diff --git "a/checkpoint-1000/trainer_state.json" "b/checkpoint-1000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1000/trainer_state.json"
@@ -0,0 +1,7033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.08888888888888889,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "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
+    }
+  ],
+  "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": 1.28811723128832e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}