {
+      "epoch": 0.21351111111111112,
+      "grad_norm": 2.875,
+      "learning_rate": 1.4234074074074075e-05,
+      "loss": 0.7544,
+      "step": 2402
+    },
+    {
+      "epoch": 0.2136,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.4240000000000001e-05,
+      "loss": 0.8028,
+      "step": 2403
+    },
+    {
+      "epoch": 0.2136888888888889,
+      "grad_norm": 2.375,
+      "learning_rate": 1.4245925925925928e-05,
+      "loss": 0.775,
+      "step": 2404
+    },
+    {
+      "epoch": 0.2137777777777778,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4251851851851852e-05,
+      "loss": 0.7012,
+      "step": 2405
+    },
+    {
+      "epoch": 0.21386666666666668,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.425777777777778e-05,
+      "loss": 0.6602,
+      "step": 2406
+    },
+    {
+      "epoch": 0.21395555555555557,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4263703703703704e-05,
+      "loss": 0.7508,
+      "step": 2407
+    },
+    {
+      "epoch": 0.21404444444444445,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.4269629629629631e-05,
+      "loss": 0.7013,
+      "step": 2408
+    },
+    {
+      "epoch": 0.21413333333333334,
+      "grad_norm": 2.125,
+      "learning_rate": 1.4275555555555555e-05,
+      "loss": 0.6824,
+      "step": 2409
+    },
+    {
+      "epoch": 0.21422222222222223,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.4281481481481483e-05,
+      "loss": 0.731,
+      "step": 2410
+    },
+    {
+      "epoch": 0.21431111111111112,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4287407407407407e-05,
+      "loss": 0.7077,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 2.625,
+      "learning_rate": 1.4293333333333334e-05,
+      "loss": 0.6932,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2144888888888889,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.429925925925926e-05,
+      "loss": 0.6064,
+      "step": 2413
+    },
+    {
+      "epoch": 0.21457777777777778,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.4305185185185186e-05,
+      "loss": 0.6373,
+      "step": 2414
+    },
+    {
+      "epoch": 0.21466666666666667,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.4311111111111111e-05,
+      "loss": 0.8894,
+      "step": 2415
+    },
+    {
+      "epoch": 0.21475555555555556,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4317037037037039e-05,
+      "loss": 0.7955,
+      "step": 2416
+    },
+    {
+      "epoch": 0.21484444444444445,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.4322962962962965e-05,
+      "loss": 0.5834,
+      "step": 2417
+    },
+    {
+      "epoch": 0.21493333333333334,
+      "grad_norm": 1.625,
+      "learning_rate": 1.432888888888889e-05,
+      "loss": 0.5943,
+      "step": 2418
+    },
+    {
+      "epoch": 0.21502222222222223,
+      "grad_norm": 2.125,
+      "learning_rate": 1.4334814814814816e-05,
+      "loss": 0.7083,
+      "step": 2419
+    },
+    {
+      "epoch": 0.21511111111111111,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.4340740740740742e-05,
+      "loss": 0.9411,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2152,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.434666666666667e-05,
+      "loss": 0.7128,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2152888888888889,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.4352592592592593e-05,
+      "loss": 0.708,
+      "step": 2422
+    },
+    {
+      "epoch": 0.21537777777777778,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.435851851851852e-05,
+      "loss": 0.574,
+      "step": 2423
+    },
+    {
+      "epoch": 0.21546666666666667,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.4364444444444445e-05,
+      "loss": 0.674,
+      "step": 2424
+    },
+    {
+      "epoch": 0.21555555555555556,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.4370370370370372e-05,
+      "loss": 0.7116,
+      "step": 2425
+    },
+    {
+      "epoch": 0.21564444444444444,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.4376296296296296e-05,
+      "loss": 0.6336,
+      "step": 2426
+    },
+    {
+      "epoch": 0.21573333333333333,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.4382222222222224e-05,
+      "loss": 0.712,
+      "step": 2427
+    },
+    {
+      "epoch": 0.21582222222222222,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.438814814814815e-05,
+      "loss": 0.7006,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2159111111111111,
+      "grad_norm": 2.625,
+      "learning_rate": 1.4394074074074075e-05,
+      "loss": 0.7583,
+      "step": 2429
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 0.7134,
+      "step": 2430
+    },
+    {
+      "epoch": 0.21608888888888889,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.4405925925925928e-05,
+      "loss": 0.828,
+      "step": 2431
+    },
+    {
+      "epoch": 0.21617777777777777,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.4411851851851853e-05,
+      "loss": 0.7488,
+      "step": 2432
+    },
+    {
+      "epoch": 0.21626666666666666,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.441777777777778e-05,
+      "loss": 0.8318,
+      "step": 2433
+    },
+    {
+      "epoch": 0.21635555555555555,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.4423703703703704e-05,
+      "loss": 0.7832,
+      "step": 2434
+    },
+    {
+      "epoch": 0.21644444444444444,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.4429629629629631e-05,
+      "loss": 0.7098,
+      "step": 2435
+    },
+    {
+      "epoch": 0.21653333333333333,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4435555555555556e-05,
+      "loss": 0.5223,
+      "step": 2436
+    },
+    {
+      "epoch": 0.21662222222222222,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.4441481481481483e-05,
+      "loss": 0.4835,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2167111111111111,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.4447407407407407e-05,
+      "loss": 0.6492,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2168,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.4453333333333334e-05,
+      "loss": 0.7823,
+      "step": 2439
+    },
+    {
+      "epoch": 0.21688888888888888,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.445925925925926e-05,
+      "loss": 0.7141,
+      "step": 2440
+    },
+    {
+      "epoch": 0.21697777777777777,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4465185185185186e-05,
+      "loss": 0.8124,
+      "step": 2441
+    },
+    {
+      "epoch": 0.21706666666666666,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.4471111111111112e-05,
+      "loss": 0.543,
+      "step": 2442
+    },
+    {
+      "epoch": 0.21715555555555555,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.4477037037037039e-05,
+      "loss": 0.832,
+      "step": 2443
+    },
+    {
+      "epoch": 0.21724444444444443,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.4482962962962963e-05,
+      "loss": 0.8118,
+      "step": 2444
+    },
+    {
+      "epoch": 0.21733333333333332,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.448888888888889e-05,
+      "loss": 0.5768,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2174222222222222,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.4494814814814815e-05,
+      "loss": 0.7504,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2175111111111111,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.4500740740740742e-05,
+      "loss": 0.7567,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.450666666666667e-05,
+      "loss": 0.5171,
+      "step": 2448
+    },
+    {
+      "epoch": 0.21768888888888888,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4512592592592594e-05,
+      "loss": 0.7293,
+      "step": 2449
+    },
+    {
+      "epoch": 0.21777777777777776,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.4518518518518521e-05,
+      "loss": 0.6365,
+      "step": 2450
+    },
+    {
+      "epoch": 0.21786666666666665,
+      "grad_norm": 2.125,
+      "learning_rate": 1.4524444444444445e-05,
+      "loss": 0.6912,
+      "step": 2451
+    },
+    {
+      "epoch": 0.21795555555555557,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.4530370370370372e-05,
+      "loss": 0.6581,
+      "step": 2452
+    },
+    {
+      "epoch": 0.21804444444444446,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.4536296296296297e-05,
+      "loss": 0.6318,
+      "step": 2453
+    },
+    {
+      "epoch": 0.21813333333333335,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4542222222222224e-05,
+      "loss": 0.6805,
+      "step": 2454
+    },
+    {
+      "epoch": 0.21822222222222223,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.454814814814815e-05,
+      "loss": 0.7154,
+      "step": 2455
+    },
+    {
+      "epoch": 0.21831111111111112,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.4554074074074075e-05,
+      "loss": 0.7709,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2184,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.4560000000000001e-05,
+      "loss": 0.6823,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2184888888888889,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.4565925925925929e-05,
+      "loss": 0.7711,
+      "step": 2458
+    },
+    {
+      "epoch": 0.2185777777777778,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.4571851851851853e-05,
+      "loss": 0.6154,
+      "step": 2459
+    },
+    {
+      "epoch": 0.21866666666666668,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.457777777777778e-05,
+      "loss": 0.6971,
+      "step": 2460
+    },
+    {
+      "epoch": 0.21875555555555556,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.4583703703703704e-05,
+      "loss": 0.6743,
+      "step": 2461
+    },
+    {
+      "epoch": 0.21884444444444445,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.4589629629629632e-05,
+      "loss": 0.5441,
+      "step": 2462
+    },
+    {
+      "epoch": 0.21893333333333334,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.4595555555555556e-05,
+      "loss": 0.6746,
+      "step": 2463
+    },
+    {
+      "epoch": 0.21902222222222223,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.4601481481481483e-05,
+      "loss": 0.7195,
+      "step": 2464
+    },
+    {
+      "epoch": 0.21911111111111112,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.4607407407407407e-05,
+      "loss": 0.6742,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2192,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.4613333333333335e-05,
+      "loss": 0.7686,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2192888888888889,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.461925925925926e-05,
+      "loss": 0.6597,
+      "step": 2467
+    },
+    {
+      "epoch": 0.21937777777777778,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.4625185185185186e-05,
+      "loss": 0.5614,
+      "step": 2468
+    },
+    {
+      "epoch": 0.21946666666666667,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.4631111111111112e-05,
+      "loss": 0.6541,
+      "step": 2469
+    },
+    {
+      "epoch": 0.21955555555555556,
+      "grad_norm": 2.25,
+      "learning_rate": 1.463703703703704e-05,
+      "loss": 0.7895,
+      "step": 2470
+    },
+    {
+      "epoch": 0.21964444444444445,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.4642962962962963e-05,
+      "loss": 0.6509,
+      "step": 2471
+    },
+    {
+      "epoch": 0.21973333333333334,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.464888888888889e-05,
+      "loss": 0.7614,
+      "step": 2472
+    },
+    {
+      "epoch": 0.21982222222222222,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.4654814814814815e-05,
+      "loss": 0.6009,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2199111111111111,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.4660740740740742e-05,
+      "loss": 0.7137,
+      "step": 2474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 0.7716,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2200888888888889,
+      "grad_norm": 2.375,
+      "learning_rate": 1.4672592592592594e-05,
+      "loss": 0.7206,
+      "step": 2476
+    },
+    {
+      "epoch": 0.22017777777777778,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.4678518518518518e-05,
+      "loss": 0.5376,
+      "step": 2477
+    },
+    {
+      "epoch": 0.22026666666666667,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.4684444444444445e-05,
+      "loss": 0.7829,
+      "step": 2478
+    },
+    {
+      "epoch": 0.22035555555555555,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.4690370370370373e-05,
+      "loss": 0.6523,
+      "step": 2479
+    },
+    {
+      "epoch": 0.22044444444444444,
+      "grad_norm": 2.375,
+      "learning_rate": 1.4696296296296297e-05,
+      "loss": 0.6829,
+      "step": 2480
+    },
+    {
+      "epoch": 0.22053333333333333,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.4702222222222224e-05,
+      "loss": 0.5825,
+      "step": 2481
+    },
+    {
+      "epoch": 0.22062222222222222,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.470814814814815e-05,
+      "loss": 0.7599,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2207111111111111,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.4714074074074076e-05,
+      "loss": 0.9018,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.4720000000000001e-05,
+      "loss": 0.5895,
+      "step": 2484
+    },
+    {
+      "epoch": 0.22088888888888888,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.4725925925925927e-05,
+      "loss": 0.6911,
+      "step": 2485
+    },
+    {
+      "epoch": 0.22097777777777777,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.4731851851851853e-05,
+      "loss": 0.7391,
+      "step": 2486
+    },
+    {
+      "epoch": 0.22106666666666666,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.473777777777778e-05,
+      "loss": 0.8305,
+      "step": 2487
+    },
+    {
+      "epoch": 0.22115555555555555,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.4743703703703704e-05,
+      "loss": 0.7554,
+      "step": 2488
+    },
+    {
+      "epoch": 0.22124444444444444,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.4749629629629632e-05,
+      "loss": 0.6379,
+      "step": 2489
+    },
+    {
+      "epoch": 0.22133333333333333,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4755555555555556e-05,
+      "loss": 0.6727,
+      "step": 2490
+    },
+    {
+      "epoch": 0.22142222222222221,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.4761481481481483e-05,
+      "loss": 0.6256,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2215111111111111,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.4767407407407407e-05,
+      "loss": 0.7085,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2216,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4773333333333335e-05,
+      "loss": 0.6508,
+      "step": 2493
+    },
+    {
+      "epoch": 0.22168888888888888,
+      "grad_norm": 2.875,
+      "learning_rate": 1.477925925925926e-05,
+      "loss": 0.8127,
+      "step": 2494
+    },
+    {
+      "epoch": 0.22177777777777777,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.4785185185185186e-05,
+      "loss": 0.6407,
+      "step": 2495
+    },
+    {
+      "epoch": 0.22186666666666666,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.4791111111111112e-05,
+      "loss": 0.7347,
+      "step": 2496
+    },
+    {
+      "epoch": 0.22195555555555554,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.479703703703704e-05,
+      "loss": 0.6727,
+      "step": 2497
+    },
+    {
+      "epoch": 0.22204444444444443,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.4802962962962963e-05,
+      "loss": 0.6469,
+      "step": 2498
+    },
+    {
+      "epoch": 0.22213333333333332,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.4808888888888891e-05,
+      "loss": 0.723,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.7065,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2223111111111111,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.4820740740740742e-05,
+      "loss": 0.7989,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2224,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.4826666666666666e-05,
+      "loss": 0.7128,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2224888888888889,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.4832592592592594e-05,
+      "loss": 0.7895,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2225777777777778,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.4838518518518518e-05,
+      "loss": 0.8318,
+      "step": 2504
+    },
+    {
+      "epoch": 0.22266666666666668,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4844444444444445e-05,
+      "loss": 0.7924,
+      "step": 2505
+    },
+    {
+      "epoch": 0.22275555555555557,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.4850370370370371e-05,
+      "loss": 0.7579,
+      "step": 2506
+    },
+    {
+      "epoch": 0.22284444444444446,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4856296296296297e-05,
+      "loss": 0.759,
+      "step": 2507
+    },
+    {
+      "epoch": 0.22293333333333334,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.4862222222222223e-05,
+      "loss": 0.7457,
+      "step": 2508
+    },
+    {
+      "epoch": 0.22302222222222223,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.486814814814815e-05,
+      "loss": 0.7416,
+      "step": 2509
+    },
+    {
+      "epoch": 0.22311111111111112,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.4874074074074076e-05,
+      "loss": 0.8179,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2232,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.4880000000000002e-05,
+      "loss": 0.7385,
+      "step": 2511
+    },
+    {
+      "epoch": 0.2232888888888889,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.4885925925925927e-05,
+      "loss": 0.6582,
+      "step": 2512
+    },
+    {
+      "epoch": 0.2233777777777778,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4891851851851853e-05,
+      "loss": 0.6524,
+      "step": 2513
+    },
+    {
+      "epoch": 0.22346666666666667,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.489777777777778e-05,
+      "loss": 0.7571,
+      "step": 2514
+    },
+    {
+      "epoch": 0.22355555555555556,
+      "grad_norm": 2.375,
+      "learning_rate": 1.4903703703703705e-05,
+      "loss": 0.8136,
+      "step": 2515
+    },
+    {
+      "epoch": 0.22364444444444445,
+      "grad_norm": 2.875,
+      "learning_rate": 1.4909629629629632e-05,
+      "loss": 0.8682,
+      "step": 2516
+    },
+    {
+      "epoch": 0.22373333333333334,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.4915555555555556e-05,
+      "loss": 0.7303,
+      "step": 2517
+    },
+    {
+      "epoch": 0.22382222222222223,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.4921481481481483e-05,
+      "loss": 0.6368,
+      "step": 2518
+    },
+    {
+      "epoch": 0.22391111111111112,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4927407407407407e-05,
+      "loss": 0.6807,
+      "step": 2519
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.4933333333333335e-05,
+      "loss": 0.9431,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2240888888888889,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.493925925925926e-05,
+      "loss": 0.8318,
+      "step": 2521
+    },
+    {
+      "epoch": 0.22417777777777778,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.4945185185185186e-05,
+      "loss": 0.5958,
+      "step": 2522
+    },
+    {
+      "epoch": 0.22426666666666667,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.4951111111111112e-05,
+      "loss": 0.7041,
+      "step": 2523
+    },
+    {
+      "epoch": 0.22435555555555556,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.495703703703704e-05,
+      "loss": 0.6345,
+      "step": 2524
+    },
+    {
+      "epoch": 0.22444444444444445,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.4962962962962964e-05,
+      "loss": 0.7115,
+      "step": 2525
+    },
+    {
+      "epoch": 0.22453333333333333,
+      "grad_norm": 2.125,
+      "learning_rate": 1.4968888888888891e-05,
+      "loss": 0.7903,
+      "step": 2526
+    },
+    {
+      "epoch": 0.22462222222222222,
+      "grad_norm": 2.5,
+      "learning_rate": 1.4974814814814815e-05,
+      "loss": 0.8929,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2247111111111111,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.4980740740740743e-05,
+      "loss": 0.7448,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2248,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.4986666666666667e-05,
+      "loss": 0.788,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2248888888888889,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4992592592592594e-05,
+      "loss": 0.6044,
+      "step": 2530
+    },
+    {
+      "epoch": 0.22497777777777778,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.4998518518518518e-05,
+      "loss": 0.8146,
+      "step": 2531
+    },
+    {
+      "epoch": 0.22506666666666666,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.5004444444444446e-05,
+      "loss": 0.7068,
+      "step": 2532
+    },
+    {
+      "epoch": 0.22515555555555555,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.5010370370370371e-05,
+      "loss": 0.8569,
+      "step": 2533
+    },
+    {
+      "epoch": 0.22524444444444444,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5016296296296297e-05,
+      "loss": 0.7058,
+      "step": 2534
+    },
+    {
+      "epoch": 0.22533333333333333,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.5022222222222223e-05,
+      "loss": 0.6364,
+      "step": 2535
+    },
+    {
+      "epoch": 0.22542222222222222,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.502814814814815e-05,
+      "loss": 0.7134,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2255111111111111,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.5034074074074074e-05,
+      "loss": 0.6766,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2256,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.5040000000000002e-05,
+      "loss": 0.6686,
+      "step": 2538
+    },
+    {
+      "epoch": 0.22568888888888888,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.5045925925925927e-05,
+      "loss": 0.7501,
+      "step": 2539
+    },
+    {
+      "epoch": 0.22577777777777777,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.5051851851851853e-05,
+      "loss": 0.6289,
+      "step": 2540
+    },
+    {
+      "epoch": 0.22586666666666666,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.505777777777778e-05,
+      "loss": 0.7716,
+      "step": 2541
+    },
+    {
+      "epoch": 0.22595555555555555,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.5063703703703705e-05,
+      "loss": 0.5756,
+      "step": 2542
+    },
+    {
+      "epoch": 0.22604444444444444,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.5069629629629632e-05,
+      "loss": 0.6686,
+      "step": 2543
+    },
+    {
+      "epoch": 0.22613333333333333,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.5075555555555556e-05,
+      "loss": 0.8068,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2262222222222222,
+      "grad_norm": 2.5,
+      "learning_rate": 1.5081481481481484e-05,
+      "loss": 0.7774,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2263111111111111,
+      "grad_norm": 2.0,
+      "learning_rate": 1.5087407407407408e-05,
+      "loss": 0.7397,
+      "step": 2546
+    },
+    {
+      "epoch": 0.2264,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.5093333333333335e-05,
+      "loss": 0.6755,
+      "step": 2547
+    },
+    {
+      "epoch": 0.22648888888888888,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.509925925925926e-05,
+      "loss": 0.6241,
+      "step": 2548
+    },
+    {
+      "epoch": 0.22657777777777777,
+      "grad_norm": 1.75,
+      "learning_rate": 1.5105185185185187e-05,
+      "loss": 0.6456,
+      "step": 2549
+    },
+    {
+      "epoch": 0.22666666666666666,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.5111111111111112e-05,
+      "loss": 0.694,
+      "step": 2550
+    },
+    {
+      "epoch": 0.22675555555555554,
+      "grad_norm": 2.0,
+      "learning_rate": 1.511703703703704e-05,
+      "loss": 0.5783,
+      "step": 2551
+    },
+    {
+      "epoch": 0.22684444444444443,
+      "grad_norm": 1.875,
+      "learning_rate": 1.5122962962962964e-05,
+      "loss": 0.7688,
+      "step": 2552
+    },
+    {
+      "epoch": 0.22693333333333332,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.5128888888888891e-05,
+      "loss": 0.7768,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2270222222222222,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.5134814814814815e-05,
+      "loss": 0.7773,
+      "step": 2554
+    },
+    {
+      "epoch": 0.22711111111111112,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.5140740740740743e-05,
+      "loss": 0.7248,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5146666666666667e-05,
+      "loss": 0.7611,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2272888888888889,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5152592592592594e-05,
+      "loss": 0.6509,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2273777777777778,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.5158518518518518e-05,
+      "loss": 0.7661,
+      "step": 2558
+    },
+    {
+      "epoch": 0.22746666666666668,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.5164444444444446e-05,
+      "loss": 0.7854,
+      "step": 2559
+    },
+    {
+      "epoch": 0.22755555555555557,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.5170370370370371e-05,
+      "loss": 0.6454,
+      "step": 2560
+    },
+    {
+      "epoch": 0.22764444444444445,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.5176296296296297e-05,
+      "loss": 0.7951,
+      "step": 2561
+    },
+    {
+      "epoch": 0.22773333333333334,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.5182222222222223e-05,
+      "loss": 0.7476,
+      "step": 2562
+    },
+    {
+      "epoch": 0.22782222222222223,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.518814814814815e-05,
+      "loss": 0.7904,
+      "step": 2563
+    },
+    {
+      "epoch": 0.22791111111111112,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.5194074074074074e-05,
+      "loss": 0.8357,
+      "step": 2564
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 0.6261,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2280888888888889,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.5205925925925926e-05,
+      "loss": 0.6267,
+      "step": 2566
+    },
+    {
+      "epoch": 0.22817777777777778,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5211851851851853e-05,
+      "loss": 0.709,
+      "step": 2567
+    },
+    {
+      "epoch": 0.22826666666666667,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5217777777777777e-05,
+      "loss": 0.6648,
+      "step": 2568
+    },
+    {
+      "epoch": 0.22835555555555556,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5223703703703705e-05,
+      "loss": 0.7837,
+      "step": 2569
+    },
+    {
+      "epoch": 0.22844444444444445,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.5229629629629632e-05,
+      "loss": 0.6959,
+      "step": 2570
+    },
+    {
+      "epoch": 0.22853333333333334,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.5235555555555556e-05,
+      "loss": 0.7652,
+      "step": 2571
+    },
+    {
+      "epoch": 0.22862222222222223,
+      "grad_norm": 2.375,
+      "learning_rate": 1.5241481481481484e-05,
+      "loss": 0.6629,
+      "step": 2572
+    },
+    {
+      "epoch": 0.22871111111111112,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.5247407407407408e-05,
+      "loss": 0.7991,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2288,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.5253333333333335e-05,
+      "loss": 0.6904,
+      "step": 2574
+    },
+    {
+      "epoch": 0.2288888888888889,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.525925925925926e-05,
+      "loss": 0.5741,
+      "step": 2575
+    },
+    {
+      "epoch": 0.22897777777777778,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.5265185185185187e-05,
+      "loss": 0.7328,
+      "step": 2576
+    },
+    {
+      "epoch": 0.22906666666666667,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.5271111111111112e-05,
+      "loss": 0.8041,
+      "step": 2577
+    },
+    {
+      "epoch": 0.22915555555555556,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.5277037037037038e-05,
+      "loss": 0.7488,
+      "step": 2578
+    },
+    {
+      "epoch": 0.22924444444444445,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.5282962962962964e-05,
+      "loss": 0.733,
+      "step": 2579
+    },
+    {
+      "epoch": 0.22933333333333333,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.528888888888889e-05,
+      "loss": 0.622,
+      "step": 2580
+    },
+    {
+      "epoch": 0.22942222222222222,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.5294814814814815e-05,
+      "loss": 0.6735,
+      "step": 2581
+    },
+    {
+      "epoch": 0.2295111111111111,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.530074074074074e-05,
+      "loss": 0.7148,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2296,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5306666666666667e-05,
+      "loss": 0.7126,
+      "step": 2583
+    },
+    {
+      "epoch": 0.2296888888888889,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.5312592592592596e-05,
+      "loss": 0.6141,
+      "step": 2584
+    },
+    {
+      "epoch": 0.22977777777777778,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.531851851851852e-05,
+      "loss": 0.7636,
+      "step": 2585
+    },
+    {
+      "epoch": 0.22986666666666666,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.5324444444444448e-05,
+      "loss": 0.8065,
+      "step": 2586
+    },
+    {
+      "epoch": 0.22995555555555555,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.533037037037037e-05,
+      "loss": 0.6733,
+      "step": 2587
+    },
+    {
+      "epoch": 0.23004444444444444,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.53362962962963e-05,
+      "loss": 0.7719,
+      "step": 2588
+    },
+    {
+      "epoch": 0.23013333333333333,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.534222222222222e-05,
+      "loss": 0.6081,
+      "step": 2589
+    },
+    {
+      "epoch": 0.23022222222222222,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.534814814814815e-05,
+      "loss": 0.7267,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2303111111111111,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.5354074074074073e-05,
+      "loss": 0.7179,
+      "step": 2591
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.5360000000000002e-05,
+      "loss": 0.6365,
+      "step": 2592
+    },
+    {
+      "epoch": 0.23048888888888888,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5365925925925928e-05,
+      "loss": 0.7005,
+      "step": 2593
+    },
+    {
+      "epoch": 0.23057777777777777,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.5371851851851854e-05,
+      "loss": 0.6717,
+      "step": 2594
+    },
+    {
+      "epoch": 0.23066666666666666,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.537777777777778e-05,
+      "loss": 0.5753,
+      "step": 2595
+    },
+    {
+      "epoch": 0.23075555555555555,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.5383703703703705e-05,
+      "loss": 0.7639,
+      "step": 2596
+    },
+    {
+      "epoch": 0.23084444444444444,
+      "grad_norm": 2.0,
+      "learning_rate": 1.538962962962963e-05,
+      "loss": 0.7543,
+      "step": 2597
+    },
+    {
+      "epoch": 0.23093333333333332,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.5395555555555556e-05,
+      "loss": 0.7447,
+      "step": 2598
+    },
+    {
+      "epoch": 0.2310222222222222,
+      "grad_norm": 2.25,
+      "learning_rate": 1.5401481481481482e-05,
+      "loss": 0.7439,
+      "step": 2599
+    },
+    {
+      "epoch": 0.2311111111111111,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5407407407407408e-05,
+      "loss": 0.8059,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2312,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.5413333333333337e-05,
+      "loss": 0.6128,
+      "step": 2601
+    },
+    {
+      "epoch": 0.23128888888888888,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.541925925925926e-05,
+      "loss": 0.8435,
+      "step": 2602
+    },
+    {
+      "epoch": 0.23137777777777777,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.542518518518519e-05,
+      "loss": 0.8036,
+      "step": 2603
+    },
+    {
+      "epoch": 0.23146666666666665,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.543111111111111e-05,
+      "loss": 0.6241,
+      "step": 2604
+    },
+    {
+      "epoch": 0.23155555555555554,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.543703703703704e-05,
+      "loss": 0.7517,
+      "step": 2605
+    },
+    {
+      "epoch": 0.23164444444444443,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.5442962962962962e-05,
+      "loss": 0.7527,
+      "step": 2606
+    },
+    {
+      "epoch": 0.23173333333333335,
+      "grad_norm": 3.0,
+      "learning_rate": 1.544888888888889e-05,
+      "loss": 0.6249,
+      "step": 2607
+    },
+    {
+      "epoch": 0.23182222222222224,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.5454814814814817e-05,
+      "loss": 0.8608,
+      "step": 2608
+    },
+    {
+      "epoch": 0.23191111111111112,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5460740740740743e-05,
+      "loss": 0.6978,
+      "step": 2609
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.546666666666667e-05,
+      "loss": 0.6973,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2320888888888889,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.5472592592592595e-05,
+      "loss": 0.6248,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2321777777777778,
+      "grad_norm": 2.125,
+      "learning_rate": 1.547851851851852e-05,
+      "loss": 0.7906,
+      "step": 2612
+    },
+    {
+      "epoch": 0.23226666666666668,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.5484444444444446e-05,
+      "loss": 0.565,
+      "step": 2613
+    },
+    {
+      "epoch": 0.23235555555555557,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.5490370370370372e-05,
+      "loss": 0.714,
+      "step": 2614
+    },
+    {
+      "epoch": 0.23244444444444445,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.5496296296296298e-05,
+      "loss": 0.6324,
+      "step": 2615
+    },
+    {
+      "epoch": 0.23253333333333334,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5502222222222223e-05,
+      "loss": 0.5791,
+      "step": 2616
+    },
+    {
+      "epoch": 0.23262222222222223,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.550814814814815e-05,
+      "loss": 0.6162,
+      "step": 2617
+    },
+    {
+      "epoch": 0.23271111111111112,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.5514074074074075e-05,
+      "loss": 0.6695,
+      "step": 2618
+    },
+    {
+      "epoch": 0.2328,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.552e-05,
+      "loss": 0.674,
+      "step": 2619
+    },
+    {
+      "epoch": 0.2328888888888889,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5525925925925926e-05,
+      "loss": 0.7192,
+      "step": 2620
+    },
+    {
+      "epoch": 0.23297777777777778,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.5531851851851852e-05,
+      "loss": 0.7196,
+      "step": 2621
+    },
+    {
+      "epoch": 0.23306666666666667,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.5537777777777778e-05,
+      "loss": 0.8802,
+      "step": 2622
+    },
+    {
+      "epoch": 0.23315555555555556,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.5543703703703707e-05,
+      "loss": 0.6806,
+      "step": 2623
+    },
+    {
+      "epoch": 0.23324444444444445,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.554962962962963e-05,
+      "loss": 0.8065,
+      "step": 2624
+    },
+    {
+      "epoch": 0.23333333333333334,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 0.6775,
+      "step": 2625
+    },
+    {
+      "epoch": 0.23342222222222223,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.556148148148148e-05,
+      "loss": 0.6273,
+      "step": 2626
+    },
+    {
+      "epoch": 0.2335111111111111,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.556740740740741e-05,
+      "loss": 0.9393,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.5573333333333332e-05,
+      "loss": 0.6173,
+      "step": 2628
+    },
+    {
+      "epoch": 0.2336888888888889,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.557925925925926e-05,
+      "loss": 0.6648,
+      "step": 2629
+    },
+    {
+      "epoch": 0.23377777777777778,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.5585185185185187e-05,
+      "loss": 0.6349,
+      "step": 2630
+    },
+    {
+      "epoch": 0.23386666666666667,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.5591111111111113e-05,
+      "loss": 0.8502,
+      "step": 2631
+    },
+    {
+      "epoch": 0.23395555555555556,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.559703703703704e-05,
+      "loss": 0.642,
+      "step": 2632
+    },
+    {
+      "epoch": 0.23404444444444444,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.5602962962962964e-05,
+      "loss": 0.7088,
+      "step": 2633
+    },
+    {
+      "epoch": 0.23413333333333333,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.560888888888889e-05,
+      "loss": 0.634,
+      "step": 2634
+    },
+    {
+      "epoch": 0.23422222222222222,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5614814814814816e-05,
+      "loss": 0.6708,
+      "step": 2635
+    },
+    {
+      "epoch": 0.2343111111111111,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.562074074074074e-05,
+      "loss": 0.7851,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2344,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.5626666666666667e-05,
+      "loss": 0.7887,
+      "step": 2637
+    },
+    {
+      "epoch": 0.23448888888888889,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.5632592592592596e-05,
+      "loss": 0.6266,
+      "step": 2638
+    },
+    {
+      "epoch": 0.23457777777777777,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.563851851851852e-05,
+      "loss": 0.6928,
+      "step": 2639
+    },
+    {
+      "epoch": 0.23466666666666666,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.5644444444444448e-05,
+      "loss": 0.9342,
+      "step": 2640
+    },
+    {
+      "epoch": 0.23475555555555555,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.565037037037037e-05,
+      "loss": 0.709,
+      "step": 2641
+    },
+    {
+      "epoch": 0.23484444444444444,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.56562962962963e-05,
+      "loss": 0.7462,
+      "step": 2642
+    },
+    {
+      "epoch": 0.23493333333333333,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.5662222222222222e-05,
+      "loss": 0.7983,
+      "step": 2643
+    },
+    {
+      "epoch": 0.23502222222222222,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.566814814814815e-05,
+      "loss": 0.5846,
+      "step": 2644
+    },
+    {
+      "epoch": 0.2351111111111111,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.5674074074074073e-05,
+      "loss": 0.7202,
+      "step": 2645
+    },
+    {
+      "epoch": 0.2352,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.5680000000000002e-05,
+      "loss": 0.6156,
+      "step": 2646
+    },
+    {
+      "epoch": 0.23528888888888888,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.5685925925925928e-05,
+      "loss": 0.7538,
+      "step": 2647
+    },
+    {
+      "epoch": 0.23537777777777777,
+      "grad_norm": 2.25,
+      "learning_rate": 1.5691851851851854e-05,
+      "loss": 0.7851,
+      "step": 2648
+    },
+    {
+      "epoch": 0.23546666666666666,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.569777777777778e-05,
+      "loss": 0.6461,
+      "step": 2649
+    },
+    {
+      "epoch": 0.23555555555555555,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.5703703703703705e-05,
+      "loss": 0.5368,
+      "step": 2650
+    },
+    {
+      "epoch": 0.23564444444444443,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.570962962962963e-05,
+      "loss": 0.9002,
+      "step": 2651
+    },
+    {
+      "epoch": 0.23573333333333332,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.5715555555555557e-05,
+      "loss": 0.7082,
+      "step": 2652
+    },
+    {
+      "epoch": 0.2358222222222222,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.5721481481481483e-05,
+      "loss": 0.6791,
+      "step": 2653
+    },
+    {
+      "epoch": 0.2359111111111111,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.5727407407407408e-05,
+      "loss": 0.7479,
+      "step": 2654
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 2.25,
+      "learning_rate": 1.5733333333333334e-05,
+      "loss": 0.626,
+      "step": 2655
+    },
+    {
+      "epoch": 0.23608888888888888,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.573925925925926e-05,
+      "loss": 0.7143,
+      "step": 2656
+    },
+    {
+      "epoch": 0.23617777777777776,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.5745185185185186e-05,
+      "loss": 0.522,
+      "step": 2657
+    },
+    {
+      "epoch": 0.23626666666666668,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.575111111111111e-05,
+      "loss": 0.5763,
+      "step": 2658
+    },
+    {
+      "epoch": 0.23635555555555557,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5757037037037037e-05,
+      "loss": 0.7631,
+      "step": 2659
+    },
+    {
+      "epoch": 0.23644444444444446,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.5762962962962963e-05,
+      "loss": 0.5851,
+      "step": 2660
+    },
+    {
+      "epoch": 0.23653333333333335,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.576888888888889e-05,
+      "loss": 0.7431,
+      "step": 2661
+    },
+    {
+      "epoch": 0.23662222222222223,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.5774814814814818e-05,
+      "loss": 0.7455,
+      "step": 2662
+    },
+    {
+      "epoch": 0.23671111111111112,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.5780740740740743e-05,
+      "loss": 0.7731,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.578666666666667e-05,
+      "loss": 0.6657,
+      "step": 2664
+    },
+    {
+      "epoch": 0.2368888888888889,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.5792592592592595e-05,
+      "loss": 0.6485,
+      "step": 2665
+    },
+    {
+      "epoch": 0.2369777777777778,
+      "grad_norm": 2.125,
+      "learning_rate": 1.579851851851852e-05,
+      "loss": 0.7235,
+      "step": 2666
+    },
+    {
+      "epoch": 0.23706666666666668,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.5804444444444446e-05,
+      "loss": 0.7642,
+      "step": 2667
+    },
+    {
+      "epoch": 0.23715555555555556,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.5810370370370372e-05,
+      "loss": 0.8741,
+      "step": 2668
+    },
+    {
+      "epoch": 0.23724444444444445,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.5816296296296298e-05,
+      "loss": 0.7383,
+      "step": 2669
+    },
+    {
+      "epoch": 0.23733333333333334,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.5822222222222224e-05,
+      "loss": 0.767,
+      "step": 2670
+    },
+    {
+      "epoch": 0.23742222222222223,
+      "grad_norm": 2.0,
+      "learning_rate": 1.582814814814815e-05,
+      "loss": 0.5641,
+      "step": 2671
+    },
+    {
+      "epoch": 0.23751111111111112,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.5834074074074075e-05,
+      "loss": 0.6924,
+      "step": 2672
+    },
+    {
+      "epoch": 0.2376,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.584e-05,
+      "loss": 0.904,
+      "step": 2673
+    },
+    {
+      "epoch": 0.2376888888888889,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.5845925925925927e-05,
+      "loss": 0.6955,
+      "step": 2674
+    },
+    {
+      "epoch": 0.23777777777777778,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.5851851851851852e-05,
+      "loss": 0.7203,
+      "step": 2675
+    },
+    {
+      "epoch": 0.23786666666666667,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.5857777777777778e-05,
+      "loss": 0.9106,
+      "step": 2676
+    },
+    {
+      "epoch": 0.23795555555555556,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.5863703703703707e-05,
+      "loss": 0.9346,
+      "step": 2677
+    },
+    {
+      "epoch": 0.23804444444444445,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.586962962962963e-05,
+      "loss": 0.72,
+      "step": 2678
+    },
+    {
+      "epoch": 0.23813333333333334,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.587555555555556e-05,
+      "loss": 0.6498,
+      "step": 2679
+    },
+    {
+      "epoch": 0.23822222222222222,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.588148148148148e-05,
+      "loss": 0.9278,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2383111111111111,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.588740740740741e-05,
+      "loss": 0.6443,
+      "step": 2681
+    },
+    {
+      "epoch": 0.2384,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.5893333333333333e-05,
+      "loss": 0.5959,
+      "step": 2682
+    },
+    {
+      "epoch": 0.2384888888888889,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.589925925925926e-05,
+      "loss": 0.738,
+      "step": 2683
+    },
+    {
+      "epoch": 0.23857777777777778,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.5905185185185187e-05,
+      "loss": 0.7249,
+      "step": 2684
+    },
+    {
+      "epoch": 0.23866666666666667,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.5911111111111113e-05,
+      "loss": 0.7346,
+      "step": 2685
+    },
+    {
+      "epoch": 0.23875555555555555,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.591703703703704e-05,
+      "loss": 0.8582,
+      "step": 2686
+    },
+    {
+      "epoch": 0.23884444444444444,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.5922962962962965e-05,
+      "loss": 0.825,
+      "step": 2687
+    },
+    {
+      "epoch": 0.23893333333333333,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.592888888888889e-05,
+      "loss": 0.8407,
+      "step": 2688
+    },
+    {
+      "epoch": 0.23902222222222222,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.5934814814814816e-05,
+      "loss": 0.6994,
+      "step": 2689
+    },
+    {
+      "epoch": 0.2391111111111111,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.5940740740740742e-05,
+      "loss": 0.6519,
+      "step": 2690
+    },
+    {
+      "epoch": 0.2392,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.5946666666666668e-05,
+      "loss": 0.6329,
+      "step": 2691
+    },
+    {
+      "epoch": 0.23928888888888888,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.5952592592592593e-05,
+      "loss": 0.7681,
+      "step": 2692
+    },
+    {
+      "epoch": 0.23937777777777777,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.595851851851852e-05,
+      "loss": 0.9749,
+      "step": 2693
+    },
+    {
+      "epoch": 0.23946666666666666,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.5964444444444448e-05,
+      "loss": 0.6799,
+      "step": 2694
+    },
+    {
+      "epoch": 0.23955555555555555,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.597037037037037e-05,
+      "loss": 0.8716,
+      "step": 2695
+    },
+    {
+      "epoch": 0.23964444444444444,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.59762962962963e-05,
+      "loss": 0.7736,
+      "step": 2696
+    },
+    {
+      "epoch": 0.23973333333333333,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.5982222222222222e-05,
+      "loss": 0.6548,
+      "step": 2697
+    },
+    {
+      "epoch": 0.23982222222222221,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.598814814814815e-05,
+      "loss": 0.5791,
+      "step": 2698
+    },
+    {
+      "epoch": 0.2399111111111111,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5994074074074074e-05,
+      "loss": 0.7358,
+      "step": 2699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.5852,
+      "step": 2700
+    },
+    {
+      "epoch": 0.24008888888888888,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.600592592592593e-05,
+      "loss": 0.7811,
+      "step": 2701
+    },
+    {
+      "epoch": 0.24017777777777777,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.6011851851851854e-05,
+      "loss": 0.6446,
+      "step": 2702
+    },
+    {
+      "epoch": 0.24026666666666666,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.601777777777778e-05,
+      "loss": 0.7076,
+      "step": 2703
+    },
+    {
+      "epoch": 0.24035555555555554,
+      "grad_norm": 2.125,
+      "learning_rate": 1.6023703703703706e-05,
+      "loss": 0.7098,
+      "step": 2704
+    },
+    {
+      "epoch": 0.24044444444444443,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.602962962962963e-05,
+      "loss": 0.854,
+      "step": 2705
+    },
+    {
+      "epoch": 0.24053333333333332,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.6035555555555557e-05,
+      "loss": 0.7941,
+      "step": 2706
+    },
+    {
+      "epoch": 0.2406222222222222,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.6041481481481483e-05,
+      "loss": 0.727,
+      "step": 2707
+    },
+    {
+      "epoch": 0.2407111111111111,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.604740740740741e-05,
+      "loss": 0.5752,
+      "step": 2708
+    },
+    {
+      "epoch": 0.2408,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.6053333333333334e-05,
+      "loss": 0.7161,
+      "step": 2709
+    },
+    {
+      "epoch": 0.2408888888888889,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.605925925925926e-05,
+      "loss": 0.808,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2409777777777778,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.6065185185185186e-05,
+      "loss": 0.5897,
+      "step": 2711
+    },
+    {
+      "epoch": 0.24106666666666668,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.607111111111111e-05,
+      "loss": 0.7065,
+      "step": 2712
+    },
+    {
+      "epoch": 0.24115555555555557,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.6077037037037037e-05,
+      "loss": 0.7988,
+      "step": 2713
+    },
+    {
+      "epoch": 0.24124444444444446,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.6082962962962963e-05,
+      "loss": 0.597,
+      "step": 2714
+    },
+    {
+      "epoch": 0.24133333333333334,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.608888888888889e-05,
+      "loss": 0.7603,
+      "step": 2715
+    },
+    {
+      "epoch": 0.24142222222222223,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.6094814814814818e-05,
+      "loss": 0.5862,
+      "step": 2716
+    },
+    {
+      "epoch": 0.24151111111111112,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.610074074074074e-05,
+      "loss": 0.6924,
+      "step": 2717
+    },
+    {
+      "epoch": 0.2416,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.610666666666667e-05,
+      "loss": 0.6957,
+      "step": 2718
+    },
+    {
+      "epoch": 0.2416888888888889,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.6112592592592592e-05,
+      "loss": 0.8397,
+      "step": 2719
+    },
+    {
+      "epoch": 0.24177777777777779,
+      "grad_norm": 2.375,
+      "learning_rate": 1.611851851851852e-05,
+      "loss": 0.6917,
+      "step": 2720
+    },
+    {
+      "epoch": 0.24186666666666667,
+      "grad_norm": 2.125,
+      "learning_rate": 1.6124444444444443e-05,
+      "loss": 0.662,
+      "step": 2721
+    },
+    {
+      "epoch": 0.24195555555555556,
+      "grad_norm": 2.125,
+      "learning_rate": 1.6130370370370372e-05,
+      "loss": 0.8046,
+      "step": 2722
+    },
+    {
+      "epoch": 0.24204444444444445,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.6136296296296298e-05,
+      "loss": 0.5749,
+      "step": 2723
+    },
+    {
+      "epoch": 0.24213333333333334,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.6142222222222224e-05,
+      "loss": 0.7701,
+      "step": 2724
+    },
+    {
+      "epoch": 0.24222222222222223,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.614814814814815e-05,
+      "loss": 0.6927,
+      "step": 2725
+    },
+    {
+      "epoch": 0.24231111111111112,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6154074074074075e-05,
+      "loss": 0.583,
+      "step": 2726
+    },
+    {
+      "epoch": 0.2424,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.616e-05,
+      "loss": 0.7611,
+      "step": 2727
+    },
+    {
+      "epoch": 0.2424888888888889,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.6165925925925927e-05,
+      "loss": 0.7336,
+      "step": 2728
+    },
+    {
+      "epoch": 0.24257777777777778,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6171851851851853e-05,
+      "loss": 0.6826,
+      "step": 2729
+    },
+    {
+      "epoch": 0.24266666666666667,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.617777777777778e-05,
+      "loss": 0.7011,
+      "step": 2730
+    },
+    {
+      "epoch": 0.24275555555555556,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.6183703703703707e-05,
+      "loss": 0.6066,
+      "step": 2731
+    },
+    {
+      "epoch": 0.24284444444444445,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.618962962962963e-05,
+      "loss": 0.7797,
+      "step": 2732
+    },
+    {
+      "epoch": 0.24293333333333333,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.619555555555556e-05,
+      "loss": 0.847,
+      "step": 2733
+    },
+    {
+      "epoch": 0.24302222222222222,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.620148148148148e-05,
+      "loss": 0.7568,
+      "step": 2734
+    },
+    {
+      "epoch": 0.2431111111111111,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.620740740740741e-05,
+      "loss": 0.5929,
+      "step": 2735
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.6213333333333333e-05,
+      "loss": 0.7596,
+      "step": 2736
+    },
+    {
+      "epoch": 0.2432888888888889,
+      "grad_norm": 2.625,
+      "learning_rate": 1.6219259259259262e-05,
+      "loss": 0.8088,
+      "step": 2737
+    },
+    {
+      "epoch": 0.24337777777777778,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.6225185185185188e-05,
+      "loss": 0.6728,
+      "step": 2738
+    },
+    {
+      "epoch": 0.24346666666666666,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.6231111111111113e-05,
+      "loss": 0.6064,
+      "step": 2739
+    },
+    {
+      "epoch": 0.24355555555555555,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.623703703703704e-05,
+      "loss": 0.7332,
+      "step": 2740
+    },
+    {
+      "epoch": 0.24364444444444444,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.6242962962962965e-05,
+      "loss": 0.7608,
+      "step": 2741
+    },
+    {
+      "epoch": 0.24373333333333333,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.624888888888889e-05,
+      "loss": 0.6868,
+      "step": 2742
+    },
+    {
+      "epoch": 0.24382222222222222,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.6254814814814816e-05,
+      "loss": 0.752,
+      "step": 2743
+    },
+    {
+      "epoch": 0.2439111111111111,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.6260740740740742e-05,
+      "loss": 0.6398,
+      "step": 2744
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.6266666666666668e-05,
+      "loss": 0.6721,
+      "step": 2745
+    },
+    {
+      "epoch": 0.24408888888888888,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.6272592592592594e-05,
+      "loss": 0.7385,
+      "step": 2746
+    },
+    {
+      "epoch": 0.24417777777777777,
+      "grad_norm": 2.25,
+      "learning_rate": 1.627851851851852e-05,
+      "loss": 0.8009,
+      "step": 2747
+    },
+    {
+      "epoch": 0.24426666666666666,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.6284444444444445e-05,
+      "loss": 0.7028,
+      "step": 2748
+    },
+    {
+      "epoch": 0.24435555555555555,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.629037037037037e-05,
+      "loss": 0.537,
+      "step": 2749
+    },
+    {
+      "epoch": 0.24444444444444444,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.6296296296296297e-05,
+      "loss": 0.7855,
+      "step": 2750
+    },
+    {
+      "epoch": 0.24453333333333332,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.6302222222222222e-05,
+      "loss": 0.7176,
+      "step": 2751
+    },
+    {
+      "epoch": 0.2446222222222222,
+      "grad_norm": 2.125,
+      "learning_rate": 1.6308148148148148e-05,
+      "loss": 0.7146,
+      "step": 2752
+    },
+    {
+      "epoch": 0.2447111111111111,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.6314074074074074e-05,
+      "loss": 0.6424,
+      "step": 2753
+    },
+    {
+      "epoch": 0.2448,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.632e-05,
+      "loss": 0.6761,
+      "step": 2754
+    },
+    {
+      "epoch": 0.24488888888888888,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.632592592592593e-05,
+      "loss": 0.6479,
+      "step": 2755
+    },
+    {
+      "epoch": 0.24497777777777777,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.6331851851851854e-05,
+      "loss": 0.6479,
+      "step": 2756
+    },
+    {
+      "epoch": 0.24506666666666665,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.633777777777778e-05,
+      "loss": 0.7997,
+      "step": 2757
+    },
+    {
+      "epoch": 0.24515555555555554,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.6343703703703706e-05,
+      "loss": 0.6931,
+      "step": 2758
+    },
+    {
+      "epoch": 0.24524444444444443,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.6349629629629632e-05,
+      "loss": 0.7187,
+      "step": 2759
+    },
+    {
+      "epoch": 0.24533333333333332,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.6355555555555557e-05,
+      "loss": 0.7105,
+      "step": 2760
+    },
+    {
+      "epoch": 0.24542222222222224,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.6361481481481483e-05,
+      "loss": 0.6421,
+      "step": 2761
+    },
+    {
+      "epoch": 0.24551111111111112,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.636740740740741e-05,
+      "loss": 0.8023,
+      "step": 2762
+    },
+    {
+      "epoch": 0.2456,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.6373333333333335e-05,
+      "loss": 0.7849,
+      "step": 2763
+    },
+    {
+      "epoch": 0.2456888888888889,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.637925925925926e-05,
+      "loss": 0.6791,
+      "step": 2764
+    },
+    {
+      "epoch": 0.2457777777777778,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.6385185185185186e-05,
+      "loss": 0.8159,
+      "step": 2765
+    },
+    {
+      "epoch": 0.24586666666666668,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.6391111111111112e-05,
+      "loss": 0.7968,
+      "step": 2766
+    },
+    {
+      "epoch": 0.24595555555555557,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6397037037037038e-05,
+      "loss": 0.7417,
+      "step": 2767
+    },
+    {
+      "epoch": 0.24604444444444445,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.6402962962962963e-05,
+      "loss": 0.5653,
+      "step": 2768
+    },
+    {
+      "epoch": 0.24613333333333334,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.640888888888889e-05,
+      "loss": 0.8405,
+      "step": 2769
+    },
+    {
+      "epoch": 0.24622222222222223,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.6414814814814818e-05,
+      "loss": 0.8244,
+      "step": 2770
+    },
+    {
+      "epoch": 0.24631111111111112,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.642074074074074e-05,
+      "loss": 0.7185,
+      "step": 2771
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.642666666666667e-05,
+      "loss": 0.6471,
+      "step": 2772
+    },
+    {
+      "epoch": 0.2464888888888889,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6432592592592592e-05,
+      "loss": 0.6455,
+      "step": 2773
+    },
+    {
+      "epoch": 0.24657777777777778,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.643851851851852e-05,
+      "loss": 0.7083,
+      "step": 2774
+    },
+    {
+      "epoch": 0.24666666666666667,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.6444444444444444e-05,
+      "loss": 0.6705,
+      "step": 2775
+    },
+    {
+      "epoch": 0.24675555555555556,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.6450370370370373e-05,
+      "loss": 0.5559,
+      "step": 2776
+    },
+    {
+      "epoch": 0.24684444444444445,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.64562962962963e-05,
+      "loss": 0.6443,
+      "step": 2777
+    },
+    {
+      "epoch": 0.24693333333333334,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.6462222222222224e-05,
+      "loss": 0.8223,
+      "step": 2778
+    },
+    {
+      "epoch": 0.24702222222222223,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.646814814814815e-05,
+      "loss": 0.6136,
+      "step": 2779
+    },
+    {
+      "epoch": 0.24711111111111111,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.6474074074074076e-05,
+      "loss": 1.0016,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2472,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.648e-05,
+      "loss": 0.7178,
+      "step": 2781
+    },
+    {
+      "epoch": 0.2472888888888889,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.6485925925925927e-05,
+      "loss": 0.6429,
+      "step": 2782
+    },
+    {
+      "epoch": 0.24737777777777778,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.6491851851851853e-05,
+      "loss": 0.8031,
+      "step": 2783
+    },
+    {
+      "epoch": 0.24746666666666667,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.649777777777778e-05,
+      "loss": 0.7108,
+      "step": 2784
+    },
+    {
+      "epoch": 0.24755555555555556,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.6503703703703704e-05,
+      "loss": 0.7851,
+      "step": 2785
+    },
+    {
+      "epoch": 0.24764444444444444,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.650962962962963e-05,
+      "loss": 0.6592,
+      "step": 2786
+    },
+    {
+      "epoch": 0.24773333333333333,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.651555555555556e-05,
+      "loss": 0.5798,
+      "step": 2787
+    },
+    {
+      "epoch": 0.24782222222222222,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.652148148148148e-05,
+      "loss": 0.7289,
+      "step": 2788
+    },
+    {
+      "epoch": 0.2479111111111111,
+      "grad_norm": 2.5,
+      "learning_rate": 1.652740740740741e-05,
+      "loss": 0.8571,
+      "step": 2789
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.6533333333333333e-05,
+      "loss": 0.5713,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2480888888888889,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.6539259259259262e-05,
+      "loss": 0.6913,
+      "step": 2791
+    },
+    {
+      "epoch": 0.24817777777777777,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.6545185185185188e-05,
+      "loss": 0.5658,
+      "step": 2792
+    },
+    {
+      "epoch": 0.24826666666666666,
+      "grad_norm": 2.375,
+      "learning_rate": 1.6551111111111114e-05,
+      "loss": 0.6941,
+      "step": 2793
+    },
+    {
+      "epoch": 0.24835555555555555,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.655703703703704e-05,
+      "loss": 0.7235,
+      "step": 2794
+    },
+    {
+      "epoch": 0.24844444444444444,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.6562962962962965e-05,
+      "loss": 0.9089,
+      "step": 2795
+    },
+    {
+      "epoch": 0.24853333333333333,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.656888888888889e-05,
+      "loss": 0.6457,
+      "step": 2796
+    },
+    {
+      "epoch": 0.24862222222222222,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.6574814814814817e-05,
+      "loss": 0.6914,
+      "step": 2797
+    },
+    {
+      "epoch": 0.2487111111111111,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.6580740740740742e-05,
+      "loss": 0.637,
+      "step": 2798
+    },
+    {
+      "epoch": 0.2488,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.6586666666666668e-05,
+      "loss": 0.8375,
+      "step": 2799
+    },
+    {
+      "epoch": 0.24888888888888888,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.6592592592592594e-05,
+      "loss": 0.7575,
+      "step": 2800
+    },
+    {
+      "epoch": 0.24897777777777777,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.659851851851852e-05,
+      "loss": 0.6379,
+      "step": 2801
+    },
+    {
+      "epoch": 0.24906666666666666,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.6604444444444445e-05,
+      "loss": 0.6691,
+      "step": 2802
+    },
+    {
+      "epoch": 0.24915555555555555,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.661037037037037e-05,
+      "loss": 0.7758,
+      "step": 2803
+    },
+    {
+      "epoch": 0.24924444444444444,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.6616296296296297e-05,
+      "loss": 0.7447,
+      "step": 2804
+    },
+    {
+      "epoch": 0.24933333333333332,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6622222222222223e-05,
+      "loss": 0.7489,
+      "step": 2805
+    },
+    {
+      "epoch": 0.2494222222222222,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.662814814814815e-05,
+      "loss": 0.6395,
+      "step": 2806
+    },
+    {
+      "epoch": 0.2495111111111111,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.6634074074074074e-05,
+      "loss": 0.6045,
+      "step": 2807
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.664e-05,
+      "loss": 0.5685,
+      "step": 2808
+    },
+    {
+      "epoch": 0.24968888888888888,
+      "grad_norm": 4.125,
+      "learning_rate": 1.664592592592593e-05,
+      "loss": 0.7053,
+      "step": 2809
+    },
+    {
+      "epoch": 0.24977777777777777,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.665185185185185e-05,
+      "loss": 0.67,
+      "step": 2810
+    },
+    {
+      "epoch": 0.24986666666666665,
+      "grad_norm": 1.875,
+      "learning_rate": 1.665777777777778e-05,
+      "loss": 0.6306,
+      "step": 2811
+    },
+    {
+      "epoch": 0.24995555555555554,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.6663703703703703e-05,
+      "loss": 0.6804,
+      "step": 2812
+    },
+    {
+      "epoch": 0.25004444444444446,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.6669629629629632e-05,
+      "loss": 0.7386,
+      "step": 2813
+    },
+    {
+      "epoch": 0.2501333333333333,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.6675555555555554e-05,
+      "loss": 0.741,
+      "step": 2814
+    },
+    {
+      "epoch": 0.25022222222222223,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.6681481481481484e-05,
+      "loss": 0.7254,
+      "step": 2815
+    },
+    {
+      "epoch": 0.2503111111111111,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.668740740740741e-05,
+      "loss": 0.7742,
+      "step": 2816
+    },
+    {
+      "epoch": 0.2504,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.6693333333333335e-05,
+      "loss": 0.6965,
+      "step": 2817
+    },
+    {
+      "epoch": 0.25048888888888887,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.669925925925926e-05,
+      "loss": 0.7117,
+      "step": 2818
+    },
+    {
+      "epoch": 0.2505777777777778,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.6705185185185186e-05,
+      "loss": 0.6298,
+      "step": 2819
+    },
+    {
+      "epoch": 0.25066666666666665,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.6711111111111112e-05,
+      "loss": 0.7588,
+      "step": 2820
+    },
+    {
+      "epoch": 0.25075555555555556,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6717037037037038e-05,
+      "loss": 0.7613,
+      "step": 2821
+    },
+    {
+      "epoch": 0.2508444444444444,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.6722962962962964e-05,
+      "loss": 0.6932,
+      "step": 2822
+    },
+    {
+      "epoch": 0.25093333333333334,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.672888888888889e-05,
+      "loss": 0.793,
+      "step": 2823
+    },
+    {
+      "epoch": 0.2510222222222222,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.673481481481482e-05,
+      "loss": 0.8297,
+      "step": 2824
+    },
+    {
+      "epoch": 0.2511111111111111,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.674074074074074e-05,
+      "loss": 0.6113,
+      "step": 2825
+    },
+    {
+      "epoch": 0.2512,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.674666666666667e-05,
+      "loss": 0.6791,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2512888888888889,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.6752592592592592e-05,
+      "loss": 0.7463,
+      "step": 2827
+    },
+    {
+      "epoch": 0.25137777777777776,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.675851851851852e-05,
+      "loss": 0.6369,
+      "step": 2828
+    },
+    {
+      "epoch": 0.25146666666666667,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.6764444444444444e-05,
+      "loss": 0.7753,
+      "step": 2829
+    },
+    {
+      "epoch": 0.25155555555555553,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.6770370370370373e-05,
+      "loss": 0.6157,
+      "step": 2830
+    },
+    {
+      "epoch": 0.25164444444444445,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.67762962962963e-05,
+      "loss": 0.5726,
+      "step": 2831
+    },
+    {
+      "epoch": 0.2517333333333333,
+      "grad_norm": 2.25,
+      "learning_rate": 1.6782222222222225e-05,
+      "loss": 0.7558,
+      "step": 2832
+    },
+    {
+      "epoch": 0.2518222222222222,
+      "grad_norm": 2.5,
+      "learning_rate": 1.678814814814815e-05,
+      "loss": 0.7319,
+      "step": 2833
+    },
+    {
+      "epoch": 0.2519111111111111,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.6794074074074076e-05,
+      "loss": 0.7487,
+      "step": 2834
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.7047,
+      "step": 2835
+    },
+    {
+      "epoch": 0.25208888888888886,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.6805925925925928e-05,
+      "loss": 0.709,
+      "step": 2836
+    },
+    {
+      "epoch": 0.2521777777777778,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.6811851851851853e-05,
+      "loss": 0.6331,
+      "step": 2837
+    },
+    {
+      "epoch": 0.25226666666666664,
+      "grad_norm": 2.375,
+      "learning_rate": 1.681777777777778e-05,
+      "loss": 0.856,
+      "step": 2838
+    },
+    {
+      "epoch": 0.25235555555555556,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.6823703703703705e-05,
+      "loss": 0.7369,
+      "step": 2839
+    },
+    {
+      "epoch": 0.25244444444444447,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.682962962962963e-05,
+      "loss": 0.8584,
+      "step": 2840
+    },
+    {
+      "epoch": 0.25253333333333333,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.6835555555555556e-05,
+      "loss": 0.6087,
+      "step": 2841
+    },
+    {
+      "epoch": 0.25262222222222225,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.6841481481481482e-05,
+      "loss": 0.7186,
+      "step": 2842
+    },
+    {
+      "epoch": 0.2527111111111111,
+      "grad_norm": 2.25,
+      "learning_rate": 1.6847407407407408e-05,
+      "loss": 0.691,
+      "step": 2843
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.6853333333333333e-05,
+      "loss": 0.686,
+      "step": 2844
+    },
+    {
+      "epoch": 0.2528888888888889,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.685925925925926e-05,
+      "loss": 0.7318,
+      "step": 2845
+    },
+    {
+      "epoch": 0.2529777777777778,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.686518518518519e-05,
+      "loss": 0.6267,
+      "step": 2846
+    },
+    {
+      "epoch": 0.25306666666666666,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.687111111111111e-05,
+      "loss": 0.6568,
+      "step": 2847
+    },
+    {
+      "epoch": 0.2531555555555556,
+      "grad_norm": 1.875,
+      "learning_rate": 1.687703703703704e-05,
+      "loss": 0.5779,
+      "step": 2848
+    },
+    {
+      "epoch": 0.25324444444444444,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.6882962962962966e-05,
+      "loss": 0.7223,
+      "step": 2849
+    },
+    {
+      "epoch": 0.25333333333333335,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.688888888888889e-05,
+      "loss": 0.6232,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2534222222222222,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.6894814814814817e-05,
+      "loss": 0.7504,
+      "step": 2851
+    },
+    {
+      "epoch": 0.25351111111111113,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.6900740740740743e-05,
+      "loss": 0.7418,
+      "step": 2852
+    },
+    {
+      "epoch": 0.2536,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.690666666666667e-05,
+      "loss": 0.665,
+      "step": 2853
+    },
+    {
+      "epoch": 0.2536888888888889,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.6912592592592594e-05,
+      "loss": 0.7993,
+      "step": 2854
+    },
+    {
+      "epoch": 0.25377777777777777,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.691851851851852e-05,
+      "loss": 0.734,
+      "step": 2855
+    },
+    {
+      "epoch": 0.2538666666666667,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.6924444444444446e-05,
+      "loss": 0.6745,
+      "step": 2856
+    },
+    {
+      "epoch": 0.25395555555555555,
+      "grad_norm": 2.5,
+      "learning_rate": 1.693037037037037e-05,
+      "loss": 0.6548,
+      "step": 2857
+    },
+    {
+      "epoch": 0.25404444444444446,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.6936296296296297e-05,
+      "loss": 0.6495,
+      "step": 2858
+    },
+    {
+      "epoch": 0.2541333333333333,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.6942222222222223e-05,
+      "loss": 0.7819,
+      "step": 2859
+    },
+    {
+      "epoch": 0.25422222222222224,
+      "grad_norm": 2.25,
+      "learning_rate": 1.694814814814815e-05,
+      "loss": 0.7633,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2543111111111111,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.6954074074074075e-05,
+      "loss": 0.6184,
+      "step": 2861
+    },
+    {
+      "epoch": 0.2544,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.696e-05,
+      "loss": 0.7155,
+      "step": 2862
+    },
+    {
+      "epoch": 0.2544888888888889,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.696592592592593e-05,
+      "loss": 0.5809,
+      "step": 2863
+    },
+    {
+      "epoch": 0.2545777777777778,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.6971851851851852e-05,
+      "loss": 0.605,
+      "step": 2864
+    },
+    {
+      "epoch": 0.25466666666666665,
+      "grad_norm": 2.0,
+      "learning_rate": 1.697777777777778e-05,
+      "loss": 0.7537,
+      "step": 2865
+    },
+    {
+      "epoch": 0.25475555555555557,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.6983703703703703e-05,
+      "loss": 0.6018,
+      "step": 2866
+    },
+    {
+      "epoch": 0.25484444444444443,
+      "grad_norm": 2.375,
+      "learning_rate": 1.6989629629629632e-05,
+      "loss": 0.7927,
+      "step": 2867
+    },
+    {
+      "epoch": 0.25493333333333335,
+      "grad_norm": 2.25,
+      "learning_rate": 1.6995555555555555e-05,
+      "loss": 0.8499,
+      "step": 2868
+    },
+    {
+      "epoch": 0.2550222222222222,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.7001481481481484e-05,
+      "loss": 0.8104,
+      "step": 2869
+    },
+    {
+      "epoch": 0.2551111111111111,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.700740740740741e-05,
+      "loss": 0.5844,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2552,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.7013333333333335e-05,
+      "loss": 0.7872,
+      "step": 2871
+    },
+    {
+      "epoch": 0.2552888888888889,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.701925925925926e-05,
+      "loss": 0.6529,
+      "step": 2872
+    },
+    {
+      "epoch": 0.25537777777777776,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.7025185185185187e-05,
+      "loss": 0.5803,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2554666666666667,
+      "grad_norm": 2.375,
+      "learning_rate": 1.7031111111111113e-05,
+      "loss": 0.637,
+      "step": 2874
+    },
+    {
+      "epoch": 0.25555555555555554,
+      "grad_norm": 3.125,
+      "learning_rate": 1.7037037037037038e-05,
+      "loss": 0.5832,
+      "step": 2875
+    },
+    {
+      "epoch": 0.25564444444444445,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.7042962962962964e-05,
+      "loss": 0.7044,
+      "step": 2876
+    },
+    {
+      "epoch": 0.2557333333333333,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.704888888888889e-05,
+      "loss": 0.7291,
+      "step": 2877
+    },
+    {
+      "epoch": 0.25582222222222223,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.7054814814814816e-05,
+      "loss": 0.8267,
+      "step": 2878
+    },
+    {
+      "epoch": 0.2559111111111111,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.706074074074074e-05,
+      "loss": 0.7412,
+      "step": 2879
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.706666666666667e-05,
+      "loss": 0.7158,
+      "step": 2880
+    },
+    {
+      "epoch": 0.25608888888888887,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.7072592592592593e-05,
+      "loss": 0.7136,
+      "step": 2881
+    },
+    {
+      "epoch": 0.2561777777777778,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.7078518518518522e-05,
+      "loss": 0.5861,
+      "step": 2882
+    },
+    {
+      "epoch": 0.25626666666666664,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.7084444444444444e-05,
+      "loss": 0.7669,
+      "step": 2883
+    },
+    {
+      "epoch": 0.25635555555555556,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.7090370370370373e-05,
+      "loss": 0.6706,
+      "step": 2884
+    },
+    {
+      "epoch": 0.2564444444444444,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.70962962962963e-05,
+      "loss": 0.7536,
+      "step": 2885
+    },
+    {
+      "epoch": 0.25653333333333334,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.7102222222222225e-05,
+      "loss": 0.782,
+      "step": 2886
+    },
+    {
+      "epoch": 0.2566222222222222,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.710814814814815e-05,
+      "loss": 0.6364,
+      "step": 2887
+    },
+    {
+      "epoch": 0.2567111111111111,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.7114074074074076e-05,
+      "loss": 0.8023,
+      "step": 2888
+    },
+    {
+      "epoch": 0.2568,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.7120000000000002e-05,
+      "loss": 0.6355,
+      "step": 2889
+    },
+    {
+      "epoch": 0.2568888888888889,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.7125925925925928e-05,
+      "loss": 0.679,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2569777777777778,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.7131851851851854e-05,
+      "loss": 0.7602,
+      "step": 2891
+    },
+    {
+      "epoch": 0.25706666666666667,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.713777777777778e-05,
+      "loss": 0.646,
+      "step": 2892
+    },
+    {
+      "epoch": 0.2571555555555556,
+      "grad_norm": 2.25,
+      "learning_rate": 1.7143703703703705e-05,
+      "loss": 0.5066,
+      "step": 2893
+    },
+    {
+      "epoch": 0.25724444444444444,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.714962962962963e-05,
+      "loss": 0.5411,
+      "step": 2894
+    },
+    {
+      "epoch": 0.25733333333333336,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.7155555555555557e-05,
+      "loss": 0.5519,
+      "step": 2895
+    },
+    {
+      "epoch": 0.2574222222222222,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.7161481481481482e-05,
+      "loss": 0.7937,
+      "step": 2896
+    },
+    {
+      "epoch": 0.25751111111111114,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.7167407407407408e-05,
+      "loss": 0.8235,
+      "step": 2897
+    },
+    {
+      "epoch": 0.2576,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.7173333333333334e-05,
+      "loss": 0.7988,
+      "step": 2898
+    },
+    {
+      "epoch": 0.2576888888888889,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.717925925925926e-05,
+      "loss": 0.5752,
+      "step": 2899
+    },
+    {
+      "epoch": 0.2577777777777778,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.7185185185185185e-05,
+      "loss": 0.6549,
+      "step": 2900
+    },
+    {
+      "epoch": 0.2578666666666667,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.719111111111111e-05,
+      "loss": 0.5221,
+      "step": 2901
+    },
+    {
+      "epoch": 0.25795555555555555,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.719703703703704e-05,
+      "loss": 0.6634,
+      "step": 2902
+    },
+    {
+      "epoch": 0.25804444444444447,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.7202962962962963e-05,
+      "loss": 0.6277,
+      "step": 2903
+    },
+    {
+      "epoch": 0.2581333333333333,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.720888888888889e-05,
+      "loss": 0.7376,
+      "step": 2904
+    },
+    {
+      "epoch": 0.25822222222222224,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.7214814814814814e-05,
+      "loss": 0.6348,
+      "step": 2905
+    },
+    {
+      "epoch": 0.2583111111111111,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.7220740740740743e-05,
+      "loss": 0.7156,
+      "step": 2906
+    },
+    {
+      "epoch": 0.2584,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.7226666666666665e-05,
+      "loss": 0.8219,
+      "step": 2907
+    },
+    {
+      "epoch": 0.2584888888888889,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.7232592592592595e-05,
+      "loss": 0.8432,
+      "step": 2908
+    },
+    {
+      "epoch": 0.2585777777777778,
+      "grad_norm": 2.25,
+      "learning_rate": 1.723851851851852e-05,
+      "loss": 0.8632,
+      "step": 2909
+    },
+    {
+      "epoch": 0.25866666666666666,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.7244444444444446e-05,
+      "loss": 0.6339,
+      "step": 2910
+    },
+    {
+      "epoch": 0.25875555555555557,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.7250370370370372e-05,
+      "loss": 0.7271,
+      "step": 2911
+    },
+    {
+      "epoch": 0.25884444444444443,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.7256296296296298e-05,
+      "loss": 0.9226,
+      "step": 2912
+    },
+    {
+      "epoch": 0.25893333333333335,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.7262222222222223e-05,
+      "loss": 0.644,
+      "step": 2913
+    },
+    {
+      "epoch": 0.2590222222222222,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.726814814814815e-05,
+      "loss": 0.7377,
+      "step": 2914
+    },
+    {
+      "epoch": 0.2591111111111111,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.7274074074074075e-05,
+      "loss": 0.7977,
+      "step": 2915
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.728e-05,
+      "loss": 0.6793,
+      "step": 2916
+    },
+    {
+      "epoch": 0.2592888888888889,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.728592592592593e-05,
+      "loss": 0.6987,
+      "step": 2917
+    },
+    {
+      "epoch": 0.25937777777777776,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7291851851851852e-05,
+      "loss": 0.6954,
+      "step": 2918
+    },
+    {
+      "epoch": 0.2594666666666667,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.729777777777778e-05,
+      "loss": 0.6095,
+      "step": 2919
+    },
+    {
+      "epoch": 0.25955555555555554,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.7303703703703704e-05,
+      "loss": 0.5838,
+      "step": 2920
+    },
+    {
+      "epoch": 0.25964444444444446,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.7309629629629633e-05,
+      "loss": 0.7395,
+      "step": 2921
+    },
+    {
+      "epoch": 0.2597333333333333,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.7315555555555555e-05,
+      "loss": 0.755,
+      "step": 2922
+    },
+    {
+      "epoch": 0.25982222222222223,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.7321481481481484e-05,
+      "loss": 0.6947,
+      "step": 2923
+    },
+    {
+      "epoch": 0.2599111111111111,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.732740740740741e-05,
+      "loss": 0.6119,
+      "step": 2924
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 0.633,
+      "step": 2925
+    },
+    {
+      "epoch": 0.26008888888888887,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.733925925925926e-05,
+      "loss": 0.7656,
+      "step": 2926
+    },
+    {
+      "epoch": 0.2601777777777778,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.7345185185185187e-05,
+      "loss": 0.8317,
+      "step": 2927
+    },
+    {
+      "epoch": 0.26026666666666665,
+      "grad_norm": 2.375,
+      "learning_rate": 1.7351111111111113e-05,
+      "loss": 0.6042,
+      "step": 2928
+    },
+    {
+      "epoch": 0.26035555555555556,
+      "grad_norm": 2.0,
+      "learning_rate": 1.735703703703704e-05,
+      "loss": 0.8029,
+      "step": 2929
+    },
+    {
+      "epoch": 0.2604444444444444,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.7362962962962964e-05,
+      "loss": 0.7063,
+      "step": 2930
+    },
+    {
+      "epoch": 0.26053333333333334,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.736888888888889e-05,
+      "loss": 0.7241,
+      "step": 2931
+    },
+    {
+      "epoch": 0.2606222222222222,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.7374814814814816e-05,
+      "loss": 0.6852,
+      "step": 2932
+    },
+    {
+      "epoch": 0.2607111111111111,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.738074074074074e-05,
+      "loss": 0.8194,
+      "step": 2933
+    },
+    {
+      "epoch": 0.2608,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.7386666666666667e-05,
+      "loss": 0.6092,
+      "step": 2934
+    },
+    {
+      "epoch": 0.2608888888888889,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.7392592592592593e-05,
+      "loss": 0.8327,
+      "step": 2935
+    },
+    {
+      "epoch": 0.26097777777777775,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.739851851851852e-05,
+      "loss": 0.7482,
+      "step": 2936
+    },
+    {
+      "epoch": 0.26106666666666667,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.7404444444444445e-05,
+      "loss": 0.7569,
+      "step": 2937
+    },
+    {
+      "epoch": 0.26115555555555553,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.741037037037037e-05,
+      "loss": 0.5873,
+      "step": 2938
+    },
+    {
+      "epoch": 0.26124444444444445,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.74162962962963e-05,
+      "loss": 0.6984,
+      "step": 2939
+    },
+    {
+      "epoch": 0.2613333333333333,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.7422222222222222e-05,
+      "loss": 0.7628,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2614222222222222,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.742814814814815e-05,
+      "loss": 0.8653,
+      "step": 2941
+    },
+    {
+      "epoch": 0.26151111111111114,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.7434074074074077e-05,
+      "loss": 0.6002,
+      "step": 2942
+    },
+    {
+      "epoch": 0.2616,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.7440000000000002e-05,
+      "loss": 0.5122,
+      "step": 2943
+    },
+    {
+      "epoch": 0.2616888888888889,
+      "grad_norm": 1.875,
+      "learning_rate": 1.7445925925925928e-05,
+      "loss": 0.7049,
+      "step": 2944
+    },
+    {
+      "epoch": 0.2617777777777778,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.7451851851851854e-05,
+      "loss": 0.5905,
+      "step": 2945
+    },
+    {
+      "epoch": 0.2618666666666667,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.745777777777778e-05,
+      "loss": 0.6491,
+      "step": 2946
+    },
+    {
+      "epoch": 0.26195555555555555,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7463703703703705e-05,
+      "loss": 0.7508,
+      "step": 2947
+    },
+    {
+      "epoch": 0.26204444444444447,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.746962962962963e-05,
+      "loss": 0.854,
+      "step": 2948
+    },
+    {
+      "epoch": 0.26213333333333333,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.7475555555555557e-05,
+      "loss": 0.7108,
+      "step": 2949
+    },
+    {
+      "epoch": 0.26222222222222225,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7481481481481483e-05,
+      "loss": 0.7106,
+      "step": 2950
+    },
+    {
+      "epoch": 0.2623111111111111,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.748740740740741e-05,
+      "loss": 0.7724,
+      "step": 2951
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.7493333333333334e-05,
+      "loss": 0.7272,
+      "step": 2952
+    },
+    {
+      "epoch": 0.2624888888888889,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.749925925925926e-05,
+      "loss": 0.7368,
+      "step": 2953
+    },
+    {
+      "epoch": 0.2625777777777778,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.7505185185185186e-05,
+      "loss": 0.6533,
+      "step": 2954
+    },
+    {
+      "epoch": 0.26266666666666666,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.751111111111111e-05,
+      "loss": 0.7349,
+      "step": 2955
+    },
+    {
+      "epoch": 0.2627555555555556,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.751703703703704e-05,
+      "loss": 0.5954,
+      "step": 2956
+    },
+    {
+      "epoch": 0.26284444444444444,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.7522962962962963e-05,
+      "loss": 0.8371,
+      "step": 2957
+    },
+    {
+      "epoch": 0.26293333333333335,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.7528888888888892e-05,
+      "loss": 0.6856,
+      "step": 2958
+    },
+    {
+      "epoch": 0.2630222222222222,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.7534814814814814e-05,
+      "loss": 0.8006,
+      "step": 2959
+    },
+    {
+      "epoch": 0.26311111111111113,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.7540740740740743e-05,
+      "loss": 0.4642,
+      "step": 2960
+    },
+    {
+      "epoch": 0.2632,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.7546666666666666e-05,
+      "loss": 1.0138,
+      "step": 2961
+    },
+    {
+      "epoch": 0.2632888888888889,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.7552592592592595e-05,
+      "loss": 0.9299,
+      "step": 2962
+    },
+    {
+      "epoch": 0.26337777777777777,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.755851851851852e-05,
+      "loss": 0.5369,
+      "step": 2963
+    },
+    {
+      "epoch": 0.2634666666666667,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7564444444444446e-05,
+      "loss": 0.711,
+      "step": 2964
+    },
+    {
+      "epoch": 0.26355555555555554,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.7570370370370372e-05,
+      "loss": 0.6385,
+      "step": 2965
+    },
+    {
+      "epoch": 0.26364444444444446,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.7576296296296298e-05,
+      "loss": 0.7183,
+      "step": 2966
+    },
+    {
+      "epoch": 0.2637333333333333,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.7582222222222224e-05,
+      "loss": 0.7584,
+      "step": 2967
+    },
+    {
+      "epoch": 0.26382222222222224,
+      "grad_norm": 2.25,
+      "learning_rate": 1.758814814814815e-05,
+      "loss": 0.7106,
+      "step": 2968
+    },
+    {
+      "epoch": 0.2639111111111111,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.7594074074074075e-05,
+      "loss": 0.833,
+      "step": 2969
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.76e-05,
+      "loss": 0.711,
+      "step": 2970
+    },
+    {
+      "epoch": 0.2640888888888889,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.7605925925925927e-05,
+      "loss": 0.6944,
+      "step": 2971
+    },
+    {
+      "epoch": 0.2641777777777778,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.7611851851851852e-05,
+      "loss": 0.7625,
+      "step": 2972
+    },
+    {
+      "epoch": 0.26426666666666665,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.761777777777778e-05,
+      "loss": 0.6704,
+      "step": 2973
+    },
+    {
+      "epoch": 0.26435555555555557,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.7623703703703704e-05,
+      "loss": 0.7563,
+      "step": 2974
+    },
+    {
+      "epoch": 0.2644444444444444,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.7629629629629633e-05,
+      "loss": 0.6958,
+      "step": 2975
+    },
+    {
+      "epoch": 0.26453333333333334,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.7635555555555555e-05,
+      "loss": 0.6802,
+      "step": 2976
+    },
+    {
+      "epoch": 0.2646222222222222,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7641481481481484e-05,
+      "loss": 0.648,
+      "step": 2977
+    },
+    {
+      "epoch": 0.2647111111111111,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.764740740740741e-05,
+      "loss": 0.7295,
+      "step": 2978
+    },
+    {
+      "epoch": 0.2648,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.7653333333333336e-05,
+      "loss": 0.5038,
+      "step": 2979
+    },
+    {
+      "epoch": 0.2648888888888889,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7659259259259262e-05,
+      "loss": 0.5676,
+      "step": 2980
+    },
+    {
+      "epoch": 0.26497777777777776,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.7665185185185187e-05,
+      "loss": 0.675,
+      "step": 2981
+    },
+    {
+      "epoch": 0.2650666666666667,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.7671111111111113e-05,
+      "loss": 0.6837,
+      "step": 2982
+    },
+    {
+      "epoch": 0.26515555555555553,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.767703703703704e-05,
+      "loss": 0.7243,
+      "step": 2983
+    },
+    {
+      "epoch": 0.26524444444444445,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.7682962962962965e-05,
+      "loss": 0.5783,
+      "step": 2984
+    },
+    {
+      "epoch": 0.2653333333333333,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.768888888888889e-05,
+      "loss": 0.7107,
+      "step": 2985
+    },
+    {
+      "epoch": 0.2654222222222222,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.7694814814814816e-05,
+      "loss": 0.6369,
+      "step": 2986
+    },
+    {
+      "epoch": 0.2655111111111111,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.7700740740740742e-05,
+      "loss": 0.6303,
+      "step": 2987
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.7706666666666668e-05,
+      "loss": 0.7872,
+      "step": 2988
+    },
+    {
+      "epoch": 0.26568888888888886,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.7712592592592593e-05,
+      "loss": 0.7929,
+      "step": 2989
+    },
+    {
+      "epoch": 0.2657777777777778,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.771851851851852e-05,
+      "loss": 0.7493,
+      "step": 2990
+    },
+    {
+      "epoch": 0.26586666666666664,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.7724444444444445e-05,
+      "loss": 0.7355,
+      "step": 2991
+    },
+    {
+      "epoch": 0.26595555555555556,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.773037037037037e-05,
+      "loss": 0.6965,
+      "step": 2992
+    },
+    {
+      "epoch": 0.2660444444444444,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.77362962962963e-05,
+      "loss": 0.5736,
+      "step": 2993
+    },
+    {
+      "epoch": 0.26613333333333333,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.7742222222222222e-05,
+      "loss": 0.8154,
+      "step": 2994
+    },
+    {
+      "epoch": 0.26622222222222225,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.774814814814815e-05,
+      "loss": 0.75,
+      "step": 2995
+    },
+    {
+      "epoch": 0.2663111111111111,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.7754074074074074e-05,
+      "loss": 0.7799,
+      "step": 2996
+    },
+    {
+      "epoch": 0.2664,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.7760000000000003e-05,
+      "loss": 0.7409,
+      "step": 2997
+    },
+    {
+      "epoch": 0.2664888888888889,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.7765925925925925e-05,
+      "loss": 0.7978,
+      "step": 2998
+    },
+    {
+      "epoch": 0.2665777777777778,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7771851851851854e-05,
+      "loss": 0.6078,
+      "step": 2999
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.7839,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2667555555555556,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.7783703703703706e-05,
+      "loss": 0.6977,
+      "step": 3001
+    },
+    {
+      "epoch": 0.26684444444444444,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.778962962962963e-05,
+      "loss": 0.6921,
+      "step": 3002
+    },
+    {
+      "epoch": 0.26693333333333336,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.7795555555555557e-05,
+      "loss": 0.7067,
+      "step": 3003
+    },
+    {
+      "epoch": 0.2670222222222222,
+      "grad_norm": 2.125,
+      "learning_rate": 1.7801481481481483e-05,
+      "loss": 0.7892,
+      "step": 3004
+    },
+    {
+      "epoch": 0.26711111111111113,
+      "grad_norm": 1.875,
+      "learning_rate": 1.780740740740741e-05,
+      "loss": 0.711,
+      "step": 3005
+    },
+    {
+      "epoch": 0.2672,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7813333333333334e-05,
+      "loss": 0.5209,
+      "step": 3006
+    },
+    {
+      "epoch": 0.2672888888888889,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.781925925925926e-05,
+      "loss": 0.7081,
+      "step": 3007
+    },
+    {
+      "epoch": 0.26737777777777777,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7825185185185186e-05,
+      "loss": 0.7771,
+      "step": 3008
+    },
+    {
+      "epoch": 0.2674666666666667,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.783111111111111e-05,
+      "loss": 0.6117,
+      "step": 3009
+    },
+    {
+      "epoch": 0.26755555555555555,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.783703703703704e-05,
+      "loss": 0.6615,
+      "step": 3010
+    },
+    {
+      "epoch": 0.26764444444444446,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.7842962962962963e-05,
+      "loss": 0.7292,
+      "step": 3011
+    },
+    {
+      "epoch": 0.2677333333333333,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.7848888888888892e-05,
+      "loss": 0.738,
+      "step": 3012
+    },
+    {
+      "epoch": 0.26782222222222224,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.7854814814814815e-05,
+      "loss": 0.6082,
+      "step": 3013
+    },
+    {
+      "epoch": 0.2679111111111111,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.7860740740740744e-05,
+      "loss": 0.6414,
+      "step": 3014
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.7866666666666666e-05,
+      "loss": 0.6956,
+      "step": 3015
+    },
+    {
+      "epoch": 0.2680888888888889,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.7872592592592595e-05,
+      "loss": 0.6564,
+      "step": 3016
+    },
+    {
+      "epoch": 0.2681777777777778,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.787851851851852e-05,
+      "loss": 0.6649,
+      "step": 3017
+    },
+    {
+      "epoch": 0.26826666666666665,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.7884444444444447e-05,
+      "loss": 0.5731,
+      "step": 3018
+    },
+    {
+      "epoch": 0.26835555555555557,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7890370370370373e-05,
+      "loss": 0.6276,
+      "step": 3019
+    },
+    {
+      "epoch": 0.26844444444444443,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.7896296296296298e-05,
+      "loss": 0.67,
+      "step": 3020
+    },
+    {
+      "epoch": 0.26853333333333335,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.7902222222222224e-05,
+      "loss": 0.5465,
+      "step": 3021
+    },
+    {
+      "epoch": 0.2686222222222222,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.790814814814815e-05,
+      "loss": 0.7028,
+      "step": 3022
+    },
+    {
+      "epoch": 0.2687111111111111,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.7914074074074075e-05,
+      "loss": 0.6254,
+      "step": 3023
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.792e-05,
+      "loss": 0.5537,
+      "step": 3024
+    },
+    {
+      "epoch": 0.2688888888888889,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.7925925925925927e-05,
+      "loss": 0.7592,
+      "step": 3025
+    },
+    {
+      "epoch": 0.26897777777777776,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7931851851851853e-05,
+      "loss": 0.6156,
+      "step": 3026
+    },
+    {
+      "epoch": 0.2690666666666667,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.793777777777778e-05,
+      "loss": 0.6185,
+      "step": 3027
+    },
+    {
+      "epoch": 0.26915555555555554,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.7943703703703704e-05,
+      "loss": 0.6942,
+      "step": 3028
+    },
+    {
+      "epoch": 0.26924444444444445,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.794962962962963e-05,
+      "loss": 0.8446,
+      "step": 3029
+    },
+    {
+      "epoch": 0.2693333333333333,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.7955555555555556e-05,
+      "loss": 0.6522,
+      "step": 3030
+    },
+    {
+      "epoch": 0.26942222222222223,
+      "grad_norm": 1.75,
+      "learning_rate": 1.796148148148148e-05,
+      "loss": 0.7319,
+      "step": 3031
+    },
+    {
+      "epoch": 0.2695111111111111,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.796740740740741e-05,
+      "loss": 0.6536,
+      "step": 3032
+    },
+    {
+      "epoch": 0.2696,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7973333333333333e-05,
+      "loss": 0.7176,
+      "step": 3033
+    },
+    {
+      "epoch": 0.26968888888888887,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.7979259259259262e-05,
+      "loss": 0.7056,
+      "step": 3034
+    },
+    {
+      "epoch": 0.2697777777777778,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.7985185185185188e-05,
+      "loss": 0.7255,
+      "step": 3035
+    },
+    {
+      "epoch": 0.26986666666666664,
+      "grad_norm": 2.375,
+      "learning_rate": 1.7991111111111114e-05,
+      "loss": 0.6119,
+      "step": 3036
+    },
+    {
+      "epoch": 0.26995555555555556,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.799703703703704e-05,
+      "loss": 0.6707,
+      "step": 3037
+    },
+    {
+      "epoch": 0.2700444444444444,
+      "grad_norm": 1.375,
+      "learning_rate": 1.8002962962962965e-05,
+      "loss": 0.5829,
+      "step": 3038
+    },
+    {
+      "epoch": 0.27013333333333334,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.800888888888889e-05,
+      "loss": 0.5972,
+      "step": 3039
+    },
+    {
+      "epoch": 0.2702222222222222,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.8014814814814817e-05,
+      "loss": 0.6695,
+      "step": 3040
+    },
+    {
+      "epoch": 0.2703111111111111,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.8020740740740742e-05,
+      "loss": 0.7238,
+      "step": 3041
+    },
+    {
+      "epoch": 0.2704,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8026666666666668e-05,
+      "loss": 0.711,
+      "step": 3042
+    },
+    {
+      "epoch": 0.2704888888888889,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.8032592592592594e-05,
+      "loss": 0.8693,
+      "step": 3043
+    },
+    {
+      "epoch": 0.27057777777777775,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.803851851851852e-05,
+      "loss": 0.8407,
+      "step": 3044
+    },
+    {
+      "epoch": 0.27066666666666667,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.8044444444444445e-05,
+      "loss": 0.5931,
+      "step": 3045
+    },
+    {
+      "epoch": 0.2707555555555556,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.805037037037037e-05,
+      "loss": 0.8298,
+      "step": 3046
+    },
+    {
+      "epoch": 0.27084444444444444,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.80562962962963e-05,
+      "loss": 0.8303,
+      "step": 3047
+    },
+    {
+      "epoch": 0.27093333333333336,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8062222222222222e-05,
+      "loss": 0.9424,
+      "step": 3048
+    },
+    {
+      "epoch": 0.2710222222222222,
+      "grad_norm": 2.375,
+      "learning_rate": 1.806814814814815e-05,
+      "loss": 0.7352,
+      "step": 3049
+    },
+    {
+      "epoch": 0.27111111111111114,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8074074074074074e-05,
+      "loss": 0.6515,
+      "step": 3050
+    },
+    {
+      "epoch": 0.2712,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.8080000000000003e-05,
+      "loss": 0.497,
+      "step": 3051
+    },
+    {
+      "epoch": 0.2712888888888889,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8085925925925925e-05,
+      "loss": 0.5991,
+      "step": 3052
+    },
+    {
+      "epoch": 0.2713777777777778,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8091851851851855e-05,
+      "loss": 0.6303,
+      "step": 3053
+    },
+    {
+      "epoch": 0.2714666666666667,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.8097777777777777e-05,
+      "loss": 0.5988,
+      "step": 3054
+    },
+    {
+      "epoch": 0.27155555555555555,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.8103703703703706e-05,
+      "loss": 0.8591,
+      "step": 3055
+    },
+    {
+      "epoch": 0.27164444444444447,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.8109629629629632e-05,
+      "loss": 0.7753,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2717333333333333,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.8115555555555558e-05,
+      "loss": 0.6117,
+      "step": 3057
+    },
+    {
+      "epoch": 0.27182222222222224,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.8121481481481483e-05,
+      "loss": 0.7238,
+      "step": 3058
+    },
+    {
+      "epoch": 0.2719111111111111,
+      "grad_norm": 2.625,
+      "learning_rate": 1.812740740740741e-05,
+      "loss": 0.742,
+      "step": 3059
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.8133333333333335e-05,
+      "loss": 0.6988,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2720888888888889,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.813925925925926e-05,
+      "loss": 0.6187,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2721777777777778,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.8145185185185186e-05,
+      "loss": 0.694,
+      "step": 3062
+    },
+    {
+      "epoch": 0.27226666666666666,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.8151111111111112e-05,
+      "loss": 0.7307,
+      "step": 3063
+    },
+    {
+      "epoch": 0.2723555555555556,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.8157037037037038e-05,
+      "loss": 0.6464,
+      "step": 3064
+    },
+    {
+      "epoch": 0.27244444444444443,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.8162962962962963e-05,
+      "loss": 0.8231,
+      "step": 3065
+    },
+    {
+      "epoch": 0.27253333333333335,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.8168888888888893e-05,
+      "loss": 0.6604,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2726222222222222,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8174814814814815e-05,
+      "loss": 0.6991,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2727111111111111,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.8180740740740744e-05,
+      "loss": 0.4793,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2728,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.8186666666666666e-05,
+      "loss": 0.6786,
+      "step": 3069
+    },
+    {
+      "epoch": 0.2728888888888889,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.8192592592592596e-05,
+      "loss": 0.6569,
+      "step": 3070
+    },
+    {
+      "epoch": 0.27297777777777776,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.819851851851852e-05,
+      "loss": 0.7665,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2730666666666667,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.8204444444444447e-05,
+      "loss": 0.6005,
+      "step": 3072
+    },
+    {
+      "epoch": 0.27315555555555554,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.8210370370370373e-05,
+      "loss": 0.6163,
+      "step": 3073
+    },
+    {
+      "epoch": 0.27324444444444446,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.82162962962963e-05,
+      "loss": 0.7327,
+      "step": 3074
+    },
+    {
+      "epoch": 0.2733333333333333,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.8222222222222224e-05,
+      "loss": 0.6473,
+      "step": 3075
+    },
+    {
+      "epoch": 0.27342222222222223,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.822814814814815e-05,
+      "loss": 0.6774,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2735111111111111,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.8234074074074076e-05,
+      "loss": 0.6308,
+      "step": 3077
+    },
+    {
+      "epoch": 0.2736,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.824e-05,
+      "loss": 0.7034,
+      "step": 3078
+    },
+    {
+      "epoch": 0.27368888888888887,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8245925925925927e-05,
+      "loss": 0.7197,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2737777777777778,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.8251851851851853e-05,
+      "loss": 0.7523,
+      "step": 3080
+    },
+    {
+      "epoch": 0.27386666666666665,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.825777777777778e-05,
+      "loss": 0.7486,
+      "step": 3081
+    },
+    {
+      "epoch": 0.27395555555555556,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.8263703703703705e-05,
+      "loss": 0.8034,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2740444444444444,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.826962962962963e-05,
+      "loss": 0.6282,
+      "step": 3083
+    },
+    {
+      "epoch": 0.27413333333333334,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.8275555555555556e-05,
+      "loss": 0.816,
+      "step": 3084
+    },
+    {
+      "epoch": 0.2742222222222222,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.8281481481481482e-05,
+      "loss": 0.6158,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2743111111111111,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.828740740740741e-05,
+      "loss": 0.6131,
+      "step": 3086
+    },
+    {
+      "epoch": 0.2744,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.8293333333333333e-05,
+      "loss": 0.7174,
+      "step": 3087
+    },
+    {
+      "epoch": 0.2744888888888889,
+      "grad_norm": 2.25,
+      "learning_rate": 1.8299259259259262e-05,
+      "loss": 0.6141,
+      "step": 3088
+    },
+    {
+      "epoch": 0.27457777777777775,
+      "grad_norm": 2.375,
+      "learning_rate": 1.8305185185185185e-05,
+      "loss": 0.7305,
+      "step": 3089
+    },
+    {
+      "epoch": 0.27466666666666667,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.8311111111111114e-05,
+      "loss": 0.6769,
+      "step": 3090
+    },
+    {
+      "epoch": 0.27475555555555553,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.8317037037037036e-05,
+      "loss": 0.7104,
+      "step": 3091
+    },
+    {
+      "epoch": 0.27484444444444445,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.8322962962962965e-05,
+      "loss": 0.8185,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2749333333333333,
+      "grad_norm": 2.5,
+      "learning_rate": 1.832888888888889e-05,
+      "loss": 0.7508,
+      "step": 3093
+    },
+    {
+      "epoch": 0.2750222222222222,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.8334814814814817e-05,
+      "loss": 0.6758,
+      "step": 3094
+    },
+    {
+      "epoch": 0.2751111111111111,
+      "grad_norm": 1.875,
+      "learning_rate": 1.8340740740740743e-05,
+      "loss": 0.5467,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.834666666666667e-05,
+      "loss": 0.6707,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2752888888888889,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.8352592592592594e-05,
+      "loss": 0.7628,
+      "step": 3097
+    },
+    {
+      "epoch": 0.2753777777777778,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.835851851851852e-05,
+      "loss": 0.7908,
+      "step": 3098
+    },
+    {
+      "epoch": 0.2754666666666667,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.8364444444444446e-05,
+      "loss": 0.7099,
+      "step": 3099
+    },
+    {
+      "epoch": 0.27555555555555555,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.837037037037037e-05,
+      "loss": 0.5461,
+      "step": 3100
+    },
+    {
+      "epoch": 0.27564444444444447,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.83762962962963e-05,
+      "loss": 0.5895,
+      "step": 3101
+    },
+    {
+      "epoch": 0.27573333333333333,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8382222222222223e-05,
+      "loss": 0.6352,
+      "step": 3102
+    },
+    {
+      "epoch": 0.27582222222222225,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.8388148148148152e-05,
+      "loss": 0.728,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2759111111111111,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.8394074074074074e-05,
+      "loss": 0.742,
+      "step": 3104
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 0.5166,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2760888888888889,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.8405925925925926e-05,
+      "loss": 0.7753,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2761777777777778,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.8411851851851855e-05,
+      "loss": 0.7316,
+      "step": 3107
+    },
+    {
+      "epoch": 0.27626666666666666,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.8417777777777777e-05,
+      "loss": 0.7333,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2763555555555556,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.8423703703703706e-05,
+      "loss": 0.6108,
+      "step": 3109
+    },
+    {
+      "epoch": 0.27644444444444444,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.8429629629629632e-05,
+      "loss": 0.8269,
+      "step": 3110
+    },
+    {
+      "epoch": 0.27653333333333335,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8435555555555558e-05,
+      "loss": 0.598,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2766222222222222,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8441481481481484e-05,
+      "loss": 0.815,
+      "step": 3112
+    },
+    {
+      "epoch": 0.27671111111111113,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.844740740740741e-05,
+      "loss": 0.7398,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2768,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8453333333333335e-05,
+      "loss": 0.6598,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2768888888888889,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.845925925925926e-05,
+      "loss": 0.8013,
+      "step": 3115
+    },
+    {
+      "epoch": 0.27697777777777777,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8465185185185187e-05,
+      "loss": 0.8256,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2770666666666667,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.8471111111111112e-05,
+      "loss": 0.7116,
+      "step": 3117
+    },
+    {
+      "epoch": 0.27715555555555554,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.8477037037037038e-05,
+      "loss": 0.7493,
+      "step": 3118
+    },
+    {
+      "epoch": 0.27724444444444446,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.8482962962962964e-05,
+      "loss": 0.5951,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2773333333333333,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.848888888888889e-05,
+      "loss": 0.7392,
+      "step": 3120
+    },
+    {
+      "epoch": 0.27742222222222224,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8494814814814815e-05,
+      "loss": 0.7294,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2775111111111111,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.850074074074074e-05,
+      "loss": 0.5708,
+      "step": 3122
+    },
+    {
+      "epoch": 0.2776,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.8506666666666667e-05,
+      "loss": 0.6999,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2776888888888889,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8512592592592593e-05,
+      "loss": 0.5472,
+      "step": 3124
+    },
+    {
+      "epoch": 0.2777777777777778,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.851851851851852e-05,
+      "loss": 0.7264,
+      "step": 3125
+    },
+    {
+      "epoch": 0.27786666666666665,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.8524444444444444e-05,
+      "loss": 0.8062,
+      "step": 3126
+    },
+    {
+      "epoch": 0.27795555555555557,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.8530370370370373e-05,
+      "loss": 0.729,
+      "step": 3127
+    },
+    {
+      "epoch": 0.2780444444444444,
+      "grad_norm": 2.125,
+      "learning_rate": 1.85362962962963e-05,
+      "loss": 0.7956,
+      "step": 3128
+    },
+    {
+      "epoch": 0.27813333333333334,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.8542222222222225e-05,
+      "loss": 0.6679,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2782222222222222,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.854814814814815e-05,
+      "loss": 0.7166,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2783111111111111,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.8554074074074076e-05,
+      "loss": 0.739,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.8560000000000002e-05,
+      "loss": 0.5257,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2784888888888889,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.8565925925925928e-05,
+      "loss": 0.5573,
+      "step": 3133
+    },
+    {
+      "epoch": 0.27857777777777776,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.8571851851851853e-05,
+      "loss": 0.6987,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2786666666666667,
+      "grad_norm": 2.125,
+      "learning_rate": 1.857777777777778e-05,
+      "loss": 0.8056,
+      "step": 3135
+    },
+    {
+      "epoch": 0.27875555555555553,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.8583703703703705e-05,
+      "loss": 0.6655,
+      "step": 3136
+    },
+    {
+      "epoch": 0.27884444444444445,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.858962962962963e-05,
+      "loss": 0.5184,
+      "step": 3137
+    },
+    {
+      "epoch": 0.2789333333333333,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8595555555555556e-05,
+      "loss": 0.5775,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2790222222222222,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.8601481481481482e-05,
+      "loss": 0.732,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2791111111111111,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.860740740740741e-05,
+      "loss": 0.6804,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2792,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.8613333333333334e-05,
+      "loss": 0.6225,
+      "step": 3141
+    },
+    {
+      "epoch": 0.27928888888888886,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8619259259259263e-05,
+      "loss": 0.7236,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2793777777777778,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.8625185185185185e-05,
+      "loss": 0.7807,
+      "step": 3143
+    },
+    {
+      "epoch": 0.27946666666666664,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.8631111111111114e-05,
+      "loss": 0.7736,
+      "step": 3144
+    },
+    {
+      "epoch": 0.27955555555555556,
+      "grad_norm": 2.25,
+      "learning_rate": 1.8637037037037037e-05,
+      "loss": 0.693,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2796444444444444,
+      "grad_norm": 2.75,
+      "learning_rate": 1.8642962962962966e-05,
+      "loss": 0.7647,
+      "step": 3146
+    },
+    {
+      "epoch": 0.27973333333333333,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.8648888888888888e-05,
+      "loss": 0.6479,
+      "step": 3147
+    },
+    {
+      "epoch": 0.27982222222222225,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.8654814814814817e-05,
+      "loss": 0.8215,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2799111111111111,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.8660740740740743e-05,
+      "loss": 0.8298,
+      "step": 3149
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 0.6483,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2800888888888889,
+      "grad_norm": 2.625,
+      "learning_rate": 1.8672592592592594e-05,
+      "loss": 0.762,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2801777777777778,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.867851851851852e-05,
+      "loss": 0.627,
+      "step": 3152
+    },
+    {
+      "epoch": 0.28026666666666666,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.8684444444444446e-05,
+      "loss": 0.7445,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2803555555555556,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.869037037037037e-05,
+      "loss": 0.9857,
+      "step": 3154
+    },
+    {
+      "epoch": 0.28044444444444444,
+      "grad_norm": 2.625,
+      "learning_rate": 1.8696296296296297e-05,
+      "loss": 0.6617,
+      "step": 3155
+    },
+    {
+      "epoch": 0.28053333333333336,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.8702222222222223e-05,
+      "loss": 0.5564,
+      "step": 3156
+    },
+    {
+      "epoch": 0.2806222222222222,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.870814814814815e-05,
+      "loss": 0.5001,
+      "step": 3157
+    },
+    {
+      "epoch": 0.28071111111111113,
+      "grad_norm": 2.25,
+      "learning_rate": 1.8714074074074075e-05,
+      "loss": 0.7676,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2808,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.8720000000000004e-05,
+      "loss": 0.7689,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2808888888888889,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.8725925925925926e-05,
+      "loss": 0.698,
+      "step": 3160
+    },
+    {
+      "epoch": 0.28097777777777777,
+      "grad_norm": 1.875,
+      "learning_rate": 1.8731851851851855e-05,
+      "loss": 0.7084,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2810666666666667,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.8737777777777778e-05,
+      "loss": 0.7708,
+      "step": 3162
+    },
+    {
+      "epoch": 0.28115555555555555,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.8743703703703707e-05,
+      "loss": 0.6847,
+      "step": 3163
+    },
+    {
+      "epoch": 0.28124444444444446,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.8749629629629632e-05,
+      "loss": 0.6563,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2813333333333333,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.8755555555555558e-05,
+      "loss": 0.6148,
+      "step": 3165
+    },
+    {
+      "epoch": 0.28142222222222224,
+      "grad_norm": 2.875,
+      "learning_rate": 1.8761481481481484e-05,
+      "loss": 0.6607,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2815111111111111,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.876740740740741e-05,
+      "loss": 0.6557,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.8773333333333335e-05,
+      "loss": 0.7711,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2816888888888889,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.877925925925926e-05,
+      "loss": 0.8321,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2817777777777778,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8785185185185187e-05,
+      "loss": 0.7531,
+      "step": 3170
+    },
+    {
+      "epoch": 0.28186666666666665,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.8791111111111113e-05,
+      "loss": 0.7405,
+      "step": 3171
+    },
+    {
+      "epoch": 0.28195555555555557,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.879703703703704e-05,
+      "loss": 0.6573,
+      "step": 3172
+    },
+    {
+      "epoch": 0.28204444444444443,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.8802962962962964e-05,
+      "loss": 0.7348,
+      "step": 3173
+    },
+    {
+      "epoch": 0.28213333333333335,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.880888888888889e-05,
+      "loss": 0.9339,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2822222222222222,
+      "grad_norm": 2.25,
+      "learning_rate": 1.8814814814814816e-05,
+      "loss": 0.7669,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2823111111111111,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.882074074074074e-05,
+      "loss": 0.6933,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2824,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.8826666666666667e-05,
+      "loss": 0.8457,
+      "step": 3177
+    },
+    {
+      "epoch": 0.2824888888888889,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8832592592592593e-05,
+      "loss": 0.6103,
+      "step": 3178
+    },
+    {
+      "epoch": 0.28257777777777776,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.8838518518518522e-05,
+      "loss": 0.5732,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2826666666666667,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8844444444444444e-05,
+      "loss": 0.7281,
+      "step": 3180
+    },
+    {
+      "epoch": 0.28275555555555554,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8850370370370373e-05,
+      "loss": 0.7394,
+      "step": 3181
+    },
+    {
+      "epoch": 0.28284444444444445,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.8856296296296296e-05,
+      "loss": 0.7442,
+      "step": 3182
+    },
+    {
+      "epoch": 0.2829333333333333,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.8862222222222225e-05,
+      "loss": 0.6822,
+      "step": 3183
+    },
+    {
+      "epoch": 0.28302222222222223,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.8868148148148147e-05,
+      "loss": 0.6836,
+      "step": 3184
+    },
+    {
+      "epoch": 0.2831111111111111,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.8874074074074076e-05,
+      "loss": 0.4885,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2832,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.8880000000000002e-05,
+      "loss": 0.6757,
+      "step": 3186
+    },
+    {
+      "epoch": 0.28328888888888887,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.8885925925925928e-05,
+      "loss": 0.7012,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2833777777777778,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.8891851851851854e-05,
+      "loss": 0.6462,
+      "step": 3188
+    },
+    {
+      "epoch": 0.28346666666666664,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.889777777777778e-05,
+      "loss": 0.7123,
+      "step": 3189
+    },
+    {
+      "epoch": 0.28355555555555556,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.8903703703703705e-05,
+      "loss": 0.6663,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2836444444444444,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.890962962962963e-05,
+      "loss": 0.6998,
+      "step": 3191
+    },
+    {
+      "epoch": 0.28373333333333334,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8915555555555557e-05,
+      "loss": 0.6907,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2838222222222222,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.8921481481481482e-05,
+      "loss": 0.613,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2839111111111111,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.892740740740741e-05,
+      "loss": 0.7343,
+      "step": 3194
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8933333333333334e-05,
+      "loss": 0.6943,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2840888888888889,
+      "grad_norm": 2.125,
+      "learning_rate": 1.8939259259259263e-05,
+      "loss": 0.7992,
+      "step": 3196
+    },
+    {
+      "epoch": 0.28417777777777775,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8945185185185185e-05,
+      "loss": 0.5354,
+      "step": 3197
+    },
+    {
+      "epoch": 0.28426666666666667,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.8951111111111115e-05,
+      "loss": 0.6216,
+      "step": 3198
+    },
+    {
+      "epoch": 0.28435555555555553,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8957037037037037e-05,
+      "loss": 0.6471,
+      "step": 3199
+    },
+    {
+      "epoch": 0.28444444444444444,
+      "grad_norm": 2.125,
+      "learning_rate": 1.8962962962962966e-05,
+      "loss": 0.7135,
+      "step": 3200
+    },
+    {
+      "epoch": 0.28453333333333336,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.896888888888889e-05,
+      "loss": 0.6803,
+      "step": 3201
+    },
+    {
+      "epoch": 0.2846222222222222,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.8974814814814817e-05,
+      "loss": 0.6041,
+      "step": 3202
+    },
+    {
+      "epoch": 0.28471111111111114,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.8980740740740743e-05,
+      "loss": 0.6816,
+      "step": 3203
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.898666666666667e-05,
+      "loss": 0.5941,
+      "step": 3204
+    },
+    {
+      "epoch": 0.2848888888888889,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8992592592592595e-05,
+      "loss": 0.5944,
+      "step": 3205
+    },
+    {
+      "epoch": 0.2849777777777778,
+      "grad_norm": 2.375,
+      "learning_rate": 1.899851851851852e-05,
+      "loss": 0.8902,
+      "step": 3206
+    },
+    {
+      "epoch": 0.2850666666666667,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9004444444444446e-05,
+      "loss": 0.8259,
+      "step": 3207
+    },
+    {
+      "epoch": 0.28515555555555555,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.9010370370370372e-05,
+      "loss": 1.0409,
+      "step": 3208
+    },
+    {
+      "epoch": 0.28524444444444447,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9016296296296298e-05,
+      "loss": 0.5323,
+      "step": 3209
+    },
+    {
+      "epoch": 0.2853333333333333,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.9022222222222223e-05,
+      "loss": 0.7015,
+      "step": 3210
+    },
+    {
+      "epoch": 0.28542222222222224,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.902814814814815e-05,
+      "loss": 0.6142,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2855111111111111,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9034074074074075e-05,
+      "loss": 0.6628,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2856,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.904e-05,
+      "loss": 0.7388,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2856888888888889,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9045925925925926e-05,
+      "loss": 0.6903,
+      "step": 3214
+    },
+    {
+      "epoch": 0.2857777777777778,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9051851851851852e-05,
+      "loss": 0.6548,
+      "step": 3215
+    },
+    {
+      "epoch": 0.28586666666666666,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9057777777777778e-05,
+      "loss": 0.7576,
+      "step": 3216
+    },
+    {
+      "epoch": 0.2859555555555556,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.9063703703703704e-05,
+      "loss": 0.7354,
+      "step": 3217
+    },
+    {
+      "epoch": 0.28604444444444443,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9069629629629633e-05,
+      "loss": 0.8553,
+      "step": 3218
+    },
+    {
+      "epoch": 0.28613333333333335,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9075555555555555e-05,
+      "loss": 0.6212,
+      "step": 3219
+    },
+    {
+      "epoch": 0.2862222222222222,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9081481481481484e-05,
+      "loss": 0.755,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2863111111111111,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.908740740740741e-05,
+      "loss": 0.8708,
+      "step": 3221
+    },
+    {
+      "epoch": 0.2864,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9093333333333336e-05,
+      "loss": 0.6944,
+      "step": 3222
+    },
+    {
+      "epoch": 0.2864888888888889,
+      "grad_norm": 2.25,
+      "learning_rate": 1.909925925925926e-05,
+      "loss": 0.7449,
+      "step": 3223
+    },
+    {
+      "epoch": 0.28657777777777776,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9105185185185187e-05,
+      "loss": 0.5336,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2866666666666667,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.9111111111111113e-05,
+      "loss": 0.8196,
+      "step": 3225
+    },
+    {
+      "epoch": 0.28675555555555554,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.911703703703704e-05,
+      "loss": 0.5211,
+      "step": 3226
+    },
+    {
+      "epoch": 0.28684444444444446,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9122962962962964e-05,
+      "loss": 0.8975,
+      "step": 3227
+    },
+    {
+      "epoch": 0.2869333333333333,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.912888888888889e-05,
+      "loss": 0.6937,
+      "step": 3228
+    },
+    {
+      "epoch": 0.28702222222222223,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9134814814814816e-05,
+      "loss": 0.6108,
+      "step": 3229
+    },
+    {
+      "epoch": 0.2871111111111111,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9140740740740742e-05,
+      "loss": 0.5358,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2872,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.9146666666666667e-05,
+      "loss": 0.9381,
+      "step": 3231
+    },
+    {
+      "epoch": 0.28728888888888887,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.9152592592592593e-05,
+      "loss": 0.6654,
+      "step": 3232
+    },
+    {
+      "epoch": 0.2873777777777778,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9158518518518522e-05,
+      "loss": 0.6295,
+      "step": 3233
+    },
+    {
+      "epoch": 0.28746666666666665,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9164444444444445e-05,
+      "loss": 0.6988,
+      "step": 3234
+    },
+    {
+      "epoch": 0.28755555555555556,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9170370370370374e-05,
+      "loss": 0.6852,
+      "step": 3235
+    },
+    {
+      "epoch": 0.2876444444444444,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9176296296296296e-05,
+      "loss": 0.8222,
+      "step": 3236
+    },
+    {
+      "epoch": 0.28773333333333334,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9182222222222225e-05,
+      "loss": 0.72,
+      "step": 3237
+    },
+    {
+      "epoch": 0.2878222222222222,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9188148148148148e-05,
+      "loss": 0.5551,
+      "step": 3238
+    },
+    {
+      "epoch": 0.2879111111111111,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9194074074074077e-05,
+      "loss": 0.6283,
+      "step": 3239
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.7266,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2880888888888889,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9205925925925928e-05,
+      "loss": 0.6778,
+      "step": 3241
+    },
+    {
+      "epoch": 0.28817777777777775,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9211851851851854e-05,
+      "loss": 0.6837,
+      "step": 3242
+    },
+    {
+      "epoch": 0.28826666666666667,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.921777777777778e-05,
+      "loss": 0.6791,
+      "step": 3243
+    },
+    {
+      "epoch": 0.28835555555555553,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9223703703703705e-05,
+      "loss": 0.5851,
+      "step": 3244
+    },
+    {
+      "epoch": 0.28844444444444445,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.922962962962963e-05,
+      "loss": 0.7037,
+      "step": 3245
+    },
+    {
+      "epoch": 0.2885333333333333,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9235555555555557e-05,
+      "loss": 0.5545,
+      "step": 3246
+    },
+    {
+      "epoch": 0.2886222222222222,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.9241481481481483e-05,
+      "loss": 0.5881,
+      "step": 3247
+    },
+    {
+      "epoch": 0.2887111111111111,
+      "grad_norm": 2.75,
+      "learning_rate": 1.924740740740741e-05,
+      "loss": 0.8804,
+      "step": 3248
+    },
+    {
+      "epoch": 0.2888,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9253333333333334e-05,
+      "loss": 0.6075,
+      "step": 3249
+    },
+    {
+      "epoch": 0.28888888888888886,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.7452,
+      "step": 3250
+    },
+    {
+      "epoch": 0.2889777777777778,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9265185185185186e-05,
+      "loss": 0.6187,
+      "step": 3251
+    },
+    {
+      "epoch": 0.2890666666666667,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9271111111111115e-05,
+      "loss": 0.5507,
+      "step": 3252
+    },
+    {
+      "epoch": 0.28915555555555555,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9277037037037037e-05,
+      "loss": 0.8396,
+      "step": 3253
+    },
+    {
+      "epoch": 0.28924444444444447,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.9282962962962966e-05,
+      "loss": 0.7152,
+      "step": 3254
+    },
+    {
+      "epoch": 0.28933333333333333,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.928888888888889e-05,
+      "loss": 0.7362,
+      "step": 3255
+    },
+    {
+      "epoch": 0.28942222222222225,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9294814814814818e-05,
+      "loss": 0.8603,
+      "step": 3256
+    },
+    {
+      "epoch": 0.2895111111111111,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9300740740740744e-05,
+      "loss": 0.6356,
+      "step": 3257
+    },
+    {
+      "epoch": 0.2896,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.930666666666667e-05,
+      "loss": 0.6976,
+      "step": 3258
+    },
+    {
+      "epoch": 0.2896888888888889,
+      "grad_norm": 2.5,
+      "learning_rate": 1.9312592592592595e-05,
+      "loss": 0.6588,
+      "step": 3259
+    },
+    {
+      "epoch": 0.2897777777777778,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.931851851851852e-05,
+      "loss": 0.753,
+      "step": 3260
+    },
+    {
+      "epoch": 0.28986666666666666,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9324444444444447e-05,
+      "loss": 0.5361,
+      "step": 3261
+    },
+    {
+      "epoch": 0.2899555555555556,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9330370370370372e-05,
+      "loss": 0.7496,
+      "step": 3262
+    },
+    {
+      "epoch": 0.29004444444444444,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9336296296296298e-05,
+      "loss": 0.5546,
+      "step": 3263
+    },
+    {
+      "epoch": 0.29013333333333335,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9342222222222224e-05,
+      "loss": 0.6949,
+      "step": 3264
+    },
+    {
+      "epoch": 0.2902222222222222,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.934814814814815e-05,
+      "loss": 0.6081,
+      "step": 3265
+    },
+    {
+      "epoch": 0.29031111111111113,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.9354074074074075e-05,
+      "loss": 0.6973,
+      "step": 3266
+    },
+    {
+      "epoch": 0.2904,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.936e-05,
+      "loss": 0.7255,
+      "step": 3267
+    },
+    {
+      "epoch": 0.2904888888888889,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9365925925925927e-05,
+      "loss": 0.8128,
+      "step": 3268
+    },
+    {
+      "epoch": 0.29057777777777777,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.9371851851851852e-05,
+      "loss": 0.6751,
+      "step": 3269
+    },
+    {
+      "epoch": 0.2906666666666667,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9377777777777778e-05,
+      "loss": 0.7116,
+      "step": 3270
+    },
+    {
+      "epoch": 0.29075555555555554,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9383703703703704e-05,
+      "loss": 0.8013,
+      "step": 3271
+    },
+    {
+      "epoch": 0.29084444444444446,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.9389629629629633e-05,
+      "loss": 0.7036,
+      "step": 3272
+    },
+    {
+      "epoch": 0.2909333333333333,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9395555555555555e-05,
+      "loss": 0.8818,
+      "step": 3273
+    },
+    {
+      "epoch": 0.29102222222222224,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9401481481481485e-05,
+      "loss": 0.679,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2911111111111111,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9407407407407407e-05,
+      "loss": 0.6136,
+      "step": 3275
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9413333333333336e-05,
+      "loss": 0.7601,
+      "step": 3276
+    },
+    {
+      "epoch": 0.2912888888888889,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.941925925925926e-05,
+      "loss": 0.786,
+      "step": 3277
+    },
+    {
+      "epoch": 0.2913777777777778,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9425185185185188e-05,
+      "loss": 0.5808,
+      "step": 3278
+    },
+    {
+      "epoch": 0.29146666666666665,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9431111111111113e-05,
+      "loss": 0.6989,
+      "step": 3279
+    },
+    {
+      "epoch": 0.29155555555555557,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.943703703703704e-05,
+      "loss": 0.8006,
+      "step": 3280
+    },
+    {
+      "epoch": 0.29164444444444443,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9442962962962965e-05,
+      "loss": 0.7015,
+      "step": 3281
+    },
+    {
+      "epoch": 0.29173333333333334,
+      "grad_norm": 2.625,
+      "learning_rate": 1.944888888888889e-05,
+      "loss": 0.8851,
+      "step": 3282
+    },
+    {
+      "epoch": 0.2918222222222222,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9454814814814816e-05,
+      "loss": 0.6301,
+      "step": 3283
+    },
+    {
+      "epoch": 0.2919111111111111,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.9460740740740742e-05,
+      "loss": 0.5262,
+      "step": 3284
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9466666666666668e-05,
+      "loss": 0.6398,
+      "step": 3285
+    },
+    {
+      "epoch": 0.2920888888888889,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.9472592592592594e-05,
+      "loss": 0.6516,
+      "step": 3286
+    },
+    {
+      "epoch": 0.29217777777777776,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.9478518518518523e-05,
+      "loss": 0.7104,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2922666666666667,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.9484444444444445e-05,
+      "loss": 0.7717,
+      "step": 3288
+    },
+    {
+      "epoch": 0.29235555555555554,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.9490370370370374e-05,
+      "loss": 0.6691,
+      "step": 3289
+    },
+    {
+      "epoch": 0.29244444444444445,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9496296296296296e-05,
+      "loss": 0.7333,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2925333333333333,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.9502222222222226e-05,
+      "loss": 0.651,
+      "step": 3291
+    },
+    {
+      "epoch": 0.29262222222222223,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.9508148148148148e-05,
+      "loss": 0.7891,
+      "step": 3292
+    },
+    {
+      "epoch": 0.2927111111111111,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.9514074074074077e-05,
+      "loss": 0.7998,
+      "step": 3293
+    },
+    {
+      "epoch": 0.2928,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9520000000000003e-05,
+      "loss": 0.7262,
+      "step": 3294
+    },
+    {
+      "epoch": 0.29288888888888887,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.952592592592593e-05,
+      "loss": 0.7504,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2929777777777778,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9531851851851854e-05,
+      "loss": 0.4961,
+      "step": 3296
+    },
+    {
+      "epoch": 0.29306666666666664,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.953777777777778e-05,
+      "loss": 0.7355,
+      "step": 3297
+    },
+    {
+      "epoch": 0.29315555555555556,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.9543703703703706e-05,
+      "loss": 0.801,
+      "step": 3298
+    },
+    {
+      "epoch": 0.2932444444444444,
+      "grad_norm": 2.125,
+      "learning_rate": 1.954962962962963e-05,
+      "loss": 0.7129,
+      "step": 3299
+    },
+    {
+      "epoch": 0.29333333333333333,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9555555555555557e-05,
+      "loss": 0.7084,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2934222222222222,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9561481481481483e-05,
+      "loss": 0.5893,
+      "step": 3301
+    },
+    {
+      "epoch": 0.2935111111111111,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.956740740740741e-05,
+      "loss": 0.6831,
+      "step": 3302
+    },
+    {
+      "epoch": 0.2936,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.9573333333333335e-05,
+      "loss": 0.6463,
+      "step": 3303
+    },
+    {
+      "epoch": 0.2936888888888889,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.957925925925926e-05,
+      "loss": 0.7659,
+      "step": 3304
+    },
+    {
+      "epoch": 0.2937777777777778,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9585185185185186e-05,
+      "loss": 0.7913,
+      "step": 3305
+    },
+    {
+      "epoch": 0.29386666666666666,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9591111111111112e-05,
+      "loss": 0.7385,
+      "step": 3306
+    },
+    {
+      "epoch": 0.2939555555555556,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.9597037037037038e-05,
+      "loss": 0.8176,
+      "step": 3307
+    },
+    {
+      "epoch": 0.29404444444444444,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9602962962962963e-05,
+      "loss": 0.769,
+      "step": 3308
+    },
+    {
+      "epoch": 0.29413333333333336,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.960888888888889e-05,
+      "loss": 0.6787,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2942222222222222,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.9614814814814815e-05,
+      "loss": 0.758,
+      "step": 3310
+    },
+    {
+      "epoch": 0.29431111111111113,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9620740740740744e-05,
+      "loss": 0.6994,
+      "step": 3311
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9626666666666666e-05,
+      "loss": 0.7609,
+      "step": 3312
+    },
+    {
+      "epoch": 0.2944888888888889,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9632592592592595e-05,
+      "loss": 0.6695,
+      "step": 3313
+    },
+    {
+      "epoch": 0.29457777777777777,
+      "grad_norm": 1.75,
+      "learning_rate": 1.963851851851852e-05,
+      "loss": 0.863,
+      "step": 3314
+    },
+    {
+      "epoch": 0.2946666666666667,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9644444444444447e-05,
+      "loss": 0.7771,
+      "step": 3315
+    },
+    {
+      "epoch": 0.29475555555555555,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9650370370370373e-05,
+      "loss": 0.723,
+      "step": 3316
+    },
+    {
+      "epoch": 0.29484444444444446,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.96562962962963e-05,
+      "loss": 0.6389,
+      "step": 3317
+    },
+    {
+      "epoch": 0.2949333333333333,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9662222222222224e-05,
+      "loss": 0.7678,
+      "step": 3318
+    },
+    {
+      "epoch": 0.29502222222222224,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.966814814814815e-05,
+      "loss": 0.7763,
+      "step": 3319
+    },
+    {
+      "epoch": 0.2951111111111111,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.9674074074074076e-05,
+      "loss": 0.9481,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2952,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.968e-05,
+      "loss": 0.7171,
+      "step": 3321
+    },
+    {
+      "epoch": 0.2952888888888889,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9685925925925927e-05,
+      "loss": 0.727,
+      "step": 3322
+    },
+    {
+      "epoch": 0.2953777777777778,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9691851851851853e-05,
+      "loss": 0.7058,
+      "step": 3323
+    },
+    {
+      "epoch": 0.29546666666666666,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.969777777777778e-05,
+      "loss": 0.7433,
+      "step": 3324
+    },
+    {
+      "epoch": 0.29555555555555557,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9703703703703704e-05,
+      "loss": 0.8516,
+      "step": 3325
+    },
+    {
+      "epoch": 0.29564444444444443,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.9709629629629633e-05,
+      "loss": 0.5749,
+      "step": 3326
+    },
+    {
+      "epoch": 0.29573333333333335,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.9715555555555556e-05,
+      "loss": 0.7709,
+      "step": 3327
+    },
+    {
+      "epoch": 0.2958222222222222,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9721481481481485e-05,
+      "loss": 0.6996,
+      "step": 3328
+    },
+    {
+      "epoch": 0.2959111111111111,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9727407407407407e-05,
+      "loss": 0.7776,
+      "step": 3329
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9733333333333336e-05,
+      "loss": 0.5761,
+      "step": 3330
+    },
+    {
+      "epoch": 0.2960888888888889,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.973925925925926e-05,
+      "loss": 0.7321,
+      "step": 3331
+    },
+    {
+      "epoch": 0.29617777777777776,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.9745185185185188e-05,
+      "loss": 0.6758,
+      "step": 3332
+    },
+    {
+      "epoch": 0.2962666666666667,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9751111111111114e-05,
+      "loss": 0.594,
+      "step": 3333
+    },
+    {
+      "epoch": 0.29635555555555554,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.975703703703704e-05,
+      "loss": 0.6206,
+      "step": 3334
+    },
+    {
+      "epoch": 0.29644444444444445,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9762962962962965e-05,
+      "loss": 0.6602,
+      "step": 3335
+    },
+    {
+      "epoch": 0.2965333333333333,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.976888888888889e-05,
+      "loss": 0.7383,
+      "step": 3336
+    },
+    {
+      "epoch": 0.29662222222222223,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9774814814814817e-05,
+      "loss": 0.7615,
+      "step": 3337
+    },
+    {
+      "epoch": 0.2967111111111111,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.9780740740740742e-05,
+      "loss": 0.7762,
+      "step": 3338
+    },
+    {
+      "epoch": 0.2968,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9786666666666668e-05,
+      "loss": 0.7191,
+      "step": 3339
+    },
+    {
+      "epoch": 0.29688888888888887,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9792592592592594e-05,
+      "loss": 0.5323,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2969777777777778,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.979851851851852e-05,
+      "loss": 0.671,
+      "step": 3341
+    },
+    {
+      "epoch": 0.29706666666666665,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9804444444444445e-05,
+      "loss": 0.6345,
+      "step": 3342
+    },
+    {
+      "epoch": 0.29715555555555556,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.981037037037037e-05,
+      "loss": 0.8184,
+      "step": 3343
+    },
+    {
+      "epoch": 0.2972444444444444,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9816296296296297e-05,
+      "loss": 0.711,
+      "step": 3344
+    },
+    {
+      "epoch": 0.29733333333333334,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9822222222222226e-05,
+      "loss": 0.8864,
+      "step": 3345
+    },
+    {
+      "epoch": 0.2974222222222222,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9828148148148148e-05,
+      "loss": 0.6949,
+      "step": 3346
+    },
+    {
+      "epoch": 0.2975111111111111,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.9834074074074077e-05,
+      "loss": 0.7508,
+      "step": 3347
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.9840000000000003e-05,
+      "loss": 0.7553,
+      "step": 3348
+    },
+    {
+      "epoch": 0.2976888888888889,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.984592592592593e-05,
+      "loss": 0.7885,
+      "step": 3349
+    },
+    {
+      "epoch": 0.29777777777777775,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9851851851851855e-05,
+      "loss": 0.6951,
+      "step": 3350
+    },
+    {
+      "epoch": 0.29786666666666667,
+      "grad_norm": 2.875,
+      "learning_rate": 1.985777777777778e-05,
+      "loss": 0.6938,
+      "step": 3351
+    },
+    {
+      "epoch": 0.29795555555555553,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.9863703703703706e-05,
+      "loss": 0.6071,
+      "step": 3352
+    },
+    {
+      "epoch": 0.29804444444444445,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.9869629629629632e-05,
+      "loss": 0.7149,
+      "step": 3353
+    },
+    {
+      "epoch": 0.2981333333333333,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9875555555555558e-05,
+      "loss": 0.8511,
+      "step": 3354
+    },
+    {
+      "epoch": 0.2982222222222222,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9881481481481483e-05,
+      "loss": 0.6265,
+      "step": 3355
+    },
+    {
+      "epoch": 0.29831111111111114,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.988740740740741e-05,
+      "loss": 0.6908,
+      "step": 3356
+    },
+    {
+      "epoch": 0.2984,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9893333333333335e-05,
+      "loss": 0.7144,
+      "step": 3357
+    },
+    {
+      "epoch": 0.2984888888888889,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.989925925925926e-05,
+      "loss": 0.6816,
+      "step": 3358
+    },
+    {
+      "epoch": 0.2985777777777778,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.9905185185185186e-05,
+      "loss": 0.6312,
+      "step": 3359
+    },
+    {
+      "epoch": 0.2986666666666667,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.9911111111111112e-05,
+      "loss": 0.8078,
+      "step": 3360
+    },
+    {
+      "epoch": 0.29875555555555555,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9917037037037038e-05,
+      "loss": 0.6207,
+      "step": 3361
+    },
+    {
+      "epoch": 0.29884444444444447,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9922962962962964e-05,
+      "loss": 0.5575,
+      "step": 3362
+    },
+    {
+      "epoch": 0.29893333333333333,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.992888888888889e-05,
+      "loss": 0.6693,
+      "step": 3363
+    },
+    {
+      "epoch": 0.29902222222222224,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.9934814814814815e-05,
+      "loss": 0.7695,
+      "step": 3364
+    },
+    {
+      "epoch": 0.2991111111111111,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9940740740740744e-05,
+      "loss": 0.807,
+      "step": 3365
+    },
+    {
+      "epoch": 0.2992,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.9946666666666667e-05,
+      "loss": 0.7515,
+      "step": 3366
+    },
+    {
+      "epoch": 0.2992888888888889,
+      "grad_norm": 2.375,
+      "learning_rate": 1.9952592592592596e-05,
+      "loss": 0.5025,
+      "step": 3367
+    },
+    {
+      "epoch": 0.2993777777777778,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9958518518518518e-05,
+      "loss": 0.7395,
+      "step": 3368
+    },
+    {
+      "epoch": 0.29946666666666666,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.9964444444444447e-05,
+      "loss": 0.697,
+      "step": 3369
+    },
+    {
+      "epoch": 0.2995555555555556,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.997037037037037e-05,
+      "loss": 0.639,
+      "step": 3370
+    },
+    {
+      "epoch": 0.29964444444444444,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.99762962962963e-05,
+      "loss": 0.6014,
+      "step": 3371
+    },
+    {
+      "epoch": 0.29973333333333335,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9982222222222224e-05,
+      "loss": 0.686,
+      "step": 3372
+    },
+    {
+      "epoch": 0.2998222222222222,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.998814814814815e-05,
+      "loss": 0.7081,
+      "step": 3373
+    },
+    {
+      "epoch": 0.29991111111111113,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9994074074074076e-05,
+      "loss": 0.6776,
+      "step": 3374
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2e-05,
+      "loss": 0.584,
+      "step": 3375
+    },
+    {
+      "epoch": 0.3000888888888889,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9999999946514364e-05,
+      "loss": 0.7292,
+      "step": 3376
+    },
+    {
+      "epoch": 0.30017777777777777,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9999999786057453e-05,
+      "loss": 0.8281,
+      "step": 3377
+    },
+    {
+      "epoch": 0.3002666666666667,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9999999518629267e-05,
+      "loss": 0.5703,
+      "step": 3378
+    },
+    {
+      "epoch": 0.30035555555555554,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9999999144229815e-05,
+      "loss": 0.7201,
+      "step": 3379
+    },
+    {
+      "epoch": 0.30044444444444446,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9999998662859093e-05,
+      "loss": 0.7381,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3005333333333333,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9999998074517113e-05,
+      "loss": 0.8093,
+      "step": 3381
+    },
+    {
+      "epoch": 0.30062222222222224,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.999999737920388e-05,
+      "loss": 0.6332,
+      "step": 3382
+    },
+    {
+      "epoch": 0.3007111111111111,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9999996576919396e-05,
+      "loss": 0.5476,
+      "step": 3383
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9999995667663676e-05,
+      "loss": 0.6684,
+      "step": 3384
+    },
+    {
+      "epoch": 0.3008888888888889,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.9999994651436728e-05,
+      "loss": 0.6515,
+      "step": 3385
+    },
+    {
+      "epoch": 0.3009777777777778,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9999993528238564e-05,
+      "loss": 0.6398,
+      "step": 3386
+    },
+    {
+      "epoch": 0.30106666666666665,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.999999229806919e-05,
+      "loss": 0.6033,
+      "step": 3387
+    },
+    {
+      "epoch": 0.30115555555555557,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9999990960928627e-05,
+      "loss": 0.7495,
+      "step": 3388
+    },
+    {
+      "epoch": 0.3012444444444444,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9999989516816883e-05,
+      "loss": 0.7309,
+      "step": 3389
+    },
+    {
+      "epoch": 0.30133333333333334,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.999998796573398e-05,
+      "loss": 0.6556,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3014222222222222,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9999986307679927e-05,
+      "loss": 0.6003,
+      "step": 3391
+    },
+    {
+      "epoch": 0.3015111111111111,
+      "grad_norm": 2.0,
+      "learning_rate": 1.999998454265475e-05,
+      "loss": 0.6569,
+      "step": 3392
+    },
+    {
+      "epoch": 0.3016,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9999982670658457e-05,
+      "loss": 0.5904,
+      "step": 3393
+    },
+    {
+      "epoch": 0.3016888888888889,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9999980691691083e-05,
+      "loss": 0.6396,
+      "step": 3394
+    },
+    {
+      "epoch": 0.30177777777777776,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.9999978605752632e-05,
+      "loss": 0.8274,
+      "step": 3395
+    },
+    {
+      "epoch": 0.30186666666666667,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.999997641284314e-05,
+      "loss": 0.6933,
+      "step": 3396
+    },
+    {
+      "epoch": 0.30195555555555553,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.9999974112962625e-05,
+      "loss": 0.7337,
+      "step": 3397
+    },
+    {
+      "epoch": 0.30204444444444445,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.999997170611111e-05,
+      "loss": 0.7047,
+      "step": 3398
+    },
+    {
+      "epoch": 0.3021333333333333,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.9999969192288626e-05,
+      "loss": 0.7323,
+      "step": 3399
+    },
+    {
+      "epoch": 0.3022222222222222,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.999996657149519e-05,
+      "loss": 0.5967,
+      "step": 3400
+    },
+    {
+      "epoch": 0.3023111111111111,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9999963843730844e-05,
+      "loss": 0.7203,
+      "step": 3401
+    },
+    {
+      "epoch": 0.3024,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.9999961008995607e-05,
+      "loss": 0.7977,
+      "step": 3402
+    },
+    {
+      "epoch": 0.30248888888888886,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.999995806728951e-05,
+      "loss": 0.7435,
+      "step": 3403
+    },
+    {
+      "epoch": 0.3025777777777778,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9999955018612592e-05,
+      "loss": 0.6411,
+      "step": 3404
+    },
+    {
+      "epoch": 0.30266666666666664,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9999951862964876e-05,
+      "loss": 0.822,
+      "step": 3405
+    },
+    {
+      "epoch": 0.30275555555555556,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.99999486003464e-05,
+      "loss": 0.6955,
+      "step": 3406
+    },
+    {
+      "epoch": 0.30284444444444447,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.9999945230757204e-05,
+      "loss": 0.6816,
+      "step": 3407
+    },
+    {
+      "epoch": 0.30293333333333333,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.9999941754197313e-05,
+      "loss": 0.7273,
+      "step": 3408
+    },
+    {
+      "epoch": 0.30302222222222225,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9999938170666777e-05,
+      "loss": 0.7164,
+      "step": 3409
+    },
+    {
+      "epoch": 0.3031111111111111,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.999993448016562e-05,
+      "loss": 0.6794,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3032,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9999930682693892e-05,
+      "loss": 0.8892,
+      "step": 3411
+    },
+    {
+      "epoch": 0.3032888888888889,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.999992677825163e-05,
+      "loss": 0.6706,
+      "step": 3412
+    },
+    {
+      "epoch": 0.3033777777777778,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9999922766838878e-05,
+      "loss": 0.6333,
+      "step": 3413
+    },
+    {
+      "epoch": 0.30346666666666666,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9999918648455675e-05,
+      "loss": 0.7175,
+      "step": 3414
+    },
+    {
+      "epoch": 0.3035555555555556,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9999914423102068e-05,
+      "loss": 0.6526,
+      "step": 3415
+    },
+    {
+      "epoch": 0.30364444444444444,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9999910090778102e-05,
+      "loss": 0.6619,
+      "step": 3416
+    },
+    {
+      "epoch": 0.30373333333333336,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.9999905651483823e-05,
+      "loss": 0.5868,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3038222222222222,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.999990110521928e-05,
+      "loss": 0.6847,
+      "step": 3418
+    },
+    {
+      "epoch": 0.30391111111111113,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.999989645198452e-05,
+      "loss": 0.5742,
+      "step": 3419
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.999989169177959e-05,
+      "loss": 0.8036,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3040888888888889,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9999886824604545e-05,
+      "loss": 0.7682,
+      "step": 3421
+    },
+    {
+      "epoch": 0.30417777777777777,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.9999881850459435e-05,
+      "loss": 0.5821,
+      "step": 3422
+    },
+    {
+      "epoch": 0.3042666666666667,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9999876769344314e-05,
+      "loss": 0.739,
+      "step": 3423
+    },
+    {
+      "epoch": 0.30435555555555555,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.9999871581259235e-05,
+      "loss": 0.7204,
+      "step": 3424
+    },
+    {
+      "epoch": 0.30444444444444446,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.999986628620426e-05,
+      "loss": 0.749,
+      "step": 3425
+    },
+    {
+      "epoch": 0.3045333333333333,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9999860884179437e-05,
+      "loss": 0.6519,
+      "step": 3426
+    },
+    {
+      "epoch": 0.30462222222222224,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.9999855375184826e-05,
+      "loss": 0.7048,
+      "step": 3427
+    },
+    {
+      "epoch": 0.3047111111111111,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.9999849759220488e-05,
+      "loss": 0.7971,
+      "step": 3428
+    },
+    {
+      "epoch": 0.3048,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9999844036286483e-05,
+      "loss": 0.7347,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3048888888888889,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.9999838206382873e-05,
+      "loss": 0.666,
+      "step": 3430
+    },
+    {
+      "epoch": 0.3049777777777778,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.9999832269509714e-05,
+      "loss": 0.7103,
+      "step": 3431
+    },
+    {
+      "epoch": 0.30506666666666665,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.999982622566708e-05,
+      "loss": 0.8208,
+      "step": 3432
+    },
+    {
+      "epoch": 0.30515555555555557,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9999820074855028e-05,
+      "loss": 0.7368,
+      "step": 3433
+    },
+    {
+      "epoch": 0.30524444444444443,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9999813817073622e-05,
+      "loss": 0.7154,
+      "step": 3434
+    },
+    {
+      "epoch": 0.30533333333333335,
+      "grad_norm": 1.875,
+      "learning_rate": 1.999980745232294e-05,
+      "loss": 0.7523,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3054222222222222,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9999800980603038e-05,
+      "loss": 0.616,
+      "step": 3436
+    },
+    {
+      "epoch": 0.3055111111111111,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.999979440191399e-05,
+      "loss": 0.626,
+      "step": 3437
+    },
+    {
+      "epoch": 0.3056,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9999787716255867e-05,
+      "loss": 0.5838,
+      "step": 3438
+    },
+    {
+      "epoch": 0.3056888888888889,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.999978092362874e-05,
+      "loss": 0.811,
+      "step": 3439
+    },
+    {
+      "epoch": 0.30577777777777776,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.9999774024032683e-05,
+      "loss": 0.6974,
+      "step": 3440
+    },
+    {
+      "epoch": 0.3058666666666667,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9999767017467765e-05,
+      "loss": 0.5519,
+      "step": 3441
+    },
+    {
+      "epoch": 0.30595555555555554,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.999975990393407e-05,
+      "loss": 0.9707,
+      "step": 3442
+    },
+    {
+      "epoch": 0.30604444444444445,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.9999752683431664e-05,
+      "loss": 0.5676,
+      "step": 3443
+    },
+    {
+      "epoch": 0.3061333333333333,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.999974535596063e-05,
+      "loss": 0.643,
+      "step": 3444
+    },
+    {
+      "epoch": 0.30622222222222223,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9999737921521044e-05,
+      "loss": 0.7964,
+      "step": 3445
+    },
+    {
+      "epoch": 0.3063111111111111,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9999730380112987e-05,
+      "loss": 0.7437,
+      "step": 3446
+    },
+    {
+      "epoch": 0.3064,
+      "grad_norm": 1.5,
+      "learning_rate": 1.999972273173654e-05,
+      "loss": 0.6611,
+      "step": 3447
+    },
+    {
+      "epoch": 0.30648888888888887,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9999714976391784e-05,
+      "loss": 0.6296,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3065777777777778,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.99997071140788e-05,
+      "loss": 0.6972,
+      "step": 3449
+    },
+    {
+      "epoch": 0.30666666666666664,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.999969914479768e-05,
+      "loss": 0.8494,
+      "step": 3450
+    },
+    {
+      "epoch": 0.30675555555555556,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.9999691068548496e-05,
+      "loss": 0.6059,
+      "step": 3451
+    },
+    {
+      "epoch": 0.3068444444444444,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9999682885331347e-05,
+      "loss": 0.7424,
+      "step": 3452
+    },
+    {
+      "epoch": 0.30693333333333334,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.999967459514631e-05,
+      "loss": 0.6586,
+      "step": 3453
+    },
+    {
+      "epoch": 0.3070222222222222,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9999666197993485e-05,
+      "loss": 0.7336,
+      "step": 3454
+    },
+    {
+      "epoch": 0.3071111111111111,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.9999657693872955e-05,
+      "loss": 0.8322,
+      "step": 3455
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.9999649082784807e-05,
+      "loss": 0.6342,
+      "step": 3456
+    },
+    {
+      "epoch": 0.3072888888888889,
+      "grad_norm": 2.0,
+      "learning_rate": 1.999964036472914e-05,
+      "loss": 0.6653,
+      "step": 3457
+    },
+    {
+      "epoch": 0.3073777777777778,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.999963153970605e-05,
+      "loss": 0.8098,
+      "step": 3458
+    },
+    {
+      "epoch": 0.30746666666666667,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.999962260771562e-05,
+      "loss": 0.8994,
+      "step": 3459
+    },
+    {
+      "epoch": 0.3075555555555556,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9999613568757955e-05,
+      "loss": 0.72,
+      "step": 3460
+    },
+    {
+      "epoch": 0.30764444444444444,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.999960442283315e-05,
+      "loss": 0.7142,
+      "step": 3461
+    },
+    {
+      "epoch": 0.30773333333333336,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9999595169941298e-05,
+      "loss": 0.674,
+      "step": 3462
+    },
+    {
+      "epoch": 0.3078222222222222,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9999585810082503e-05,
+      "loss": 0.6468,
+      "step": 3463
+    },
+    {
+      "epoch": 0.30791111111111114,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9999576343256868e-05,
+      "loss": 0.6716,
+      "step": 3464
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9999566769464483e-05,
+      "loss": 0.7259,
+      "step": 3465
+    },
+    {
+      "epoch": 0.3080888888888889,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9999557088705464e-05,
+      "loss": 0.5901,
+      "step": 3466
+    },
+    {
+      "epoch": 0.3081777777777778,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9999547300979906e-05,
+      "loss": 0.7426,
+      "step": 3467
+    },
+    {
+      "epoch": 0.3082666666666667,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9999537406287913e-05,
+      "loss": 0.6722,
+      "step": 3468
+    },
+    {
+      "epoch": 0.30835555555555555,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9999527404629595e-05,
+      "loss": 0.7422,
+      "step": 3469
+    },
+    {
+      "epoch": 0.30844444444444447,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.999951729600506e-05,
+      "loss": 0.8377,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3085333333333333,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.999950708041441e-05,
+      "loss": 0.6326,
+      "step": 3471
+    },
+    {
+      "epoch": 0.30862222222222224,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.999949675785776e-05,
+      "loss": 0.7256,
+      "step": 3472
+    },
+    {
+      "epoch": 0.3087111111111111,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9999486328335218e-05,
+      "loss": 0.6066,
+      "step": 3473
+    },
+    {
+      "epoch": 0.3088,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.9999475791846896e-05,
+      "loss": 0.6452,
+      "step": 3474
+    },
+    {
+      "epoch": 0.3088888888888889,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9999465148392906e-05,
+      "loss": 0.5889,
+      "step": 3475
+    },
+    {
+      "epoch": 0.3089777777777778,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9999454397973363e-05,
+      "loss": 0.7326,
+      "step": 3476
+    },
+    {
+      "epoch": 0.30906666666666666,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.999944354058838e-05,
+      "loss": 0.6563,
+      "step": 3477
+    },
+    {
+      "epoch": 0.3091555555555556,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9999432576238077e-05,
+      "loss": 0.7053,
+      "step": 3478
+    },
+    {
+      "epoch": 0.30924444444444443,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.999942150492257e-05,
+      "loss": 0.6349,
+      "step": 3479
+    },
+    {
+      "epoch": 0.30933333333333335,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9999410326641974e-05,
+      "loss": 0.8925,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3094222222222222,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.999939904139641e-05,
+      "loss": 0.7421,
+      "step": 3481
+    },
+    {
+      "epoch": 0.3095111111111111,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.9999387649185998e-05,
+      "loss": 0.5909,
+      "step": 3482
+    },
+    {
+      "epoch": 0.3096,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.9999376150010868e-05,
+      "loss": 0.6228,
+      "step": 3483
+    },
+    {
+      "epoch": 0.3096888888888889,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.9999364543871133e-05,
+      "loss": 0.7311,
+      "step": 3484
+    },
+    {
+      "epoch": 0.30977777777777776,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.999935283076692e-05,
+      "loss": 0.7351,
+      "step": 3485
+    },
+    {
+      "epoch": 0.3098666666666667,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9999341010698354e-05,
+      "loss": 0.6558,
+      "step": 3486
+    },
+    {
+      "epoch": 0.30995555555555554,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9999329083665564e-05,
+      "loss": 0.7299,
+      "step": 3487
+    },
+    {
+      "epoch": 0.31004444444444446,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9999317049668677e-05,
+      "loss": 0.6262,
+      "step": 3488
+    },
+    {
+      "epoch": 0.3101333333333333,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.999930490870782e-05,
+      "loss": 0.782,
+      "step": 3489
+    },
+    {
+      "epoch": 0.31022222222222223,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9999292660783123e-05,
+      "loss": 0.5227,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3103111111111111,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9999280305894716e-05,
+      "loss": 0.7571,
+      "step": 3491
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9999267844042734e-05,
+      "loss": 0.6235,
+      "step": 3492
+    },
+    {
+      "epoch": 0.31048888888888887,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.999925527522731e-05,
+      "loss": 0.6471,
+      "step": 3493
+    },
+    {
+      "epoch": 0.3105777777777778,
+      "grad_norm": 2.0,
+      "learning_rate": 1.999924259944858e-05,
+      "loss": 0.7058,
+      "step": 3494
+    },
+    {
+      "epoch": 0.31066666666666665,
+      "grad_norm": 2.25,
+      "learning_rate": 1.999922981670667e-05,
+      "loss": 0.6309,
+      "step": 3495
+    },
+    {
+      "epoch": 0.31075555555555556,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.999921692700173e-05,
+      "loss": 0.658,
+      "step": 3496
+    },
+    {
+      "epoch": 0.3108444444444444,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.999920393033389e-05,
+      "loss": 0.6541,
+      "step": 3497
+    },
+    {
+      "epoch": 0.31093333333333334,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.999919082670329e-05,
+      "loss": 0.5905,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3110222222222222,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9999177616110072e-05,
+      "loss": 0.614,
+      "step": 3499
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9999164298554375e-05,
+      "loss": 0.6915,
+      "step": 3500
+    },
+    {
+      "epoch": 0.3112,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9999150874036344e-05,
+      "loss": 0.5873,
+      "step": 3501
+    },
+    {
+      "epoch": 0.3112888888888889,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.999913734255612e-05,
+      "loss": 0.7937,
+      "step": 3502
+    },
+    {
+      "epoch": 0.31137777777777775,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.999912370411385e-05,
+      "loss": 0.7254,
+      "step": 3503
+    },
+    {
+      "epoch": 0.31146666666666667,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.999910995870968e-05,
+      "loss": 0.6465,
+      "step": 3504
+    },
+    {
+      "epoch": 0.31155555555555553,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9999096106343755e-05,
+      "loss": 0.6362,
+      "step": 3505
+    },
+    {
+      "epoch": 0.31164444444444445,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9999082147016223e-05,
+      "loss": 0.6513,
+      "step": 3506
+    },
+    {
+      "epoch": 0.3117333333333333,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9999068080727236e-05,
+      "loss": 0.812,
+      "step": 3507
+    },
+    {
+      "epoch": 0.3118222222222222,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.999905390747694e-05,
+      "loss": 0.6888,
+      "step": 3508
+    },
+    {
+      "epoch": 0.3119111111111111,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.999903962726549e-05,
+      "loss": 0.6261,
+      "step": 3509
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9999025240093045e-05,
+      "loss": 0.6245,
+      "step": 3510
+    },
+    {
+      "epoch": 0.3120888888888889,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.9999010745959745e-05,
+      "loss": 0.7402,
+      "step": 3511
+    },
+    {
+      "epoch": 0.3121777777777778,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9998996144865756e-05,
+      "loss": 0.6526,
+      "step": 3512
+    },
+    {
+      "epoch": 0.3122666666666667,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.9998981436811233e-05,
+      "loss": 0.5962,
+      "step": 3513
+    },
+    {
+      "epoch": 0.31235555555555555,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9998966621796322e-05,
+      "loss": 0.5494,
+      "step": 3514
+    },
+    {
+      "epoch": 0.31244444444444447,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9998951699821195e-05,
+      "loss": 0.5683,
+      "step": 3515
+    },
+    {
+      "epoch": 0.31253333333333333,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.999893667088601e-05,
+      "loss": 0.5563,
+      "step": 3516
+    },
+    {
+      "epoch": 0.31262222222222225,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9998921534990922e-05,
+      "loss": 0.6722,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3127111111111111,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9998906292136096e-05,
+      "loss": 0.7879,
+      "step": 3518
+    },
+    {
+      "epoch": 0.3128,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.9998890942321693e-05,
+      "loss": 0.6837,
+      "step": 3519
+    },
+    {
+      "epoch": 0.3128888888888889,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9998875485547882e-05,
+      "loss": 0.6428,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3129777777777778,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.999885992181482e-05,
+      "loss": 0.6742,
+      "step": 3521
+    },
+    {
+      "epoch": 0.31306666666666666,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9998844251122683e-05,
+      "loss": 0.8241,
+      "step": 3522
+    },
+    {
+      "epoch": 0.3131555555555556,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9998828473471633e-05,
+      "loss": 0.6218,
+      "step": 3523
+    },
+    {
+      "epoch": 0.31324444444444444,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.999881258886184e-05,
+      "loss": 0.7611,
+      "step": 3524
+    },
+    {
+      "epoch": 0.31333333333333335,
+      "grad_norm": 2.375,
+      "learning_rate": 1.9998796597293477e-05,
+      "loss": 0.7894,
+      "step": 3525
+    },
+    {
+      "epoch": 0.3134222222222222,
+      "grad_norm": 2.0,
+      "learning_rate": 1.999878049876671e-05,
+      "loss": 0.6579,
+      "step": 3526
+    },
+    {
+      "epoch": 0.31351111111111113,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.999876429328171e-05,
+      "loss": 0.7201,
+      "step": 3527
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.9998747980838655e-05,
+      "loss": 0.516,
+      "step": 3528
+    },
+    {
+      "epoch": 0.3136888888888889,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.999873156143772e-05,
+      "loss": 0.7047,
+      "step": 3529
+    },
+    {
+      "epoch": 0.31377777777777777,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.999871503507908e-05,
+      "loss": 0.6564,
+      "step": 3530
+    },
+    {
+      "epoch": 0.3138666666666667,
+      "grad_norm": 2.125,
+      "learning_rate": 1.999869840176291e-05,
+      "loss": 0.6448,
+      "step": 3531
+    },
+    {
+      "epoch": 0.31395555555555554,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9998681661489384e-05,
+      "loss": 0.7336,
+      "step": 3532
+    },
+    {
+      "epoch": 0.31404444444444446,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.999866481425869e-05,
+      "loss": 0.695,
+      "step": 3533
+    },
+    {
+      "epoch": 0.3141333333333333,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.9998647860071e-05,
+      "loss": 0.6745,
+      "step": 3534
+    },
+    {
+      "epoch": 0.31422222222222224,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9998630798926505e-05,
+      "loss": 0.8224,
+      "step": 3535
+    },
+    {
+      "epoch": 0.3143111111111111,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9998613630825376e-05,
+      "loss": 0.6857,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3144,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9998596355767805e-05,
+      "loss": 0.8142,
+      "step": 3537
+    },
+    {
+      "epoch": 0.3144888888888889,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9998578973753974e-05,
+      "loss": 0.7448,
+      "step": 3538
+    },
+    {
+      "epoch": 0.3145777777777778,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.999856148478407e-05,
+      "loss": 0.944,
+      "step": 3539
+    },
+    {
+      "epoch": 0.31466666666666665,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.9998543888858278e-05,
+      "loss": 0.7107,
+      "step": 3540
+    },
+    {
+      "epoch": 0.31475555555555557,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.9998526185976786e-05,
+      "loss": 0.7519,
+      "step": 3541
+    },
+    {
+      "epoch": 0.3148444444444444,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.9998508376139783e-05,
+      "loss": 0.6665,
+      "step": 3542
+    },
+    {
+      "epoch": 0.31493333333333334,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9998490459347466e-05,
+      "loss": 0.7386,
+      "step": 3543
+    },
+    {
+      "epoch": 0.3150222222222222,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.999847243560002e-05,
+      "loss": 0.7321,
+      "step": 3544
+    },
+    {
+      "epoch": 0.3151111111111111,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.999845430489764e-05,
+      "loss": 0.6533,
+      "step": 3545
+    },
+    {
+      "epoch": 0.3152,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9998436067240518e-05,
+      "loss": 0.507,
+      "step": 3546
+    },
+    {
+      "epoch": 0.3152888888888889,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9998417722628854e-05,
+      "loss": 0.6226,
+      "step": 3547
+    },
+    {
+      "epoch": 0.31537777777777776,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9998399271062837e-05,
+      "loss": 0.6904,
+      "step": 3548
+    },
+    {
+      "epoch": 0.3154666666666667,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9998380712542672e-05,
+      "loss": 0.6665,
+      "step": 3549
+    },
+    {
+      "epoch": 0.31555555555555553,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9998362047068548e-05,
+      "loss": 0.6808,
+      "step": 3550
+    },
+    {
+      "epoch": 0.31564444444444445,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9998343274640674e-05,
+      "loss": 0.5385,
+      "step": 3551
+    },
+    {
+      "epoch": 0.3157333333333333,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9998324395259248e-05,
+      "loss": 0.7169,
+      "step": 3552
+    },
+    {
+      "epoch": 0.3158222222222222,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9998305408924473e-05,
+      "loss": 0.489,
+      "step": 3553
+    },
+    {
+      "epoch": 0.3159111111111111,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9998286315636548e-05,
+      "loss": 0.616,
+      "step": 3554
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.999826711539568e-05,
+      "loss": 0.7381,
+      "step": 3555
+    },
+    {
+      "epoch": 0.31608888888888886,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.999824780820207e-05,
+      "loss": 0.7734,
+      "step": 3556
+    },
+    {
+      "epoch": 0.3161777777777778,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9998228394055937e-05,
+      "loss": 0.6677,
+      "step": 3557
+    },
+    {
+      "epoch": 0.31626666666666664,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9998208872957474e-05,
+      "loss": 0.716,
+      "step": 3558
+    },
+    {
+      "epoch": 0.31635555555555556,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.99981892449069e-05,
+      "loss": 0.685,
+      "step": 3559
+    },
+    {
+      "epoch": 0.3164444444444444,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9998169509904418e-05,
+      "loss": 0.6783,
+      "step": 3560
+    },
+    {
+      "epoch": 0.31653333333333333,
+      "grad_norm": 2.0,
+      "learning_rate": 1.999814966795024e-05,
+      "loss": 0.7766,
+      "step": 3561
+    },
+    {
+      "epoch": 0.31662222222222225,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.9998129719044582e-05,
+      "loss": 0.7213,
+      "step": 3562
+    },
+    {
+      "epoch": 0.3167111111111111,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9998109663187657e-05,
+      "loss": 0.9048,
+      "step": 3563
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.999808950037968e-05,
+      "loss": 0.6231,
+      "step": 3564
+    },
+    {
+      "epoch": 0.3168888888888889,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.9998069230620864e-05,
+      "loss": 0.5537,
+      "step": 3565
+    },
+    {
+      "epoch": 0.3169777777777778,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9998048853911426e-05,
+      "loss": 0.7235,
+      "step": 3566
+    },
+    {
+      "epoch": 0.31706666666666666,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.999802837025158e-05,
+      "loss": 0.7031,
+      "step": 3567
+    },
+    {
+      "epoch": 0.3171555555555556,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.9998007779641556e-05,
+      "loss": 0.7474,
+      "step": 3568
+    },
+    {
+      "epoch": 0.31724444444444444,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9997987082081567e-05,
+      "loss": 0.689,
+      "step": 3569
+    },
+    {
+      "epoch": 0.31733333333333336,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9997966277571837e-05,
+      "loss": 0.6341,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3174222222222222,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9997945366112583e-05,
+      "loss": 0.5863,
+      "step": 3571
+    },
+    {
+      "epoch": 0.31751111111111113,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9997924347704036e-05,
+      "loss": 0.5973,
+      "step": 3572
+    },
+    {
+      "epoch": 0.3176,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9997903222346418e-05,
+      "loss": 0.587,
+      "step": 3573
+    },
+    {
+      "epoch": 0.3176888888888889,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.999788199003995e-05,
+      "loss": 0.829,
+      "step": 3574
+    },
+    {
+      "epoch": 0.31777777777777777,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.9997860650784867e-05,
+      "loss": 0.6523,
+      "step": 3575
+    },
+    {
+      "epoch": 0.3178666666666667,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9997839204581395e-05,
+      "loss": 0.8199,
+      "step": 3576
+    },
+    {
+      "epoch": 0.31795555555555555,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9997817651429763e-05,
+      "loss": 0.8038,
+      "step": 3577
+    },
+    {
+      "epoch": 0.31804444444444446,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9997795991330197e-05,
+      "loss": 0.6777,
+      "step": 3578
+    },
+    {
+      "epoch": 0.3181333333333333,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9997774224282936e-05,
+      "loss": 0.5971,
+      "step": 3579
+    },
+    {
+      "epoch": 0.31822222222222224,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9997752350288208e-05,
+      "loss": 0.7338,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3183111111111111,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9997730369346252e-05,
+      "loss": 0.565,
+      "step": 3581
+    },
+    {
+      "epoch": 0.3184,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9997708281457293e-05,
+      "loss": 0.61,
+      "step": 3582
+    },
+    {
+      "epoch": 0.3184888888888889,
+      "grad_norm": 1.875,
+      "learning_rate": 1.999768608662158e-05,
+      "loss": 0.6258,
+      "step": 3583
+    },
+    {
+      "epoch": 0.3185777777777778,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9997663784839342e-05,
+      "loss": 0.5313,
+      "step": 3584
+    },
+    {
+      "epoch": 0.31866666666666665,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9997641376110816e-05,
+      "loss": 0.6513,
+      "step": 3585
+    },
+    {
+      "epoch": 0.31875555555555557,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.9997618860436253e-05,
+      "loss": 0.6822,
+      "step": 3586
+    },
+    {
+      "epoch": 0.31884444444444443,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9997596237815883e-05,
+      "loss": 0.7051,
+      "step": 3587
+    },
+    {
+      "epoch": 0.31893333333333335,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.999757350824995e-05,
+      "loss": 0.8623,
+      "step": 3588
+    },
+    {
+      "epoch": 0.3190222222222222,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9997550671738703e-05,
+      "loss": 0.6637,
+      "step": 3589
+    },
+    {
+      "epoch": 0.3191111111111111,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9997527728282374e-05,
+      "loss": 0.6189,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3192,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9997504677881224e-05,
+      "loss": 0.6035,
+      "step": 3591
+    },
+    {
+      "epoch": 0.3192888888888889,
+      "grad_norm": 2.5,
+      "learning_rate": 1.999748152053549e-05,
+      "loss": 0.7209,
+      "step": 3592
+    },
+    {
+      "epoch": 0.31937777777777776,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9997458256245418e-05,
+      "loss": 0.6041,
+      "step": 3593
+    },
+    {
+      "epoch": 0.3194666666666667,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9997434885011264e-05,
+      "loss": 0.6669,
+      "step": 3594
+    },
+    {
+      "epoch": 0.31955555555555554,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9997411406833275e-05,
+      "loss": 0.7016,
+      "step": 3595
+    },
+    {
+      "epoch": 0.31964444444444445,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.99973878217117e-05,
+      "loss": 0.7097,
+      "step": 3596
+    },
+    {
+      "epoch": 0.3197333333333333,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.999736412964679e-05,
+      "loss": 0.7624,
+      "step": 3597
+    },
+    {
+      "epoch": 0.31982222222222223,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.9997340330638805e-05,
+      "loss": 0.6298,
+      "step": 3598
+    },
+    {
+      "epoch": 0.3199111111111111,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9997316424687998e-05,
+      "loss": 0.6457,
+      "step": 3599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.999729241179462e-05,
+      "loss": 0.6958,
+      "step": 3600
+    },
+    {
+      "epoch": 0.32008888888888887,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.999726829195893e-05,
+      "loss": 0.5394,
+      "step": 3601
+    },
+    {
+      "epoch": 0.3201777777777778,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9997244065181184e-05,
+      "loss": 0.5557,
+      "step": 3602
+    },
+    {
+      "epoch": 0.32026666666666664,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.999721973146165e-05,
+      "loss": 0.9113,
+      "step": 3603
+    },
+    {
+      "epoch": 0.32035555555555556,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.999719529080058e-05,
+      "loss": 0.5581,
+      "step": 3604
+    },
+    {
+      "epoch": 0.3204444444444444,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9997170743198235e-05,
+      "loss": 0.6661,
+      "step": 3605
+    },
+    {
+      "epoch": 0.32053333333333334,
+      "grad_norm": 2.125,
+      "learning_rate": 1.999714608865488e-05,
+      "loss": 0.7594,
+      "step": 3606
+    },
+    {
+      "epoch": 0.3206222222222222,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.999712132717078e-05,
+      "loss": 0.7895,
+      "step": 3607
+    },
+    {
+      "epoch": 0.3207111111111111,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9997096458746198e-05,
+      "loss": 0.7078,
+      "step": 3608
+    },
+    {
+      "epoch": 0.3208,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.99970714833814e-05,
+      "loss": 0.6943,
+      "step": 3609
+    },
+    {
+      "epoch": 0.3208888888888889,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9997046401076658e-05,
+      "loss": 0.853,
+      "step": 3610
+    },
+    {
+      "epoch": 0.32097777777777775,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9997021211832233e-05,
+      "loss": 0.6296,
+      "step": 3611
+    },
+    {
+      "epoch": 0.32106666666666667,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.9996995915648398e-05,
+      "loss": 0.6849,
+      "step": 3612
+    },
+    {
+      "epoch": 0.3211555555555556,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.9996970512525423e-05,
+      "loss": 0.5959,
+      "step": 3613
+    },
+    {
+      "epoch": 0.32124444444444444,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.999694500246358e-05,
+      "loss": 0.8167,
+      "step": 3614
+    },
+    {
+      "epoch": 0.32133333333333336,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.999691938546314e-05,
+      "loss": 0.8312,
+      "step": 3615
+    },
+    {
+      "epoch": 0.3214222222222222,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.9996893661524383e-05,
+      "loss": 0.699,
+      "step": 3616
+    },
+    {
+      "epoch": 0.32151111111111114,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9996867830647577e-05,
+      "loss": 0.9276,
+      "step": 3617
+    },
+    {
+      "epoch": 0.3216,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9996841892833e-05,
+      "loss": 0.5655,
+      "step": 3618
+    },
+    {
+      "epoch": 0.3216888888888889,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.9996815848080936e-05,
+      "loss": 0.6716,
+      "step": 3619
+    },
+    {
+      "epoch": 0.3217777777777778,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9996789696391656e-05,
+      "loss": 0.5602,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3218666666666667,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9996763437765438e-05,
+      "loss": 0.6692,
+      "step": 3621
+    },
+    {
+      "epoch": 0.32195555555555555,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.999673707220257e-05,
+      "loss": 0.7671,
+      "step": 3622
+    },
+    {
+      "epoch": 0.32204444444444447,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.999671059970333e-05,
+      "loss": 0.6256,
+      "step": 3623
+    },
+    {
+      "epoch": 0.3221333333333333,
+      "grad_norm": 1.75,
+      "learning_rate": 1.999668402026801e-05,
+      "loss": 0.7165,
+      "step": 3624
+    },
+    {
+      "epoch": 0.32222222222222224,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9996657333896875e-05,
+      "loss": 0.6803,
+      "step": 3625
+    },
+    {
+      "epoch": 0.3223111111111111,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9996630540590226e-05,
+      "loss": 0.6858,
+      "step": 3626
+    },
+    {
+      "epoch": 0.3224,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.999660364034835e-05,
+      "loss": 0.7241,
+      "step": 3627
+    },
+    {
+      "epoch": 0.3224888888888889,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9996576633171526e-05,
+      "loss": 0.5894,
+      "step": 3628
+    },
+    {
+      "epoch": 0.3225777777777778,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.999654951906005e-05,
+      "loss": 0.5262,
+      "step": 3629
+    },
+    {
+      "epoch": 0.32266666666666666,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9996522298014208e-05,
+      "loss": 0.6568,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3227555555555556,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.999649497003429e-05,
+      "loss": 0.8191,
+      "step": 3631
+    },
+    {
+      "epoch": 0.32284444444444443,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.9996467535120593e-05,
+      "loss": 0.7683,
+      "step": 3632
+    },
+    {
+      "epoch": 0.32293333333333335,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.999643999327341e-05,
+      "loss": 0.8161,
+      "step": 3633
+    },
+    {
+      "epoch": 0.3230222222222222,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9996412344493033e-05,
+      "loss": 0.6683,
+      "step": 3634
+    },
+    {
+      "epoch": 0.3231111111111111,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9996384588779757e-05,
+      "loss": 0.6889,
+      "step": 3635
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 1.625,
+      "learning_rate": 1.999635672613388e-05,
+      "loss": 0.5457,
+      "step": 3636
+    },
+    {
+      "epoch": 0.3232888888888889,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9996328756555704e-05,
+      "loss": 0.5606,
+      "step": 3637
+    },
+    {
+      "epoch": 0.32337777777777776,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9996300680045525e-05,
+      "loss": 0.791,
+      "step": 3638
+    },
+    {
+      "epoch": 0.3234666666666667,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.999627249660364e-05,
+      "loss": 0.6469,
+      "step": 3639
+    },
+    {
+      "epoch": 0.32355555555555554,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9996244206230355e-05,
+      "loss": 0.6168,
+      "step": 3640
+    },
+    {
+      "epoch": 0.32364444444444446,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9996215808925973e-05,
+      "loss": 0.6693,
+      "step": 3641
+    },
+    {
+      "epoch": 0.3237333333333333,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9996187304690797e-05,
+      "loss": 0.7177,
+      "step": 3642
+    },
+    {
+      "epoch": 0.32382222222222223,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9996158693525127e-05,
+      "loss": 0.6654,
+      "step": 3643
+    },
+    {
+      "epoch": 0.3239111111111111,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9996129975429274e-05,
+      "loss": 0.6741,
+      "step": 3644
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.9996101150403543e-05,
+      "loss": 0.7455,
+      "step": 3645
+    },
+    {
+      "epoch": 0.32408888888888887,
+      "grad_norm": 2.375,
+      "learning_rate": 1.999607221844825e-05,
+      "loss": 0.5863,
+      "step": 3646
+    },
+    {
+      "epoch": 0.3241777777777778,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.999604317956369e-05,
+      "loss": 0.8115,
+      "step": 3647
+    },
+    {
+      "epoch": 0.32426666666666665,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9996014033750185e-05,
+      "loss": 0.6702,
+      "step": 3648
+    },
+    {
+      "epoch": 0.32435555555555556,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9995984781008046e-05,
+      "loss": 0.6459,
+      "step": 3649
+    },
+    {
+      "epoch": 0.3244444444444444,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9995955421337583e-05,
+      "loss": 0.6036,
+      "step": 3650
+    },
+    {
+      "epoch": 0.32453333333333334,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9995925954739107e-05,
+      "loss": 0.7277,
+      "step": 3651
+    },
+    {
+      "epoch": 0.3246222222222222,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.999589638121294e-05,
+      "loss": 0.6389,
+      "step": 3652
+    },
+    {
+      "epoch": 0.3247111111111111,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9995866700759393e-05,
+      "loss": 0.7309,
+      "step": 3653
+    },
+    {
+      "epoch": 0.3248,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9995836913378786e-05,
+      "loss": 0.7387,
+      "step": 3654
+    },
+    {
+      "epoch": 0.3248888888888889,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.9995807019071436e-05,
+      "loss": 0.7236,
+      "step": 3655
+    },
+    {
+      "epoch": 0.32497777777777775,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.9995777017837666e-05,
+      "loss": 0.7741,
+      "step": 3656
+    },
+    {
+      "epoch": 0.32506666666666667,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9995746909677795e-05,
+      "loss": 0.7484,
+      "step": 3657
+    },
+    {
+      "epoch": 0.32515555555555553,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9995716694592146e-05,
+      "loss": 0.6341,
+      "step": 3658
+    },
+    {
+      "epoch": 0.32524444444444445,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.999568637258104e-05,
+      "loss": 0.6761,
+      "step": 3659
+    },
+    {
+      "epoch": 0.3253333333333333,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.99956559436448e-05,
+      "loss": 0.8466,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3254222222222222,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.9995625407783755e-05,
+      "loss": 0.6128,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3255111111111111,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.999559476499823e-05,
+      "loss": 0.6461,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3256,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9995564015288555e-05,
+      "loss": 0.6287,
+      "step": 3663
+    },
+    {
+      "epoch": 0.3256888888888889,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9995533158655057e-05,
+      "loss": 0.7851,
+      "step": 3664
+    },
+    {
+      "epoch": 0.3257777777777778,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9995502195098065e-05,
+      "loss": 0.6265,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3258666666666667,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.999547112461792e-05,
+      "loss": 0.5256,
+      "step": 3666
+    },
+    {
+      "epoch": 0.32595555555555555,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.9995439947214937e-05,
+      "loss": 0.5884,
+      "step": 3667
+    },
+    {
+      "epoch": 0.32604444444444447,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9995408662889462e-05,
+      "loss": 0.721,
+      "step": 3668
+    },
+    {
+      "epoch": 0.32613333333333333,
+      "grad_norm": 2.5,
+      "learning_rate": 1.9995377271641825e-05,
+      "loss": 0.8275,
+      "step": 3669
+    },
+    {
+      "epoch": 0.32622222222222225,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9995345773472362e-05,
+      "loss": 0.7747,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3263111111111111,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.999531416838141e-05,
+      "loss": 0.7972,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.9995282456369313e-05,
+      "loss": 0.6885,
+      "step": 3672
+    },
+    {
+      "epoch": 0.3264888888888889,
+      "grad_norm": 2.875,
+      "learning_rate": 1.9995250637436404e-05,
+      "loss": 0.609,
+      "step": 3673
+    },
+    {
+      "epoch": 0.3265777777777778,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.999521871158302e-05,
+      "loss": 0.5862,
+      "step": 3674
+    },
+    {
+      "epoch": 0.32666666666666666,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9995186678809513e-05,
+      "loss": 0.6006,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3267555555555556,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.999515453911622e-05,
+      "loss": 0.6359,
+      "step": 3676
+    },
+    {
+      "epoch": 0.32684444444444444,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.999512229250348e-05,
+      "loss": 0.7695,
+      "step": 3677
+    },
+    {
+      "epoch": 0.32693333333333335,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9995089938971642e-05,
+      "loss": 0.6658,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3270222222222222,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9995057478521055e-05,
+      "loss": 0.7147,
+      "step": 3679
+    },
+    {
+      "epoch": 0.32711111111111113,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9995024911152067e-05,
+      "loss": 0.7346,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3272,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.9994992236865017e-05,
+      "loss": 0.6207,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3272888888888889,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9994959455660265e-05,
+      "loss": 0.6857,
+      "step": 3682
+    },
+    {
+      "epoch": 0.32737777777777777,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9994926567538158e-05,
+      "loss": 0.6707,
+      "step": 3683
+    },
+    {
+      "epoch": 0.3274666666666667,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9994893572499047e-05,
+      "loss": 0.6591,
+      "step": 3684
+    },
+    {
+      "epoch": 0.32755555555555554,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9994860470543283e-05,
+      "loss": 0.6269,
+      "step": 3685
+    },
+    {
+      "epoch": 0.32764444444444446,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9994827261671223e-05,
+      "loss": 0.574,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3277333333333333,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.999479394588322e-05,
+      "loss": 0.5935,
+      "step": 3687
+    },
+    {
+      "epoch": 0.32782222222222224,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.9994760523179635e-05,
+      "loss": 0.7204,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3279111111111111,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9994726993560822e-05,
+      "loss": 0.594,
+      "step": 3689
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.999469335702714e-05,
+      "loss": 0.8412,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3280888888888889,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.999465961357895e-05,
+      "loss": 0.7419,
+      "step": 3691
+    },
+    {
+      "epoch": 0.3281777777777778,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.999462576321661e-05,
+      "loss": 0.7752,
+      "step": 3692
+    },
+    {
+      "epoch": 0.32826666666666665,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.999459180594048e-05,
+      "loss": 0.7048,
+      "step": 3693
+    },
+    {
+      "epoch": 0.32835555555555557,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9994557741750932e-05,
+      "loss": 0.6796,
+      "step": 3694
+    },
+    {
+      "epoch": 0.32844444444444443,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.999452357064833e-05,
+      "loss": 0.6739,
+      "step": 3695
+    },
+    {
+      "epoch": 0.32853333333333334,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9994489292633027e-05,
+      "loss": 0.7866,
+      "step": 3696
+    },
+    {
+      "epoch": 0.3286222222222222,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.99944549077054e-05,
+      "loss": 0.6225,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3287111111111111,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.999442041586582e-05,
+      "loss": 0.6453,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3288,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9994385817114644e-05,
+      "loss": 0.6364,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3288888888888889,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9994351111452252e-05,
+      "loss": 0.7243,
+      "step": 3700
+    },
+    {
+      "epoch": 0.32897777777777776,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.999431629887901e-05,
+      "loss": 0.6383,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3290666666666667,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9994281379395294e-05,
+      "loss": 0.7024,
+      "step": 3702
+    },
+    {
+      "epoch": 0.32915555555555553,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9994246353001477e-05,
+      "loss": 0.7296,
+      "step": 3703
+    },
+    {
+      "epoch": 0.32924444444444445,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.999421121969793e-05,
+      "loss": 0.718,
+      "step": 3704
+    },
+    {
+      "epoch": 0.3293333333333333,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.999417597948503e-05,
+      "loss": 0.6316,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3294222222222222,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9994140632363162e-05,
+      "loss": 0.5877,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3295111111111111,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.9994105178332693e-05,
+      "loss": 0.7131,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 2.0,
+      "learning_rate": 1.999406961739401e-05,
+      "loss": 0.6754,
+      "step": 3708
+    },
+    {
+      "epoch": 0.32968888888888886,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9994033949547484e-05,
+      "loss": 0.8837,
+      "step": 3709
+    },
+    {
+      "epoch": 0.3297777777777778,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9993998174793505e-05,
+      "loss": 0.6757,
+      "step": 3710
+    },
+    {
+      "epoch": 0.32986666666666664,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9993962293132456e-05,
+      "loss": 0.6087,
+      "step": 3711
+    },
+    {
+      "epoch": 0.32995555555555556,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9993926304564716e-05,
+      "loss": 0.7367,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3300444444444444,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.9993890209090674e-05,
+      "loss": 0.6839,
+      "step": 3713
+    },
+    {
+      "epoch": 0.33013333333333333,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.9993854006710713e-05,
+      "loss": 0.5588,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3302222222222222,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9993817697425222e-05,
+      "loss": 0.6143,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3303111111111111,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.999378128123459e-05,
+      "loss": 0.8005,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3304,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9993744758139203e-05,
+      "loss": 0.7111,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3304888888888889,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9993708128139456e-05,
+      "loss": 0.6453,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3305777777777778,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.999367139123574e-05,
+      "loss": 0.7429,
+      "step": 3719
+    },
+    {
+      "epoch": 0.33066666666666666,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9993634547428443e-05,
+      "loss": 0.8892,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3307555555555556,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.999359759671797e-05,
+      "loss": 0.6823,
+      "step": 3721
+    },
+    {
+      "epoch": 0.33084444444444444,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.9993560539104702e-05,
+      "loss": 0.708,
+      "step": 3722
+    },
+    {
+      "epoch": 0.33093333333333336,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9993523374589046e-05,
+      "loss": 0.644,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3310222222222222,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.9993486103171394e-05,
+      "loss": 0.6739,
+      "step": 3724
+    },
+    {
+      "epoch": 0.33111111111111113,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.999344872485215e-05,
+      "loss": 0.6395,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3312,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9993411239631713e-05,
+      "loss": 0.5541,
+      "step": 3726
+    },
+    {
+      "epoch": 0.3312888888888889,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9993373647510478e-05,
+      "loss": 0.6442,
+      "step": 3727
+    },
+    {
+      "epoch": 0.33137777777777777,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9993335948488856e-05,
+      "loss": 0.6237,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3314666666666667,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9993298142567238e-05,
+      "loss": 0.71,
+      "step": 3729
+    },
+    {
+      "epoch": 0.33155555555555555,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.999326022974604e-05,
+      "loss": 0.5995,
+      "step": 3730
+    },
+    {
+      "epoch": 0.33164444444444446,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9993222210025665e-05,
+      "loss": 0.6352,
+      "step": 3731
+    },
+    {
+      "epoch": 0.3317333333333333,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9993184083406518e-05,
+      "loss": 0.7795,
+      "step": 3732
+    },
+    {
+      "epoch": 0.33182222222222224,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9993145849889003e-05,
+      "loss": 0.7148,
+      "step": 3733
+    },
+    {
+      "epoch": 0.3319111111111111,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9993107509473532e-05,
+      "loss": 0.7047,
+      "step": 3734
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.999306906216052e-05,
+      "loss": 0.8091,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3320888888888889,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9993030507950373e-05,
+      "loss": 0.6202,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3321777777777778,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9992991846843505e-05,
+      "loss": 0.7111,
+      "step": 3737
+    },
+    {
+      "epoch": 0.33226666666666665,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9992953078840327e-05,
+      "loss": 0.7022,
+      "step": 3738
+    },
+    {
+      "epoch": 0.33235555555555557,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.999291420394126e-05,
+      "loss": 0.626,
+      "step": 3739
+    },
+    {
+      "epoch": 0.33244444444444443,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.999287522214671e-05,
+      "loss": 0.6979,
+      "step": 3740
+    },
+    {
+      "epoch": 0.33253333333333335,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9992836133457106e-05,
+      "loss": 0.7236,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3326222222222222,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9992796937872855e-05,
+      "loss": 0.5716,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3327111111111111,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9992757635394384e-05,
+      "loss": 0.6808,
+      "step": 3743
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.999271822602211e-05,
+      "loss": 0.7541,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3328888888888889,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9992678709756458e-05,
+      "loss": 0.7593,
+      "step": 3745
+    },
+    {
+      "epoch": 0.33297777777777776,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.999263908659784e-05,
+      "loss": 0.6509,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3330666666666667,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.9992599356546694e-05,
+      "loss": 0.5433,
+      "step": 3747
+    },
+    {
+      "epoch": 0.33315555555555554,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9992559519603437e-05,
+      "loss": 0.9097,
+      "step": 3748
+    },
+    {
+      "epoch": 0.33324444444444445,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9992519575768495e-05,
+      "loss": 0.64,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9992479525042305e-05,
+      "loss": 0.6749,
+      "step": 3750
+    },
+    {
+      "epoch": 0.33342222222222223,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.999243936742528e-05,
+      "loss": 0.5725,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3335111111111111,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.999239910291786e-05,
+      "loss": 0.6221,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3336,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.999235873152047e-05,
+      "loss": 0.5148,
+      "step": 3753
+    },
+    {
+      "epoch": 0.33368888888888887,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9992318253233548e-05,
+      "loss": 0.6828,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3337777777777778,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9992277668057524e-05,
+      "loss": 0.7534,
+      "step": 3755
+    },
+    {
+      "epoch": 0.33386666666666664,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9992236975992826e-05,
+      "loss": 0.5599,
+      "step": 3756
+    },
+    {
+      "epoch": 0.33395555555555556,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.99921961770399e-05,
+      "loss": 0.7859,
+      "step": 3757
+    },
+    {
+      "epoch": 0.3340444444444444,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.999215527119918e-05,
+      "loss": 0.7321,
+      "step": 3758
+    },
+    {
+      "epoch": 0.33413333333333334,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9992114258471097e-05,
+      "loss": 0.7268,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3342222222222222,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.9992073138856094e-05,
+      "loss": 0.7759,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3343111111111111,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9992031912354613e-05,
+      "loss": 0.6606,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3344,
+      "grad_norm": 2.5,
+      "learning_rate": 1.999199057896709e-05,
+      "loss": 0.8266,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3344888888888889,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.9991949138693973e-05,
+      "loss": 0.7065,
+      "step": 3763
+    },
+    {
+      "epoch": 0.33457777777777775,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.99919075915357e-05,
+      "loss": 0.7398,
+      "step": 3764
+    },
+    {
+      "epoch": 0.33466666666666667,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.999186593749272e-05,
+      "loss": 0.584,
+      "step": 3765
+    },
+    {
+      "epoch": 0.33475555555555553,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9991824176565474e-05,
+      "loss": 0.6534,
+      "step": 3766
+    },
+    {
+      "epoch": 0.33484444444444444,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.999178230875441e-05,
+      "loss": 0.5725,
+      "step": 3767
+    },
+    {
+      "epoch": 0.33493333333333336,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.999174033405998e-05,
+      "loss": 0.7936,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3350222222222222,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9991698252482627e-05,
+      "loss": 0.5687,
+      "step": 3769
+    },
+    {
+      "epoch": 0.33511111111111114,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9991656064022806e-05,
+      "loss": 0.6042,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3352,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9991613768680965e-05,
+      "loss": 0.7346,
+      "step": 3771
+    },
+    {
+      "epoch": 0.3352888888888889,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9991571366457557e-05,
+      "loss": 0.6838,
+      "step": 3772
+    },
+    {
+      "epoch": 0.3353777777777778,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.999152885735304e-05,
+      "loss": 0.5837,
+      "step": 3773
+    },
+    {
+      "epoch": 0.3354666666666667,
+      "grad_norm": 2.5,
+      "learning_rate": 1.9991486241367857e-05,
+      "loss": 0.6235,
+      "step": 3774
+    },
+    {
+      "epoch": 0.33555555555555555,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.999144351850248e-05,
+      "loss": 0.7509,
+      "step": 3775
+    },
+    {
+      "epoch": 0.33564444444444447,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9991400688757356e-05,
+      "loss": 0.6822,
+      "step": 3776
+    },
+    {
+      "epoch": 0.33573333333333333,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.9991357752132943e-05,
+      "loss": 0.8595,
+      "step": 3777
+    },
+    {
+      "epoch": 0.33582222222222224,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9991314708629703e-05,
+      "loss": 0.7272,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3359111111111111,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.99912715582481e-05,
+      "loss": 0.5646,
+      "step": 3779
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9991228300988586e-05,
+      "loss": 0.7957,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3360888888888889,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.999118493685163e-05,
+      "loss": 0.578,
+      "step": 3781
+    },
+    {
+      "epoch": 0.3361777777777778,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.9991141465837697e-05,
+      "loss": 0.758,
+      "step": 3782
+    },
+    {
+      "epoch": 0.33626666666666666,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9991097887947245e-05,
+      "loss": 0.6548,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3363555555555556,
+      "grad_norm": 2.25,
+      "learning_rate": 1.999105420318075e-05,
+      "loss": 0.8525,
+      "step": 3784
+    },
+    {
+      "epoch": 0.33644444444444443,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9991010411538675e-05,
+      "loss": 0.6829,
+      "step": 3785
+    },
+    {
+      "epoch": 0.33653333333333335,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9990966513021485e-05,
+      "loss": 0.7479,
+      "step": 3786
+    },
+    {
+      "epoch": 0.3366222222222222,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9990922507629653e-05,
+      "loss": 0.7399,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3367111111111111,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9990878395363648e-05,
+      "loss": 0.6536,
+      "step": 3788
+    },
+    {
+      "epoch": 0.3368,
+      "grad_norm": 2.125,
+      "learning_rate": 1.999083417622395e-05,
+      "loss": 0.7393,
+      "step": 3789
+    },
+    {
+      "epoch": 0.3368888888888889,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9990789850211017e-05,
+      "loss": 0.5359,
+      "step": 3790
+    },
+    {
+      "epoch": 0.33697777777777776,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.9990745417325335e-05,
+      "loss": 0.8339,
+      "step": 3791
+    },
+    {
+      "epoch": 0.3370666666666667,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9990700877567372e-05,
+      "loss": 0.7831,
+      "step": 3792
+    },
+    {
+      "epoch": 0.33715555555555554,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9990656230937612e-05,
+      "loss": 0.6616,
+      "step": 3793
+    },
+    {
+      "epoch": 0.33724444444444446,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.999061147743653e-05,
+      "loss": 0.6326,
+      "step": 3794
+    },
+    {
+      "epoch": 0.3373333333333333,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9990566617064598e-05,
+      "loss": 0.7129,
+      "step": 3795
+    },
+    {
+      "epoch": 0.33742222222222223,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9990521649822307e-05,
+      "loss": 0.642,
+      "step": 3796
+    },
+    {
+      "epoch": 0.3375111111111111,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9990476575710126e-05,
+      "loss": 0.671,
+      "step": 3797
+    },
+    {
+      "epoch": 0.3376,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9990431394728547e-05,
+      "loss": 0.6781,
+      "step": 3798
+    },
+    {
+      "epoch": 0.33768888888888887,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.999038610687805e-05,
+      "loss": 0.7772,
+      "step": 3799
+    },
+    {
+      "epoch": 0.3377777777777778,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.999034071215912e-05,
+      "loss": 0.6215,
+      "step": 3800
+    },
+    {
+      "epoch": 0.33786666666666665,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.999029521057224e-05,
+      "loss": 0.6448,
+      "step": 3801
+    },
+    {
+      "epoch": 0.33795555555555556,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9990249602117898e-05,
+      "loss": 0.6216,
+      "step": 3802
+    },
+    {
+      "epoch": 0.3380444444444444,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9990203886796582e-05,
+      "loss": 0.7648,
+      "step": 3803
+    },
+    {
+      "epoch": 0.33813333333333334,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.9990158064608782e-05,
+      "loss": 0.7643,
+      "step": 3804
+    },
+    {
+      "epoch": 0.3382222222222222,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9990112135554985e-05,
+      "loss": 0.8405,
+      "step": 3805
+    },
+    {
+      "epoch": 0.3383111111111111,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.999006609963569e-05,
+      "loss": 0.6351,
+      "step": 3806
+    },
+    {
+      "epoch": 0.3384,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9990019956851384e-05,
+      "loss": 0.6902,
+      "step": 3807
+    },
+    {
+      "epoch": 0.3384888888888889,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9989973707202556e-05,
+      "loss": 0.5174,
+      "step": 3808
+    },
+    {
+      "epoch": 0.33857777777777776,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.998992735068971e-05,
+      "loss": 0.7548,
+      "step": 3809
+    },
+    {
+      "epoch": 0.33866666666666667,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9989880887313337e-05,
+      "loss": 0.6934,
+      "step": 3810
+    },
+    {
+      "epoch": 0.33875555555555553,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9989834317073932e-05,
+      "loss": 0.6528,
+      "step": 3811
+    },
+    {
+      "epoch": 0.33884444444444445,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.9989787639971997e-05,
+      "loss": 0.8423,
+      "step": 3812
+    },
+    {
+      "epoch": 0.3389333333333333,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.998974085600803e-05,
+      "loss": 0.7044,
+      "step": 3813
+    },
+    {
+      "epoch": 0.3390222222222222,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9989693965182534e-05,
+      "loss": 0.7745,
+      "step": 3814
+    },
+    {
+      "epoch": 0.3391111111111111,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9989646967496003e-05,
+      "loss": 0.6671,
+      "step": 3815
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.998959986294895e-05,
+      "loss": 0.6709,
+      "step": 3816
+    },
+    {
+      "epoch": 0.33928888888888886,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.998955265154187e-05,
+      "loss": 0.6481,
+      "step": 3817
+    },
+    {
+      "epoch": 0.3393777777777778,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9989505333275273e-05,
+      "loss": 0.7867,
+      "step": 3818
+    },
+    {
+      "epoch": 0.3394666666666667,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9989457908149666e-05,
+      "loss": 0.6888,
+      "step": 3819
+    },
+    {
+      "epoch": 0.33955555555555555,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9989410376165553e-05,
+      "loss": 0.6403,
+      "step": 3820
+    },
+    {
+      "epoch": 0.33964444444444447,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9989362737323442e-05,
+      "loss": 0.5825,
+      "step": 3821
+    },
+    {
+      "epoch": 0.33973333333333333,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9989314991623847e-05,
+      "loss": 0.6486,
+      "step": 3822
+    },
+    {
+      "epoch": 0.33982222222222225,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.9989267139067274e-05,
+      "loss": 0.5911,
+      "step": 3823
+    },
+    {
+      "epoch": 0.3399111111111111,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9989219179654235e-05,
+      "loss": 0.7726,
+      "step": 3824
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.998917111338525e-05,
+      "loss": 0.7583,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3400888888888889,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.998912294026082e-05,
+      "loss": 0.7606,
+      "step": 3826
+    },
+    {
+      "epoch": 0.3401777777777778,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9989074660281475e-05,
+      "loss": 0.6386,
+      "step": 3827
+    },
+    {
+      "epoch": 0.34026666666666666,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9989026273447723e-05,
+      "loss": 0.7039,
+      "step": 3828
+    },
+    {
+      "epoch": 0.3403555555555556,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9988977779760085e-05,
+      "loss": 0.582,
+      "step": 3829
+    },
+    {
+      "epoch": 0.34044444444444444,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9988929179219077e-05,
+      "loss": 0.693,
+      "step": 3830
+    },
+    {
+      "epoch": 0.34053333333333335,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9988880471825222e-05,
+      "loss": 0.5461,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3406222222222222,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.9988831657579035e-05,
+      "loss": 0.619,
+      "step": 3832
+    },
+    {
+      "epoch": 0.34071111111111113,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9988782736481044e-05,
+      "loss": 0.6737,
+      "step": 3833
+    },
+    {
+      "epoch": 0.3408,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.9988733708531772e-05,
+      "loss": 0.5664,
+      "step": 3834
+    },
+    {
+      "epoch": 0.3408888888888889,
+      "grad_norm": 2.25,
+      "learning_rate": 1.998868457373174e-05,
+      "loss": 0.6507,
+      "step": 3835
+    },
+    {
+      "epoch": 0.34097777777777777,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9988635332081475e-05,
+      "loss": 0.6833,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3410666666666667,
+      "grad_norm": 1.5,
+      "learning_rate": 1.998858598358151e-05,
+      "loss": 0.4849,
+      "step": 3837
+    },
+    {
+      "epoch": 0.34115555555555555,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9988536528232364e-05,
+      "loss": 0.7889,
+      "step": 3838
+    },
+    {
+      "epoch": 0.34124444444444446,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9988486966034566e-05,
+      "loss": 0.7991,
+      "step": 3839
+    },
+    {
+      "epoch": 0.3413333333333333,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9988437296988655e-05,
+      "loss": 0.7302,
+      "step": 3840
+    },
+    {
+      "epoch": 0.34142222222222224,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.9988387521095155e-05,
+      "loss": 0.7098,
+      "step": 3841
+    },
+    {
+      "epoch": 0.3415111111111111,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.9988337638354598e-05,
+      "loss": 0.7445,
+      "step": 3842
+    },
+    {
+      "epoch": 0.3416,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.9988287648767523e-05,
+      "loss": 0.8575,
+      "step": 3843
+    },
+    {
+      "epoch": 0.3416888888888889,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.998823755233446e-05,
+      "loss": 0.677,
+      "step": 3844
+    },
+    {
+      "epoch": 0.3417777777777778,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9988187349055942e-05,
+      "loss": 0.9379,
+      "step": 3845
+    },
+    {
+      "epoch": 0.34186666666666665,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9988137038932518e-05,
+      "loss": 0.6178,
+      "step": 3846
+    },
+    {
+      "epoch": 0.34195555555555557,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9988086621964714e-05,
+      "loss": 0.6454,
+      "step": 3847
+    },
+    {
+      "epoch": 0.34204444444444443,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9988036098153078e-05,
+      "loss": 0.6122,
+      "step": 3848
+    },
+    {
+      "epoch": 0.34213333333333334,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.9987985467498142e-05,
+      "loss": 0.5339,
+      "step": 3849
+    },
+    {
+      "epoch": 0.3422222222222222,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9987934730000457e-05,
+      "loss": 0.6708,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3423111111111111,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9987883885660556e-05,
+      "loss": 0.8308,
+      "step": 3851
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.9987832934478992e-05,
+      "loss": 0.7603,
+      "step": 3852
+    },
+    {
+      "epoch": 0.3424888888888889,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9987781876456304e-05,
+      "loss": 0.8929,
+      "step": 3853
+    },
+    {
+      "epoch": 0.34257777777777776,
+      "grad_norm": 2.0,
+      "learning_rate": 1.998773071159304e-05,
+      "loss": 0.8809,
+      "step": 3854
+    },
+    {
+      "epoch": 0.3426666666666667,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9987679439889747e-05,
+      "loss": 0.7838,
+      "step": 3855
+    },
+    {
+      "epoch": 0.34275555555555554,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9987628061346973e-05,
+      "loss": 0.7319,
+      "step": 3856
+    },
+    {
+      "epoch": 0.34284444444444445,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.9987576575965266e-05,
+      "loss": 0.6526,
+      "step": 3857
+    },
+    {
+      "epoch": 0.3429333333333333,
+      "grad_norm": 2.375,
+      "learning_rate": 1.9987524983745183e-05,
+      "loss": 0.5733,
+      "step": 3858
+    },
+    {
+      "epoch": 0.34302222222222223,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9987473284687273e-05,
+      "loss": 0.5422,
+      "step": 3859
+    },
+    {
+      "epoch": 0.3431111111111111,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9987421478792087e-05,
+      "loss": 0.5829,
+      "step": 3860
+    },
+    {
+      "epoch": 0.3432,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.998736956606018e-05,
+      "loss": 0.7097,
+      "step": 3861
+    },
+    {
+      "epoch": 0.34328888888888887,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.9987317546492102e-05,
+      "loss": 0.8506,
+      "step": 3862
+    },
+    {
+      "epoch": 0.3433777777777778,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.998726542008842e-05,
+      "loss": 0.8031,
+      "step": 3863
+    },
+    {
+      "epoch": 0.34346666666666664,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.9987213186849684e-05,
+      "loss": 0.7727,
+      "step": 3864
+    },
+    {
+      "epoch": 0.34355555555555556,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.998716084677646e-05,
+      "loss": 0.6564,
+      "step": 3865
+    },
+    {
+      "epoch": 0.3436444444444444,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9987108399869298e-05,
+      "loss": 0.7166,
+      "step": 3866
+    },
+    {
+      "epoch": 0.34373333333333334,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9987055846128767e-05,
+      "loss": 0.7365,
+      "step": 3867
+    },
+    {
+      "epoch": 0.3438222222222222,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.998700318555543e-05,
+      "loss": 0.8189,
+      "step": 3868
+    },
+    {
+      "epoch": 0.3439111111111111,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.9986950418149842e-05,
+      "loss": 0.613,
+      "step": 3869
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.998689754391257e-05,
+      "loss": 0.6677,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3440888888888889,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9986844562844188e-05,
+      "loss": 0.7773,
+      "step": 3871
+    },
+    {
+      "epoch": 0.3441777777777778,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.998679147494525e-05,
+      "loss": 0.6899,
+      "step": 3872
+    },
+    {
+      "epoch": 0.34426666666666667,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9986738280216334e-05,
+      "loss": 0.8521,
+      "step": 3873
+    },
+    {
+      "epoch": 0.3443555555555556,
+      "grad_norm": 2.75,
+      "learning_rate": 1.9986684978658007e-05,
+      "loss": 0.703,
+      "step": 3874
+    },
+    {
+      "epoch": 0.34444444444444444,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9986631570270835e-05,
+      "loss": 0.6838,
+      "step": 3875
+    },
+    {
+      "epoch": 0.34453333333333336,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.998657805505539e-05,
+      "loss": 0.6397,
+      "step": 3876
+    },
+    {
+      "epoch": 0.3446222222222222,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.998652443301225e-05,
+      "loss": 0.6755,
+      "step": 3877
+    },
+    {
+      "epoch": 0.34471111111111113,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9986470704141983e-05,
+      "loss": 0.7027,
+      "step": 3878
+    },
+    {
+      "epoch": 0.3448,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9986416868445167e-05,
+      "loss": 0.7283,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3448888888888889,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9986362925922376e-05,
+      "loss": 0.6761,
+      "step": 3880
+    },
+    {
+      "epoch": 0.34497777777777777,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.9986308876574188e-05,
+      "loss": 0.6308,
+      "step": 3881
+    },
+    {
+      "epoch": 0.3450666666666667,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.998625472040118e-05,
+      "loss": 0.787,
+      "step": 3882
+    },
+    {
+      "epoch": 0.34515555555555555,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9986200457403933e-05,
+      "loss": 0.7038,
+      "step": 3883
+    },
+    {
+      "epoch": 0.34524444444444446,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9986146087583025e-05,
+      "loss": 0.5002,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3453333333333333,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.998609161093904e-05,
+      "loss": 0.6719,
+      "step": 3885
+    },
+    {
+      "epoch": 0.34542222222222224,
+      "grad_norm": 2.125,
+      "learning_rate": 1.998603702747256e-05,
+      "loss": 0.6958,
+      "step": 3886
+    },
+    {
+      "epoch": 0.3455111111111111,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.998598233718417e-05,
+      "loss": 0.6263,
+      "step": 3887
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9985927540074453e-05,
+      "loss": 0.8622,
+      "step": 3888
+    },
+    {
+      "epoch": 0.3456888888888889,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9985872636143998e-05,
+      "loss": 0.7993,
+      "step": 3889
+    },
+    {
+      "epoch": 0.3457777777777778,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.9985817625393388e-05,
+      "loss": 0.7883,
+      "step": 3890
+    },
+    {
+      "epoch": 0.34586666666666666,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.998576250782321e-05,
+      "loss": 0.4934,
+      "step": 3891
+    },
+    {
+      "epoch": 0.34595555555555557,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9985707283434063e-05,
+      "loss": 0.6213,
+      "step": 3892
+    },
+    {
+      "epoch": 0.34604444444444443,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.998565195222653e-05,
+      "loss": 0.7843,
+      "step": 3893
+    },
+    {
+      "epoch": 0.34613333333333335,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.998559651420121e-05,
+      "loss": 0.6949,
+      "step": 3894
+    },
+    {
+      "epoch": 0.3462222222222222,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9985540969358684e-05,
+      "loss": 0.7575,
+      "step": 3895
+    },
+    {
+      "epoch": 0.3463111111111111,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9985485317699557e-05,
+      "loss": 0.5727,
+      "step": 3896
+    },
+    {
+      "epoch": 0.3464,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.998542955922442e-05,
+      "loss": 0.7813,
+      "step": 3897
+    },
+    {
+      "epoch": 0.3464888888888889,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9985373693933866e-05,
+      "loss": 0.6444,
+      "step": 3898
+    },
+    {
+      "epoch": 0.34657777777777776,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.99853177218285e-05,
+      "loss": 0.7325,
+      "step": 3899
+    },
+    {
+      "epoch": 0.3466666666666667,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9985261642908917e-05,
+      "loss": 0.6126,
+      "step": 3900
+    },
+    {
+      "epoch": 0.34675555555555554,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9985205457175716e-05,
+      "loss": 0.6512,
+      "step": 3901
+    },
+    {
+      "epoch": 0.34684444444444446,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.9985149164629503e-05,
+      "loss": 0.6649,
+      "step": 3902
+    },
+    {
+      "epoch": 0.3469333333333333,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9985092765270874e-05,
+      "loss": 0.6703,
+      "step": 3903
+    },
+    {
+      "epoch": 0.34702222222222223,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9985036259100434e-05,
+      "loss": 0.5444,
+      "step": 3904
+    },
+    {
+      "epoch": 0.3471111111111111,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.998497964611879e-05,
+      "loss": 0.4779,
+      "step": 3905
+    },
+    {
+      "epoch": 0.3472,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9984922926326543e-05,
+      "loss": 0.7524,
+      "step": 3906
+    },
+    {
+      "epoch": 0.34728888888888887,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.9984866099724305e-05,
+      "loss": 0.7717,
+      "step": 3907
+    },
+    {
+      "epoch": 0.3473777777777778,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.998480916631268e-05,
+      "loss": 0.71,
+      "step": 3908
+    },
+    {
+      "epoch": 0.34746666666666665,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.9984752126092276e-05,
+      "loss": 0.6727,
+      "step": 3909
+    },
+    {
+      "epoch": 0.34755555555555556,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.998469497906371e-05,
+      "loss": 0.7041,
+      "step": 3910
+    },
+    {
+      "epoch": 0.3476444444444444,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.998463772522759e-05,
+      "loss": 0.7319,
+      "step": 3911
+    },
+    {
+      "epoch": 0.34773333333333334,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9984580364584523e-05,
+      "loss": 0.5296,
+      "step": 3912
+    },
+    {
+      "epoch": 0.3478222222222222,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9984522897135127e-05,
+      "loss": 0.7264,
+      "step": 3913
+    },
+    {
+      "epoch": 0.3479111111111111,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.9984465322880017e-05,
+      "loss": 0.7011,
+      "step": 3914
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9984407641819812e-05,
+      "loss": 0.6734,
+      "step": 3915
+    },
+    {
+      "epoch": 0.3480888888888889,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9984349853955123e-05,
+      "loss": 0.7731,
+      "step": 3916
+    },
+    {
+      "epoch": 0.34817777777777775,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.998429195928657e-05,
+      "loss": 0.5958,
+      "step": 3917
+    },
+    {
+      "epoch": 0.34826666666666667,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9984233957814775e-05,
+      "loss": 0.7259,
+      "step": 3918
+    },
+    {
+      "epoch": 0.34835555555555553,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9984175849540354e-05,
+      "loss": 0.7067,
+      "step": 3919
+    },
+    {
+      "epoch": 0.34844444444444445,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9984117634463934e-05,
+      "loss": 0.6485,
+      "step": 3920
+    },
+    {
+      "epoch": 0.3485333333333333,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9984059312586134e-05,
+      "loss": 0.63,
+      "step": 3921
+    },
+    {
+      "epoch": 0.3486222222222222,
+      "grad_norm": 2.125,
+      "learning_rate": 1.998400088390758e-05,
+      "loss": 0.6849,
+      "step": 3922
+    },
+    {
+      "epoch": 0.34871111111111114,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9983942348428892e-05,
+      "loss": 0.6517,
+      "step": 3923
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 1.625,
+      "learning_rate": 1.99838837061507e-05,
+      "loss": 0.5432,
+      "step": 3924
+    },
+    {
+      "epoch": 0.3488888888888889,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.9983824957073632e-05,
+      "loss": 0.6829,
+      "step": 3925
+    },
+    {
+      "epoch": 0.3489777777777778,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9983766101198316e-05,
+      "loss": 0.7134,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3490666666666667,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.998370713852538e-05,
+      "loss": 0.7035,
+      "step": 3927
+    },
+    {
+      "epoch": 0.34915555555555555,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.998364806905546e-05,
+      "loss": 0.7514,
+      "step": 3928
+    },
+    {
+      "epoch": 0.34924444444444447,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.998358889278918e-05,
+      "loss": 0.8498,
+      "step": 3929
+    },
+    {
+      "epoch": 0.34933333333333333,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9983529609727176e-05,
+      "loss": 0.7575,
+      "step": 3930
+    },
+    {
+      "epoch": 0.34942222222222225,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9983470219870086e-05,
+      "loss": 0.7413,
+      "step": 3931
+    },
+    {
+      "epoch": 0.3495111111111111,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.9983410723218543e-05,
+      "loss": 0.7443,
+      "step": 3932
+    },
+    {
+      "epoch": 0.3496,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.998335111977318e-05,
+      "loss": 0.7476,
+      "step": 3933
+    },
+    {
+      "epoch": 0.3496888888888889,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9983291409534637e-05,
+      "loss": 0.7484,
+      "step": 3934
+    },
+    {
+      "epoch": 0.3497777777777778,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9983231592503555e-05,
+      "loss": 0.5856,
+      "step": 3935
+    },
+    {
+      "epoch": 0.34986666666666666,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.998317166868057e-05,
+      "loss": 0.7192,
+      "step": 3936
+    },
+    {
+      "epoch": 0.3499555555555556,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.998311163806633e-05,
+      "loss": 0.6192,
+      "step": 3937
+    },
+    {
+      "epoch": 0.35004444444444444,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9983051500661464e-05,
+      "loss": 0.7905,
+      "step": 3938
+    },
+    {
+      "epoch": 0.35013333333333335,
+      "grad_norm": 1.5,
+      "learning_rate": 1.998299125646663e-05,
+      "loss": 0.5434,
+      "step": 3939
+    },
+    {
+      "epoch": 0.3502222222222222,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9982930905482463e-05,
+      "loss": 0.6615,
+      "step": 3940
+    },
+    {
+      "epoch": 0.35031111111111113,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.998287044770961e-05,
+      "loss": 0.7554,
+      "step": 3941
+    },
+    {
+      "epoch": 0.3504,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.998280988314872e-05,
+      "loss": 0.6955,
+      "step": 3942
+    },
+    {
+      "epoch": 0.3504888888888889,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9982749211800444e-05,
+      "loss": 0.652,
+      "step": 3943
+    },
+    {
+      "epoch": 0.35057777777777777,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9982688433665424e-05,
+      "loss": 0.64,
+      "step": 3944
+    },
+    {
+      "epoch": 0.3506666666666667,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9982627548744313e-05,
+      "loss": 0.6283,
+      "step": 3945
+    },
+    {
+      "epoch": 0.35075555555555554,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9982566557037758e-05,
+      "loss": 0.5929,
+      "step": 3946
+    },
+    {
+      "epoch": 0.35084444444444446,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9982505458546423e-05,
+      "loss": 0.7852,
+      "step": 3947
+    },
+    {
+      "epoch": 0.3509333333333333,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.9982444253270953e-05,
+      "loss": 0.585,
+      "step": 3948
+    },
+    {
+      "epoch": 0.35102222222222224,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.9982382941212e-05,
+      "loss": 0.6984,
+      "step": 3949
+    },
+    {
+      "epoch": 0.3511111111111111,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.998232152237023e-05,
+      "loss": 0.7238,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3512,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.998225999674629e-05,
+      "loss": 0.858,
+      "step": 3951
+    },
+    {
+      "epoch": 0.3512888888888889,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9982198364340842e-05,
+      "loss": 0.615,
+      "step": 3952
+    },
+    {
+      "epoch": 0.3513777777777778,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9982136625154545e-05,
+      "loss": 0.68,
+      "step": 3953
+    },
+    {
+      "epoch": 0.35146666666666665,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9982074779188063e-05,
+      "loss": 0.8349,
+      "step": 3954
+    },
+    {
+      "epoch": 0.35155555555555557,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9982012826442053e-05,
+      "loss": 0.923,
+      "step": 3955
+    },
+    {
+      "epoch": 0.3516444444444444,
+      "grad_norm": 1.875,
+      "learning_rate": 1.998195076691718e-05,
+      "loss": 0.5721,
+      "step": 3956
+    },
+    {
+      "epoch": 0.35173333333333334,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9981888600614105e-05,
+      "loss": 0.6329,
+      "step": 3957
+    },
+    {
+      "epoch": 0.3518222222222222,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.99818263275335e-05,
+      "loss": 0.7623,
+      "step": 3958
+    },
+    {
+      "epoch": 0.3519111111111111,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9981763947676017e-05,
+      "loss": 0.6295,
+      "step": 3959
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.998170146104234e-05,
+      "loss": 0.555,
+      "step": 3960
+    },
+    {
+      "epoch": 0.3520888888888889,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.998163886763313e-05,
+      "loss": 0.8223,
+      "step": 3961
+    },
+    {
+      "epoch": 0.35217777777777776,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9981576167449052e-05,
+      "loss": 0.5646,
+      "step": 3962
+    },
+    {
+      "epoch": 0.3522666666666667,
+      "grad_norm": 2.375,
+      "learning_rate": 1.9981513360490786e-05,
+      "loss": 0.6968,
+      "step": 3963
+    },
+    {
+      "epoch": 0.35235555555555553,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.9981450446758995e-05,
+      "loss": 0.9002,
+      "step": 3964
+    },
+    {
+      "epoch": 0.35244444444444445,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.9981387426254358e-05,
+      "loss": 0.6273,
+      "step": 3965
+    },
+    {
+      "epoch": 0.3525333333333333,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9981324298977545e-05,
+      "loss": 0.653,
+      "step": 3966
+    },
+    {
+      "epoch": 0.3526222222222222,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.9981261064929237e-05,
+      "loss": 0.6284,
+      "step": 3967
+    },
+    {
+      "epoch": 0.3527111111111111,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9981197724110102e-05,
+      "loss": 0.8125,
+      "step": 3968
+    },
+    {
+      "epoch": 0.3528,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9981134276520828e-05,
+      "loss": 0.7998,
+      "step": 3969
+    },
+    {
+      "epoch": 0.35288888888888886,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9981070722162085e-05,
+      "loss": 0.7303,
+      "step": 3970
+    },
+    {
+      "epoch": 0.3529777777777778,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9981007061034558e-05,
+      "loss": 0.5873,
+      "step": 3971
+    },
+    {
+      "epoch": 0.35306666666666664,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.9980943293138924e-05,
+      "loss": 0.7231,
+      "step": 3972
+    },
+    {
+      "epoch": 0.35315555555555556,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.9980879418475868e-05,
+      "loss": 0.6543,
+      "step": 3973
+    },
+    {
+      "epoch": 0.35324444444444447,
+      "grad_norm": 2.375,
+      "learning_rate": 1.998081543704607e-05,
+      "loss": 0.7646,
+      "step": 3974
+    },
+    {
+      "epoch": 0.35333333333333333,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.998075134885022e-05,
+      "loss": 0.7613,
+      "step": 3975
+    },
+    {
+      "epoch": 0.35342222222222225,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.9980687153889e-05,
+      "loss": 0.7136,
+      "step": 3976
+    },
+    {
+      "epoch": 0.3535111111111111,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9980622852163097e-05,
+      "loss": 0.7459,
+      "step": 3977
+    },
+    {
+      "epoch": 0.3536,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.99805584436732e-05,
+      "loss": 0.6851,
+      "step": 3978
+    },
+    {
+      "epoch": 0.3536888888888889,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9980493928419997e-05,
+      "loss": 0.7237,
+      "step": 3979
+    },
+    {
+      "epoch": 0.3537777777777778,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.9980429306404176e-05,
+      "loss": 0.7429,
+      "step": 3980
+    },
+    {
+      "epoch": 0.35386666666666666,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.998036457762643e-05,
+      "loss": 0.5982,
+      "step": 3981
+    },
+    {
+      "epoch": 0.3539555555555556,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9980299742087452e-05,
+      "loss": 0.7025,
+      "step": 3982
+    },
+    {
+      "epoch": 0.35404444444444444,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.9980234799787935e-05,
+      "loss": 0.7298,
+      "step": 3983
+    },
+    {
+      "epoch": 0.35413333333333336,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.9980169750728575e-05,
+      "loss": 0.8605,
+      "step": 3984
+    },
+    {
+      "epoch": 0.3542222222222222,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.9980104594910066e-05,
+      "loss": 0.8796,
+      "step": 3985
+    },
+    {
+      "epoch": 0.35431111111111113,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.9980039332333104e-05,
+      "loss": 0.6475,
+      "step": 3986
+    },
+    {
+      "epoch": 0.3544,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.9979973962998396e-05,
+      "loss": 0.625,
+      "step": 3987
+    },
+    {
+      "epoch": 0.3544888888888889,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9979908486906627e-05,
+      "loss": 0.7357,
+      "step": 3988
+    },
+    {
+      "epoch": 0.35457777777777777,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9979842904058508e-05,
+      "loss": 0.9192,
+      "step": 3989
+    },
+    {
+      "epoch": 0.3546666666666667,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.9979777214454738e-05,
+      "loss": 0.7137,
+      "step": 3990
+    },
+    {
+      "epoch": 0.35475555555555555,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9979711418096016e-05,
+      "loss": 0.8496,
+      "step": 3991
+    },
+    {
+      "epoch": 0.35484444444444446,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.997964551498305e-05,
+      "loss": 0.7399,
+      "step": 3992
+    },
+    {
+      "epoch": 0.3549333333333333,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.997957950511654e-05,
+      "loss": 0.8683,
+      "step": 3993
+    },
+    {
+      "epoch": 0.35502222222222224,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.99795133884972e-05,
+      "loss": 0.6868,
+      "step": 3994
+    },
+    {
+      "epoch": 0.3551111111111111,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9979447165125735e-05,
+      "loss": 0.7039,
+      "step": 3995
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9979380835002846e-05,
+      "loss": 0.69,
+      "step": 3996
+    },
+    {
+      "epoch": 0.3552888888888889,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.997931439812925e-05,
+      "loss": 0.6695,
+      "step": 3997
+    },
+    {
+      "epoch": 0.3553777777777778,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9979247854505657e-05,
+      "loss": 0.6548,
+      "step": 3998
+    },
+    {
+      "epoch": 0.35546666666666665,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9979181204132777e-05,
+      "loss": 0.6563,
+      "step": 3999
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.9979114447011323e-05,
+      "loss": 0.7393,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 33750,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.15246892515328e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}